WAM2

Core Science & Biosynthesis

what is the WAM2layers model

An In-depth Technical Guide to the WAM2layers Model

Introduction

The Water Accounting Model - 2 layers (WAM2layers) is a sophisticated Eulerian atmospheric moisture tracking model designed for researchers and scientists in hydrology, meteorology, and climate science.[1][2] It provides a quantitative framework to understand the complete atmospheric branch of the hydrological cycle. The model's primary function is to trace the path of atmospheric moisture, either by identifying the ultimate destination of evaporated water (forward tracking) or by determining the original evaporative source of precipitation (backward tracking).[3][4][5] Developed to study atmospheric dynamics, rainfall patterns, and the impacts of climate and land-use change, WAM2layers has been utilized in over 60 academic studies.[3][6][7]

The model operates on gridded atmospheric datasets, such as reanalysis data or outputs from climate models, by solving a prognostic equation for "tagged" moisture.[3][6][8] Its evolution from a single-layer to a two-layer model significantly improved its accuracy, particularly in scenarios with strong directional wind shear.[6][9] The most recent major release, WAM2layers v3, is a complete rewrite of the original, incorporating performance optimizations and improved software sustainability practices to handle high-resolution input data.[3][6][7][8]

Core Principles and Governing Equations

WAM2layers is fundamentally a water accounting model that meticulously balances the moisture budget within a defined atmospheric column.[1][9] It operates on an offline basis, meaning it uses pre-existing data from atmospheric models rather than running concurrently with them.[2]

The core of the model is its two-layer vertical structure, which was introduced to better represent the complexity of atmospheric moisture transport. The single-layer predecessor was found to produce large errors in cases of strong directional wind shear, a limitation that the two-layer version effectively addresses.[9][10]

The model's logic is based on the principle of moisture conservation. For each grid cell in the model's domain, it balances the moisture that enters, leaves, evaporates, precipitates, and is exchanged between the two vertical layers.[1] The governing equations, as implemented in WAM2layers v3, can be summarized as follows:

• Total Moisture Budget: The model first ensures a balanced budget for the total atmospheric moisture (both tagged and untagged) in each grid cell. This involves accounting for all fluxes into and out of the cell (horizontal and vertical) as well as surface fluxes (evaporation and precipitation).[9]

• Tagged Moisture Budget: Once the total budget is balanced, the model solves a prognostic equation for the "tagged" moisture.[3][6] This equation tracks the fraction of total moisture in a grid cell that originated from (in backward tracking) or is destined for (in forward tracking) a specific source/sink region. The movement of tagged moisture is assumed to be proportional to the total moisture fluxes.[11]

The equations are solved numerically over a defined time period using gridded input data. The model is written in Python and is designed to be used as a command-line tool.[4]

Model Workflow and Tracking Modes

A typical WAM2layers experiment involves several key steps, from preparing the input data to running the tracking simulation and analyzing the output.

Caption: General experimental workflow in WAM2layers.

Tracking Modes

The model can be run in two primary modes depending on the research question.[5]

1. Backward Tracking: This mode is used to determine the evaporative sources of precipitation in a specific "tagging region". The model takes precipitation data from the target region as input and tracks the moisture backward in time to identify where it evaporated.[5]

Caption: Conceptual diagram of backward moisture tracking.

2. Forward Tracking: This mode is used to determine where moisture evaporated from a specific region eventually precipitates. The model takes evaporation data from the "tagging region" as input and tracks the moisture forward in time to map its subsequent precipitation locations.[5]

Caption: Conceptual diagram of forward moisture tracking.

Data Requirements

WAM2layers is an offline model and relies on external gridded atmospheric data. The quality and resolution of this input data are critical for the accuracy of the tracking results. The model has been adapted to work with several major reanalysis and climate model datasets.

Parameter Category	Required Variables	Common Data Sources	Typical Resolution
Atmospheric State	Specific Humidity, Surface Pressure, Total Column Water	ERA5, ERA-Interim, MERRA-2, Climate Model Output	1.5° x 1.5° (ERA-Interim) to higher resolutions[1]
Atmospheric Fluxes	Zonal Wind (u), Meridional Wind (v) at multiple levels	ERA5, ERA-Interim, MERRA-2, Climate Model Output	6-hourly (ERA-Interim)[1]
Surface Fluxes	Evaporation, Precipitation	ERA5, ERA-Interim, MERRA-2, Climate Model Output	3-hourly (ERA-Interim)[1]

Experimental Protocol: Tracking Moisture Sources for a Heatwave Event

This section outlines a detailed methodology for a typical research application of WAM2layers: identifying the terrestrial moisture sources that sustain precipitation during a European heatwave, a process known as moisture recycling.[1]

1. Objective: To quantify the contribution of different terrestrial and oceanic sources to the precipitation falling over a specific European region during a defined heatwave period.

2. Study Period and Region Definition:

• Heatwave Identification: Define heatwave periods using meteorological data (e.g., daily maximum temperature exceeding a certain percentile for a minimum number of consecutive days). For this example, 20 heatwave periods between 1979 and 2018 are selected.[1]

• Tagging Region: Define the geographical boundaries of the region of interest where precipitation will be tracked (e.g., Northern, Western, or Southern Europe).[1]

3. Input Data Preparation:

• Data Source: Utilize the ERA-Interim reanalysis dataset provided by the ECMWF.[1]

• Variable Extraction: Retrieve all required variables (see table above) at a 1.5° x 1.5° spatial resolution and 6-hourly temporal resolution (3-hourly for evaporation and precipitation).[1]

• Preprocessing: Run the WAM2layers preprocessing scripts. This step calculates the necessary states (e.g., water content in the upper and lower layers) and fluxes (e.g., horizontal and vertical transport between grid cells) from the raw reanalysis data and prepares them for the tracking algorithm.

4. Tracking Simulation:

• Mode Selection: Configure the model to run in backward tracking mode.

• Configuration: Create a configuration file specifying the start and end dates of the heatwave periods, the geographical coordinates of the tagging region (where precipitation is tracked), the paths to the preprocessed input data, and the desired output location.[5]

• Execution: Run the WAM2layers tracking script from the command line, referencing the configuration file. The model will ingest the precipitation data over the tagging region for each time step and calculate the corresponding evaporation source locations.[5]

5. Post-processing and Analysis:

• Output Aggregation: The model output will consist of gridded data showing the amount of evaporation from each grid cell that contributed to the precipitation in the tagging region. Aggregate these results over the entire heatwave period.

• Source Attribution: Analyze the aggregated output maps to identify the key terrestrial and oceanic source regions. Quantify the contribution from specific areas (e.g., forests, agricultural land, North Atlantic Ocean).[1]

• Comparison: Compare the moisture sources during the heatwave periods to climatological averages to identify anomalies in atmospheric transport patterns, such as a reduction in westerly moisture supply and an enhancement of local recycling.[1]

Model Versions and Adaptations

The official WAM2layers codebase has evolved, and several versions have been adapted by the research community to work with different input datasets.

Model Version	Adapted For	Key Characteristics
WAM2layers v1	ERA-Interim	The original two-layer version, an evolution from the single-layer WAM.[9]
WAM2layers v3	General Use	A complete rewrite in Python with performance optimizations for high-resolution data.[3][6][7]
Community Adaptations	EC-Earth, MERRA2, ERA5	Versions modified by various researchers to handle the specific data structures and levels of different climate models and reanalysis products.[12][13]

Assumptions and Future Development

While powerful, WAM2layers operates on a set of key assumptions. The model assumes that tagged moisture is well-mixed within each of the two atmospheric layers of a grid cell. The accuracy is also dependent on the quality of the input reanalysis data, which itself is a model-based product. Furthermore, using the model on a regular latitude-longitude grid near the poles (>80° N/S) can lead to numerical instability and unphysical results due to the violation of the Courant-criterion.[14]

Future development is aimed at further improving the model's user-friendliness, expanding support for more input datasets, and potentially adding new tracking capabilities such as tracking by distance or time.[5] An ongoing intercomparison study with other moisture tracking models will help to better understand the effects of different modeling decisions and assumptions.[9]

References

• 1. Moisture recycling and the potential role of forests as moisture source during European heatwaves - PMC [pmc.ncbi.nlm.nih.gov]
• 2. gmd.copernicus.org [gmd.copernicus.org]
• 3. doaj.org [doaj.org]
• 4. Welcome to WAM2layers! — WAM2layers documentation [wam2layers.readthedocs.io]
• 5. Tracking — WAM2layers documentation [wam2layers.readthedocs.io]
• 6. GMD - Atmospheric moisture tracking with WAM2layers v3 [gmd.copernicus.org]
• 7. researchgate.net [researchgate.net]
• 8. egusphere.copernicus.org [egusphere.copernicus.org]
• 9. research.tudelft.nl [research.tudelft.nl]
• 10. journals.ametsoc.org [journals.ametsoc.org]
• 11. Theory — WAM2layers documentation [wam2layers.readthedocs.io]
• 12. GitHub - WAM2layers/WAM2layers: Atmospheric moisture tracking [github.com]
• 13. GitHub - pkeys/WAM2layersPythonMerra2: Cite name of the model as: "WAM-2layers-MERRA2-v1.0-beta [github.com]
• 14. GitHub - ruudvdent/WAM2layersPython: This version still works, but all development has been moved to: https://github.com/WAM2layers/WAM2layers [github.com]

An In-depth Technical Guide to the Core Principles of Eulerian Moisture Tracking

For Researchers, Scientists, and Drug Development Professionals

This technical guide provides a comprehensive overview of the core principles of Eulerian moisture tracking, a fundamental method in atmospheric science for understanding the transport of water vapor throughout the Earth's atmosphere. The Eulerian approach offers a fixed-grid perspective on moisture movement, providing critical insights for climate modeling, weather forecasting, and related scientific disciplines. This document details the governing equations, numerical methods, and validation protocols essential for a thorough understanding of this methodology.

Fundamental Principles of Eulerian Moisture Tracking

The Eulerian approach to fluid dynamics examines fluid motion by observing changes in properties at fixed points in space. In the context of atmospheric moisture, this translates to tracking the concentration of water vapor within a defined grid system that covers a specific region or the entire globe. The core of this method lies in solving the continuity equation for water vapor, which is a statement of the conservation of mass.

The Governing Equation: Continuity of Water Vapor

The fundamental principle of Eulerian moisture tracking is encapsulated in the continuity equation for water vapor. This equation describes the rate of change of water vapor concentration at a fixed point as a function of the movement of water vapor into and out of that point (advection), as well as sources and sinks of moisture within that point.

The generalized form of the continuity equation for water vapor (q) is:

Where:

• ∂q/∂t represents the local rate of change of specific humidity (the mass of water vapor per unit mass of moist air) over time.

• ∇ ⋅ (vq) is the divergence of the moisture flux, which represents the net rate of transport of water vapor out of a given volume. Here, v is the three-dimensional wind vector. This term is central to the concept of advection.

• E represents the rate of evaporation, which acts as a source of atmospheric moisture.

• P represents the rate of precipitation (condensation and deposition), which acts as a sink of atmospheric moisture.

This equation forms the bedrock of Eulerian moisture tracking within atmospheric models.

The Atmospheric Water Vapor Budget

The continuity equation is integral to understanding the atmospheric water vapor budget. This budget provides a quantitative accounting of the inputs, outputs, and storage of water vapor in the atmosphere. Over a given region and time period, the change in atmospheric water storage is balanced by the net inflow of moisture from surrounding areas and the net difference between surface evaporation and precipitation.

The atmospheric water budget can be expressed as:

Where:

• ΔS is the change in atmospheric water vapor storage.

• Qin is the moisture flux into the region.

• Qout is the moisture flux out of the region.

• E is the total evaporation from the surface within the region.

• P is the total precipitation over the region.

Eulerian models are designed to solve the components of this budget for each grid cell in the model domain.

Numerical Methods for Eulerian Moisture Tracking

Solving the continuity equation for water vapor across a three-dimensional grid requires numerical methods to approximate the continuous partial differential equations. The choice of numerical scheme is critical as it affects the accuracy, stability, and computational cost of the model.

Discretization: The Foundation of Numerical Modeling

The first step in numerically solving the governing equations is discretization, which involves dividing the continuous domain of the atmosphere into a finite number of grid cells. The properties of the atmosphere, including water vapor concentration and wind velocity, are then calculated at the centers or boundaries of these cells. Common discretization methods include the Finite Volume Method, Finite Difference Method, and Spectral Methods.

The Finite Volume Method

The Finite Volume Method (FVM) is a widely used approach in atmospheric modeling due to its inherent conservation properties.[1] In FVM, the computational domain is divided into a set of control volumes, and the governing equations are integrated over each of these volumes.[1] The change in the average value of a variable within a control volume is determined by the net flux of that variable across its boundaries.[1]

Advection Schemes

A key challenge in Eulerian moisture tracking is accurately simulating the advection of water vapor by the wind. The choice of advection scheme determines how the movement of moisture between grid cells is calculated and can significantly impact the model's results.

A foundational and straightforward advection scheme is the first-order upwind scheme. This scheme determines the value of a property at a cell face based on the value in the cell "upwind" of the flow direction.[2][3][4] This approach ensures stability but can introduce numerical diffusion, which can artificially smooth out sharp gradients in moisture concentration.[2]

Formulation for a 1D Case:

Consider a one-dimensional grid with a constant velocity u in the positive x-direction. The flux of moisture (F) across the face i+1/2 between cells i and i+1 is given by:

Where q_i is the moisture concentration in cell i. The rate of change of moisture in cell i is then the difference between the flux into the cell and the flux out of the cell, divided by the volume of the cell.

While simple, the first-order upwind scheme provides a robust starting point for understanding more complex and accurate advection schemes used in modern atmospheric models, such as higher-order upwind schemes and semi-Lagrangian methods.[1][5]

Data Presentation: The Global Atmospheric Water Budget

The following table summarizes the key components of the global atmospheric water budget as estimated by various studies. These values represent the long-term annual mean fluxes and storage of atmospheric water vapor.

Component	Value (10^3 km³/year)	Source
Over Land
Precipitation	99 - 122	[1]
Evapotranspiration	52 - 97	[1]
Runoff	25 - 48	[1]
Over Oceans
Precipitation	242 - 412	[1]
Evaporation	273 - 458	[1]

Note: The range of values reflects the uncertainties inherent in different measurement and modeling techniques.

Experimental Protocols for Model Validation

Validating the performance of Eulerian moisture tracking models is a critical step to ensure their accuracy and reliability. A common approach involves comparing the model's output with observational data, reanalysis datasets, or the results from other established models, such as Lagrangian moisture tracking models.

Step-by-Step Validation Protocol

• Model Setup and Simulation:

  • Define the model domain, spatial resolution, and vertical levels.

  • Specify the initial and boundary conditions for meteorological fields (e.g., from reanalysis data like ERA5).[6]

  • Select the physical parameterization schemes for processes like convection and cloud microphysics.

  • Run the Eulerian model (e.g., WRF-WVTs) for a specified period to generate moisture transport and precipitation fields.[6][7]

• Acquisition of Validation Data:

  • Obtain observational precipitation data from ground-based gauges or satellite-derived products.

  • Acquire atmospheric moisture data from reanalysis datasets (e.g., ERA5, MERRA-2) which assimilate a wide range of observations.

  • If comparing with a Lagrangian model, run a Lagrangian model (e.g., FLEXPART) driven by the same meteorological fields as the Eulerian model.[6]

• Data Processing and Comparison:

  • Interpolate all datasets to a common grid for a consistent comparison.

  • Calculate relevant diagnostic variables, such as moisture flux convergence and vertically integrated water vapor.

  • Perform a point-to-point or regional comparison of key variables (e.g., precipitation, specific humidity) between the model output and the validation data.

• Evaluation Metrics:

  • Calculate statistical metrics to quantify the model's performance. Common metrics include:

    • Root Mean Square Error (RMSE): Measures the average magnitude of the errors.

    • Bias: Indicates whether the model consistently overestimates or underestimates the observed values.

    • Mean Absolute Error Skill Score (MAESS): Compares the model's performance to a reference forecast.[8] A higher MAESS indicates better performance.[8]

    • Correlation Coefficient: Measures the linear relationship between the modeled and observed values.

• Analysis and Interpretation:

  • Analyze the spatial and temporal patterns of errors and biases.

  • Identify potential sources of error in the model, such as inaccuracies in the input data, limitations of the physical parameterizations, or numerical artifacts.

  • Refine the model setup or parameterizations based on the validation results to improve its performance.

Mandatory Visualizations

The following diagrams illustrate key concepts and workflows in Eulerian moisture tracking.

References

• 1. iapjournals.ac.cn [iapjournals.ac.cn]
• 2. Upwind_scheme [chemeurope.com]
• 3. Numerical Methods for Engineers [leifh.folk.ntnu.no]
• 4. Upwind scheme - Wikipedia [en.wikipedia.org]
• 5. Upwind Schemes to Evaluate the Advection Term | McGraw-Hill Education - Access Engineering [accessengineeringlibrary.com]
• 6. ESD - Simple physics-based adjustments reconcile the results of Eulerian and Lagrangian techniques for moisture tracking in atmospheric rivers [esd.copernicus.org]
• 7. esd.copernicus.org [esd.copernicus.org]
• 8. Simple physics-based adjustments reconcile the results of Eulerian and Lagrangian techniques for moisture tracking in atmospheric rivers | Publications [ifisc.uib-csic.es]

WAM2layers: An In-depth Technical Guide to Atmospheric Moisture Transport Modeling

For Researchers, Scientists, and Climate Analysts

This technical guide provides a comprehensive overview of the WAM2layers model, a powerful open-source tool for atmospheric moisture transport analysis. This document delves into the core principles, experimental workflows, and technical specifications of the model, offering a detailed resource for researchers and scientists in atmospheric science, hydrology, and related fields. The content is structured to provide clarity on the model's application, from data input to the interpretation of results.

Core Principles of the WAM2layers Model

The Water Accounting Model - 2 layers (WAM2layers) is an Eulerian atmospheric moisture tracking model designed to understand the movement of water vapor through the atmosphere.[1][2][3] It operates on gridded atmospheric datasets, such as reanalysis data or climate model output, to trace the origin and fate of atmospheric moisture.[1][2][3] The fundamental principle behind WAM2layers is the conservation of moisture, where the model balances the moisture entering, leaving, and stored within each grid cell.

The model simplifies the complex vertical structure of the atmosphere into two distinct layers, a crucial feature that significantly improves computational efficiency compared to multi-layer models.[2] This two-layer approach was adopted to overcome the limitations of the original single-layer model, particularly in scenarios with strong directional wind shear.[2][3] WAM2layers can be operated in two primary modes:

• Forward Tracking: This mode tracks where evaporated moisture from a specific source region subsequently precipitates.[1][4][5]

• Backward Tracking: This mode determines the evaporative sources of precipitation that falls over a defined sink region.[1][4][5]

The latest major release, WAM2layers v3, represents a complete rewrite of the original model, incorporating performance optimizations to handle the increasing resolution of modern atmospheric datasets and to enhance user-friendliness and software sustainability.[1][2][6]

Governing Equations

The core of the WAM2layers model is a set of prognostic equations that describe the budget of both total and "tagged" moisture within each grid cell for both the upper and lower atmospheric layers. Tagged moisture refers to the water vapor originating from or destined for a specific region of interest. The fundamental equation for the change in tagged moisture over time in a grid cell is a function of horizontal and vertical fluxes, as well as sources (evaporation) and sinks (precipitation).

Data Requirements and Model Specifications

The accuracy and resolution of the WAM2layers model's output are intrinsically linked to the quality of the input data. The model is designed to be flexible with different gridded atmospheric datasets, with a primary focus on the ERA5 reanalysis dataset in its latest version.[1]

Input Data Specifications

The following table summarizes the typical input data required for a WAM2layers experiment, with a focus on ERA5 data.

Parameter	Description	Typical Spatial Resolution (ERA5)	Temporal Resolution (ERA5)	Data Level
U-wind component (u)	Zonal (east-west) component of wind velocity.	0.25° x 0.25°	Hourly	Pressure or Model Levels
V-wind component (v)	Meridional (north-south) component of wind velocity.	0.25° x 0.25°	Hourly	Pressure or Model Levels
Specific humidity (q)	The mass of water vapor per unit mass of moist air.	0.25° x 0.25°	Hourly	Pressure or Model Levels
Surface pressure (ps)	Atmospheric pressure at the surface of the Earth.	0.25° x 0.25°	Hourly	Surface Level
Evaporation (e)	The flux of water vapor from the Earth's surface to the atmosphere.	0.25° x 0.25°	Hourly	Surface Level
Precipitation (p)	The flux of water (in liquid or solid form) from the atmosphere to the Earth's surface.	0.25° x 0.25°	Hourly	Surface Level
Total column water vapor	The total amount of water vapor in a column of air from the surface to the top of the atmosphere.	0.25° x 0.25°	Hourly	Single Level

Note: While ERA5 is the primary focus for WAM2layers v3, other datasets like ERA-Interim and MERRA2 have been used with previous versions.[6]

Model Configuration and Assumptions

The configuration of a WAM2layers experiment involves defining a set of parameters that govern the simulation. The key assumptions of the model are outlined in the table below.

Parameter/Assumption	Description	Rationale
Two-layer vertical structure	The atmosphere is divided into an upper and a lower layer.	Balances computational efficiency with the need to represent vertical wind shear.[2]
Eulerian framework	The model calculates fluxes of moisture through fixed grid cells.	Computationally efficient for tracking moisture over large areas and long periods.
Well-mixed atmosphere	Within each grid cell, tagged and untagged moisture are assumed to be perfectly mixed.	A simplifying assumption necessary for the Eulerian approach.
Offline model	WAM2layers uses pre-existing atmospheric data and does not dynamically interact with a climate model.	Allows for rapid analysis of different scenarios and time periods without the need for computationally expensive climate model runs.
Neglect of liquid and ice water	The model primarily tracks water in its vapor phase.	Simplifies the physics and reduces data requirements, though this can be a source of error.

Experimental Protocols

The execution of a WAM2layers experiment follows a structured, two-step process: preprocessing and tracking .[1][3] This workflow is consistent for both forward and backward tracking experiments.

Preprocessing Protocol

The preprocessing step is designed to prepare the input atmospheric data for the tracking algorithm. This involves collapsing the multi-level atmospheric data into the two-layer vertical grid of the model and calculating the necessary fluxes.

• Data Acquisition: Obtain the required atmospheric data (see Table 1) for the desired temporal and spatial domain from a data repository such as the Copernicus Climate Data Store for ERA5.[7]

• Vertical Integration: Vertically integrate the high-resolution atmospheric data (e.g., wind components and specific humidity on multiple pressure or model levels) into the two model layers.

• Flux Calculation: Calculate the horizontal moisture fluxes for each of the two layers and the vertical flux between them.

• Output Generation: Store the preprocessed data, which includes the moisture content of each layer and the fluxes between grid cells, in a format ready for the tracking step.

Tracking Protocol

The tracking step utilizes the preprocessed data to trace the movement of tagged moisture through the atmosphere over time.

• Configuration: Define the parameters of the tracking experiment in a configuration file. This includes:

  • Tracking direction: Forward or backward.

  • Tagging region: The geographical area of interest for the source (forward) or sink (backward) of moisture.

  • Time period: The start and end dates for the tracking simulation.

  • Input data path: The location of the preprocessed data.

  • Output data path: The directory where the tracking results will be saved.

• Model Execution: Run the WAM2layers tracking script, which iteratively calculates the movement of tagged moisture from one time step to the next based on the preprocessed fluxes.

• Output Analysis: The model outputs gridded data that quantifies the amount of tracked moisture that precipitates (forward) or evaporates (backward) in each grid cell. This data can then be visualized and analyzed to identify moisture pathways and source-sink relationships.

Visualization of Workflows and Logical Relationships

The following diagrams, generated using the DOT language, illustrate the key workflows and logical relationships within the WAM2layers model.

Core logical flow of the WAM2layers model.

References

• 1. research.tudelft.nl [research.tudelft.nl]
• 2. GMD - Atmospheric moisture tracking with WAM2layers v3 [gmd.copernicus.org]
• 3. egusphere.copernicus.org [egusphere.copernicus.org]
• 4. WAM2layers - WUR [wur.nl]
• 5. Tracking — WAM2layers documentation [wam2layers.readthedocs.io]
• 6. researchgate.net [researchgate.net]
• 7. ERA5 hourly data on single levels from 1940 to present [cds.climate.copernicus.eu]

A Technical Guide to the Atmospheric Water Cycle with WAM-2layers

This in-depth technical guide provides a comprehensive overview of the Water Accounting Model-2layers (WAM-2layers), a crucial tool for researchers and scientists in atmospheric science and hydrology. The guide details the model's core principles, experimental workflows, and data outputs, offering a complete framework for understanding and utilizing this powerful atmospheric moisture tracking model. While the primary audience is environmental and atmospheric scientists, the methodologies of data handling, simulation, and systematic analysis may be of interest to professionals in other data-intensive scientific fields.

Core Principles of WAM-2layers

The WAM-2layers model is an Eulerian tool designed to track the movement of atmospheric moisture across the globe. By solving the atmospheric water balance equation for "tagged" moisture in gridded datasets, it allows scientists to investigate the complete lifecycle of precipitation and evaporation.[1] The model operates by dividing the atmosphere into two distinct layers to provide a more accurate representation of moisture transport, especially in areas with significant wind shear, a limitation of earlier single-layer models.[2][3]

WAM-2layers can be run in two primary modes:[2][4][5]

• Backward Tracking: This mode identifies the evaporative sources of precipitation that falls over a predefined region (a "tagging region"). Researchers can use this to answer questions like, "Where did the water for this specific rainfall event come from?"[6][7]

• Forward Tracking: Conversely, this mode tracks where moisture evaporated from a specific region eventually precipitates. This is useful for understanding the downwind impacts of land-use changes (e.g., deforestation or irrigation) on rainfall patterns.[4][5][6]

The model is driven by meteorological reanalysis data, such as ERA5 from the European Centre for Medium-Range Weather Forecasts (ECMWF), which provides the necessary atmospheric variables like wind, humidity, and pressure.[2][8]

WAM-2layers Experimental Workflow

The operational workflow of a WAM-2layers experiment is a structured, three-stage process executed via a command-line interface. The entire process is controlled by a central configuration file (config.yaml) that specifies all parameters for the experiment, ensuring reproducibility.[2][5]

References

• 1. community.data.4tu.nl [community.data.4tu.nl]
• 2. research.tudelft.nl [research.tudelft.nl]
• 3. GMD - Atmospheric moisture tracking with WAM2layers v3 [gmd.copernicus.org]
• 4. researchgate.net [researchgate.net]
• 5. Tracking — WAM2layers documentation [wam2layers.readthedocs.io]
• 6. researchgate.net [researchgate.net]
• 7. GitHub - WAM2layers/WAM2layers: Atmospheric moisture tracking [github.com]
• 8. Moisture recycling and the potential role of forests as moisture source during European heatwaves - PMC [pmc.ncbi.nlm.nih.gov]

WAM2layers for Climate Change Studies: A Technical Guide

Audience: Researchers, Scientists, and Climate Analysts

This technical guide provides an in-depth overview of the Water Accounting Model - 2 layers (WAM2layers), a powerful Eulerian moisture tracking model used to investigate the atmospheric branch of the hydrological cycle. In the context of climate change, WAM2layers is an essential tool for understanding how alterations in atmospheric dynamics and land use impact precipitation patterns, drought, and extreme weather events by tracing moisture from its evaporative source to its precipitation sink.[1][2][3]

Core Principles of WAM2layers

WAM2layers is an offline, Eulerian atmospheric moisture tracking model that solves a prognostic equation for "tagged" moisture within gridded atmospheric datasets, such as reanalysis data (e.g., ERA5) or climate model outputs.[1][2] The fundamental goal is to account for all water entering and leaving a defined atmospheric column over a grid cell.

The model operates in two primary modes:

• Backward Tracking: Identifies the evaporative sources of precipitation that falls in a specific target region (the "tagging region").[1][4][5] This is crucial for answering questions like, "Where did the moisture for this extreme rainfall event come from?"

• Forward Tracking: Determines where evaporated or transpired moisture from a specific source region ultimately precipitates.[1][4][5] This helps in assessing the downwind impacts of land-use changes, such as deforestation or irrigation projects.

The "2layers" in its name refers to the model's vertical structure, which divides the atmospheric column into an upper and a lower layer to better account for vertical wind shear, a significant improvement over single-layer models.[4] This structure is based on the principle of moisture conservation, balancing all fluxes—horizontal transport by wind, surface evaporation, precipitation, and the exchange between the two layers.[6]

Conceptual Model of WAM2layers

The diagram below illustrates the core logic of the WAM2layers model for a single grid cell. It shows the division of the atmosphere into two layers and the various moisture fluxes that the model accounts for. Horizontal fluxes represent moisture transported by wind between adjacent grid cells, while vertical fluxes include evaporation from the surface into the lower layer, precipitation, and the exchange of moisture between the layers.

Conceptual diagram of the WAM2layers two-layer atmospheric moisture balance.

Experimental Protocols & Methodology

Conducting a climate study with WAM2layers involves a systematic workflow, from data acquisition to the final analysis of moisture transport patterns.

Data Input Requirements

The model requires gridded atmospheric data, typically from reanalysis products like ERA5, which provide a comprehensive, physically consistent picture of the global atmosphere. The necessary input variables are summarized in the table below.

Parameter	Variable Name (Typical)	Temporal Resolution	Description
Specific Humidity	q	6-hourly	Mass of water vapor per unit mass of moist air at various pressure levels.
Zonal Wind	u	6-hourly	East-west component of wind velocity at various pressure levels.
Meridional Wind	v	6-hourly	North-south component of wind velocity at various pressure levels.
Surface Pressure	sp	6-hourly	Atmospheric pressure at the surface of the grid cell.
Precipitation	tp	3-hourly (accumulated)	Total precipitation (rain and snow) falling on the surface.
Evaporation	e	3-hourly (accumulated)	Total moisture evaporated from the surface into the atmosphere.
Vertically Integrated Moisture Flux	viwvd	6-hourly	Total water vapor transport integrated over the atmospheric column.

Experimental Workflow

The following diagram outlines the logical workflow for a typical backward tracking experiment designed to identify the moisture sources for a specific region and time period (e.g., an extreme precipitation event).

Typical experimental workflow for a WAM2layers backward tracking study.

Methodological Steps:

• Define Scope: The researcher first defines the target region (tagging region) and the time period of interest for the analysis.

• Data Acquisition: Relevant atmospheric variables are downloaded from a reanalysis dataset like ERA5.

• Preprocessing: The 3D atmospheric data (e.g., specific humidity and wind on multiple pressure levels) are vertically integrated into the two layers of the model. This is a critical step that prepares the input data for the core tracking algorithm.

• Model Execution: WAM2layers is run. In a backward experiment, the model calculates how much of the precipitation in the tagging region can be attributed to moisture present in each grid cell of the domain at previous time steps.

• Post-processing: The model output, which represents tracked moisture, is used to calculate the originating evaporation. This step connects the atmospheric moisture back to its surface source.

• Analysis and Visualization: The results are analyzed to create maps showing the spatial distribution of moisture sources and to quantify the percentage contribution from different geographical areas.

Quantitative Data Presentation: A Case Study

WAM2layers enables the precise quantification of moisture sources. The results of these analyses are often summarized to compare the influence of different regions. The table below, adapted from a study on precipitation over the Tibetan Plateau, demonstrates how WAM2layers results can be presented.

Table 1: Annual Mean Moisture Source Contributions for Precipitation over the Tibetan Plateau (TP)

Target Region	Source Region	Contribution (%) - Observation Data	Contribution (%) - ERA-Suite Data
Northern TP (NTP)	Northwest (NW) of TP	38.9%	42.6%
Southeast (SE) of TP	17.9%	16.3%
Tibetan Plateau (TP)	26.1%	29.6%
Local (NTP) Recycling	12.9%	16.6%
Southern TP (STP)	Northwest (NW) of TP	16.9%	18.3%
Southeast (SE) of TP	54.2%	54.2%
Tibetan Plateau (TP)	16.2%	13.9%
Local (STP) Recycling	11.9%	9.8%
Source: Adapted from "Moisture Source Changes Contributed to Different Precipitation Changes over the Northern and Southern Tibetan Plateau," Journal of Climate, 2019.[7]

This table clearly shows that for the Northern Tibetan Plateau, the dominant moisture source is the region to the northwest, influenced by the westerlies.[7] In contrast, the Southern Tibetan Plateau receives the majority of its moisture from the southeast, associated with monsoon systems.[7] Such quantitative insights are vital for predicting how changes in these large-scale circulation patterns due to climate change will affect water resources in this critical region.

Conclusion and Climate Change Applications

WAM2layers is a critical tool for hydrometeorological and climate change research. Its ability to quantitatively link precipitation to its evaporative sources provides invaluable insights into the workings of the climate system.[1][8] Key applications in climate change studies include:

• Understanding Extreme Events: Tracing the moisture sources of atmospheric rivers and extreme rainfall events to determine how warming oceans or land surfaces may be intensifying them.[9]

• Assessing Land-Use Change Impacts: Quantifying how deforestation in one region (e.g., the Amazon) might reduce rainfall in a downwind agricultural area through reduced moisture recycling.[10]

• Analyzing Drought: Investigating whether droughts are caused by a lack of local evaporation or a shift in long-range moisture transport patterns.

• Improving Climate Models: Providing a method to validate the representation of the atmospheric water cycle in global and regional climate models.[9]

As climate change continues to alter the global hydrological cycle, tools like WAM2layers will become increasingly essential for researchers and scientists to diagnose the causes of regional climate shifts and predict future changes in water availability.

References

• 1. hess.copernicus.org [hess.copernicus.org]
• 2. mdpi.com [mdpi.com]
• 3. researchgate.net [researchgate.net]
• 4. researchgate.net [researchgate.net]
• 5. cw3e.ucsd.edu [cw3e.ucsd.edu]
• 6. researchgate.net [researchgate.net]
• 7. research-portal.uu.nl [research-portal.uu.nl]
• 8. research.tudelft.nl [research.tudelft.nl]
• 9. osti.gov [osti.gov]
• 10. acp.copernicus.org [acp.copernicus.org]

The WAM2layers Model: A Technical Guide to its History, Development, and Core Principles

Introduction

The Water Accounting Model - 2 layers (WAM2layers) is a powerful Eulerian atmospheric moisture tracking model that has become a vital tool for researchers and scientists in hydrology, meteorology, and climate science. It is designed to identify the sources and sinks of atmospheric moisture, enabling a deeper understanding of precipitation patterns, extreme weather events, and the intricate connections within the global water cycle. This in-depth technical guide provides a comprehensive overview of the history, development, core principles, and practical application of the WAM2layers model.

History and Development

The WAM2layers model has undergone significant evolution since its inception, driven by the need for more accurate and efficient atmospheric moisture tracking.[1][2][3]

The Genesis: WAM v1

The foundation of WAM2layers was laid with the single-layer Water Accounting Model (WAM v1), introduced in 2010.[2] This initial version was developed to track atmospheric moisture using data from the ERA-Interim reanalysis dataset.[2]

The Evolution to Two Layers: WAM2layers v1

A key limitation of the single-layer model was its susceptibility to errors in areas with strong directional wind shear. To address this, the model was updated to a two-layer version, which significantly improved its accuracy.[2] This version, referred to as WAM2layers v1, was further adapted for use with global climate and reanalysis data.[2]

A Rewrite for Modern Research: WAM2layers v2 and v3

The second version of the model, WAM2layers v2, marked a significant technological shift, with the codebase being completely rewritten from MATLAB to Python. This change enhanced the model's accessibility and usability within the scientific community.

The most recent major release, WAM2layers v3, represents another complete rewrite of the model.[1][2][3] This version was engineered to optimize performance and handle the increasing resolution of modern input datasets like ERA5.[1][2][3] It also incorporates best practices in software development to improve user-friendliness and the sustainability of the software.[1][2][3]

Version	Key Developments	Primary Input Data
WAM v1	Initial single-layer model.	ERA-Interim
WAM2layers v1	Introduction of a two-layer system to account for wind shear.	ERA-Interim, Global Climate Models
WAM2layers v2	Complete rewrite from MATLAB to Python.	Various
WAM2layers v3	Complete rewrite for performance optimization and user-friendliness.	ERA5, various

Core Principles and Formulation

WAM2layers is an offline, Eulerian atmospheric moisture tracking model.[2] It operates on gridded atmospheric datasets and solves a prognostic equation for "tagged" moisture to determine its origin or destination.[1][2][3]

The model can be run in two primary modes:

• Backward Tracking: This mode is used to determine the evaporative sources of precipitation in a specific region.

• Forward Tracking: This mode is used to track where evaporated water from a particular area eventually precipitates.

The core of the model is a set of governing equations that describe the conservation of total and tagged moisture within two atmospheric layers. This two-layer approach is a critical feature of the model, allowing for a more accurate representation of atmospheric transport, especially in the presence of wind shear.[2]

WAM2layers Workflow

The process of conducting a moisture tracking experiment with WAM2layers can be broken down into three main stages: pre-processing, tracking, and post-processing.

A high-level overview of the WAM2layers experimental workflow.

Experimental Protocols

The following provides a generalized methodology for conducting a moisture tracking experiment using WAM2layers v3.

Installation and Setup

The WAM2layers model is available as a Python package and can be installed using pip. It is recommended to set up a dedicated virtual environment for the installation.

The user documentation also provides a quickstart guide that includes the automatic download of sample data to run example cases.

Data Pre-processing

The first step in any WAM2layers experiment is to prepare the input data. WAM2layers v3 has a dedicated pre-processing module that collapses the input data onto a two-layer grid. This module is designed to be dataset-dependent, with a primary focus on ERA5 reanalysis data.

The pre-processing step reconstructs the terms of the moisture budget equation from the available input data.

Configuration

The user defines the specifics of the tracking experiment in a YAML configuration file. This file includes parameters such as:

• Tracking direction: forward or backward

• Date range: The start and end dates for the analysis.

• Tagging region: The geographical area of interest for which to track moisture sources or sinks.

• Paths: The locations of the pre-processed input data and the desired output directory.

Tracking

With the pre-processed data and the configuration file in place, the tracking experiment is initiated through the command line.

The tracking module then integrates the prognostic equation for tagged moisture numerically.

Post-processing and Analysis

The output of the tracking module is a dataset containing the tracked moisture. This data can then be analyzed and visualized using various tools to, for example, create maps of moisture sources for a particular precipitation event.

Data Presentation

The following table summarizes the typical input data requirements for a WAM2layers experiment using ERA5 reanalysis data.

Parameter	Description
u and v wind components	Zonal and meridional wind components at various pressure levels.
q specific humidity	Specific humidity at various pressure levels.
sp surface pressure	Surface pressure.
evap evaporation	Surface evaporation.
precip precipitation	Total precipitation.
tcwv total column water vapor	Total column water vapor.

Logical Relationships in WAM2layers

The core of the WAM2layers model is the interaction between the two atmospheric layers and the surface, governed by the principles of moisture conservation. The following diagram illustrates these fundamental relationships.

References

• 1. GitHub - WAM2layers/WAM2layers: Atmospheric moisture tracking [github.com]
• 2. research.tudelft.nl [research.tudelft.nl]
• 3. GMD - Atmospheric moisture tracking with WAM2layers v3 [gmd.copernicus.org]

Unveiling the Foundation: A Technical Guide to the Theoretical Basis of Water Accounting Models

For Researchers, Scientists, and Drug Development Professionals

This technical guide provides an in-depth exploration of the theoretical underpinnings of Water Accounting Models. It is designed for a scientific audience, offering a comprehensive overview of the core principles, methodologies, and frameworks that form the basis of modern water resource assessment.

Core Principles: The Water Balance Equation

The fundamental theoretical basis of all water accounting models is the principle of conservation of mass, applied through the water balance equation . This equation stipulates that for any given spatial domain (e.g., a river basin) and time period, the total water inflows must equal the total water outflows plus the change in water storage.

The generalized water balance equation can be expressed as:

P + Q_in - ET - Q_out - ΔS = 0

Where:

• P = Precipitation (rain, snow, etc.)

• Q_in = Surface and subsurface water inflow

• ET = Evapotranspiration (the sum of evaporation from soil and water surfaces and transpiration from plants)

• Q_out = Surface and subsurface water outflow

• ΔS = Change in water storage (in soil, groundwater, reservoirs, snow, and glaciers)

This equation provides the foundational logic for tracking the movement and availability of water within a defined system.

Key Water Accounting Frameworks

Two prominent frameworks have emerged to standardize and operationalize the principles of water accounting: Water Accounting Plus (WA+) and the System of Environmental-Economic Accounting for Water (SEEA-Water) .

Water Accounting Plus (WA+)

Developed by the International Water Management Institute (IWMI) and its partners, WA+ is a framework designed to provide explicit spatial information on water depletion and net withdrawal processes in complex river basins.[1] A key feature of WA+ is its extensive use of publicly available remote sensing data to overcome the challenge of data scarcity in many regions.[2]

The WA+ framework is structured around four main sheets:

• Resource Base Sheet: This sheet provides a summary of the overall water resources situation, including precipitation, outflows, and changes in storage.

• Evapotranspiration Sheet: This sheet details the water depletion through evapotranspiration, distinguishing between beneficial and non-beneficial consumption.

• Productivity Sheet: This sheet assesses the benefits derived from water consumption, often in terms of agricultural biomass production.

• Withdrawal Sheet: This sheet quantifies the gross and net water withdrawals by different sectors.

System of Environmental-Economic Accounting for Water (SEEA-Water)

The SEEA-Water is a statistical framework developed by the United Nations to integrate hydrological and economic information. It provides a standardized system for organizing water-related data in a way that is consistent with the System of National Accounts, allowing for the analysis of the relationship between water and the economy.

The core of SEEA-Water is the Physical Supply and Use Tables (PSUT) . These tables track the flows of water from the environment into the economy, within the economy, and back to the environment.

Methodologies and Data Sources

Water accounting models rely on a variety of data sources and methodologies to quantify the components of the water balance.

Experimental Protocols: Remote Sensing for Water Accounting

A significant advancement in water accounting has been the integration of remote sensing data, which provides spatially and temporally consistent information.

Protocol for Estimating Evapotranspiration (ET) using Remote Sensing:

One of the most critical and challenging components to quantify is evapotranspiration. Surface energy balance models are commonly used for this purpose. Two widely applied algorithms are:

• Surface Energy Balance Algorithm for Land (SEBAL): SEBAL estimates ET by solving the surface energy balance equation for each pixel of a satellite image. It uses data on surface temperature, albedo, and vegetation indices derived from satellite sensors like MODIS and Landsat.

• Mapping Evapotranspiration with Internalized Calibration (METRIC): METRIC is a variant of SEBAL that uses a ground-based reference crop evapotranspiration to calibrate the sensible heat flux computation, which can improve accuracy.

Data Inputs for Remote Sensing-based Water Accounting:

Data Component	Remote Sensing Product Example
Precipitation	Tropical Rainfall Measuring Mission (TRMM), Global Precipitation Measurement (GPM)
Evapotranspiration	Moderate Resolution Imaging Spectroradiometer (MODIS) thermal and optical data
Land Use/Land Cover	MODIS Land Cover Type Product (MCD12Q1)
Snow Cover	MODIS Snow Cover Products (MOD10A1)
Surface Water Extent	Landsat, Sentinel-2

Hydrological Modeling

Hydrological models are often used in conjunction with remote sensing data to simulate water flows and storage changes. These models can range from simple conceptual models to complex, physically-based distributed models.

Quantitative Data Summary: A Case Study of the Indus Basin (2007)

The following tables summarize the quantitative data from a Water Accounting Plus (WA+) study of the Indus Basin for the year 2007.[1][3] This provides a concrete example of the outputs of a water accounting model.

Table 1: Annual Water Balance of the Indus Basin (2007)

Component	Volume (km³)
Inflows
Precipitation	482
Outflows
Total Water Depletion	501
Basin Outflow	21
Change in Storage
Groundwater Storage	-30
Surface Water Storage	-9
Glacier and Snow Storage	-2

Table 2: Water Depletion by Source and Type in the Indus Basin (2007)

Depletion Source/Type	Volume (km³)	Percentage of Total Depletion
Landscape ET (from rainfall)	344	69%
Incremental ET (from utilized flow)	157	31%
Total Depletion	501	100%

Table 3: Agricultural Water Depletion in the Indus Basin (2007)

Agricultural System	Volume (km³)	Percentage of Total Agricultural Depletion
Irrigated Agriculture	254	85%
Rainfed Systems	44	15%
Total Agricultural Depletion	298	100%

Visualizing the Theoretical Frameworks

Diagrams are essential for understanding the logical relationships and workflows within water accounting models.

The Water Balance Conceptual Diagram

Caption: Conceptual flow of the water balance equation.

Water Accounting Plus (WA+) Framework Workflow

Caption: Workflow of the Water Accounting Plus (WA+) framework.

SEEA-Water Physical Supply and Use Table Logic

Caption: Logical flows in the SEEA-Water Physical Supply and Use Tables.

References

• 1. hess.copernicus.org [hess.copernicus.org]
• 2. iwmi.org [iwmi.org]
• 3. unstats.un.org [unstats.un.org]

WAM2layers v3: A Technical Deep Dive for Scientific Modeling

An in-depth technical guide for researchers, scientists, and drug development professionals on the core features of the WAM2layers v3 atmospheric moisture tracking model.

WAM2layers v3 is a powerful, open-source tool for atmospheric moisture tracking, enabling researchers to understand the sources, sinks, and pathways of atmospheric water vapor.[1][2][3] This guide provides a technical overview of its key features, data handling, and operational workflows, with a focus on aspects relevant to scientific modeling and data analysis professionals.

Core Functionalities and Key Features

WAM2layers v3 is a complete rewrite of its predecessors, designed for enhanced performance, user-friendliness, and reproducibility.[1][2] It operates as an offline, Eulerian atmospheric-moisture-tracking model that solves a prognostic equation for tagged moisture using gridded atmospheric datasets.[1][2][4]

The model offers two primary modes of operation:

• Forward Tracking: Determines the eventual destination (precipitation) of evaporated moisture from a defined source region.[1][3][5][6]

• Backward Tracking: Identifies the original evaporative source of precipitation in a specific area.[1][3][5][6]

A key innovation in v3 is its implementation as an installable command-line program, configured via a simple text file. This approach enhances the reproducibility of experiments.[4][7]

Data Presentation: Quantitative Overview

While specific performance benchmarks can vary based on hardware and dataset size, the following tables summarize key quantitative aspects of WAM2layers v3.

Feature	Specification	Description
Primary Input Data	ERA5 Reanalysis	The model is optimized for use with the ERA5 dataset from the European Centre for Medium-Range Weather Forecasts (ECMWF).[1]
Temporal Resolution	User-defined	The internal time step of the model is configurable by the user.[4][7]
Spatial Resolution	Dependent on input data	The model operates on the grid of the input atmospheric dataset.
Output Data Format	NetCDF	Output files are generated in the NetCDF format, a standard for scientific data.

Parameter	Data Type	Description
tracking_direction	String	forward or backward
input_folder	String	Path to the preprocessed input data.
output_folder	String	Path to store the output files.
start_date	Date	Start date for the tracking experiment.
end_date	Date	End date for the tracking experiment.
tagging_region	Bounding Box or Raster	Defines the geographical area for tagging evaporation (forward) or precipitation (backward).[7]

Experimental Protocols

The following protocols outline the typical workflows for conducting experiments with WAM2layers v3.

Protocol 1: Forward Tracking of Evaporation from a Source Region

Objective: To determine where moisture evaporated from a specific region ultimately precipitates.

Methodology:

• Installation: Install WAM2layers v3 from the Python Package Index (PyPI) using pip.

• Configuration: Create a configuration file specifying the following:

  • tracking_direction: forward

  • input_folder: Path to the preprocessed ERA5 data.

  • output_folder: A dedicated directory for the experiment's output.

  • start_date and end_date: The time period for the analysis.

  • tagging_region: The geographical coordinates of the evaporation source region.

• Preprocessing: Execute the wam2layers preprocess command, pointing to the raw input data and the configuration file. This step collapses the input data onto a two-layer grid.[1][4][7]

• Tracking: Run the wam2layers track command with the configuration file. This initiates the forward tracking of tagged moisture.

• Visualization: Use the wam2layers visualize command to generate preliminary plots of the output, or load the NetCDF output files into a scientific visualization tool for in-depth analysis.

Protocol 2: Backward Tracking of Precipitation to its Source

Objective: To identify the evaporative sources of precipitation for a given region.

Methodology:

• Installation: As per Protocol 1.

• Configuration: Create a configuration file with the following key settings:

  • tracking_direction: backward

  • tagging_region: The geographical coordinates of the precipitation sink region.

  • Other parameters (input_folder, output_folder, start_date, end_date) are set as in Protocol 1.

• Preprocessing: Identical to the preprocessing step in Protocol 1.

• Tracking: Execute the wam2layers track command with the backward tracking configuration.

• Visualization: Analyze the output to identify and quantify the contribution of different source regions to the precipitation in the tagged area.

Mandatory Visualization

Signaling Pathways and Workflows

The following diagrams illustrate the logical flow and experimental workflow of WAM2layers v3.

Logical flow of data within the WAM2layers v3 model.

A typical experimental workflow using WAM2layers v3.

References

• 1. gmd.copernicus.org [gmd.copernicus.org]
• 2. egusphere.copernicus.org [egusphere.copernicus.org]
• 3. Theory — WAM2layers documentation [wam2layers.readthedocs.io]
• 4. research.tudelft.nl [research.tudelft.nl]
• 5. Welcome to WAM2layers! — WAM2layers documentation [wam2layers.readthedocs.io]
• 6. GitHub - WAM2layers/WAM2layers: Atmospheric moisture tracking [github.com]
• 7. GMD - Atmospheric moisture tracking with WAM2layers v3 [gmd.copernicus.org]

Protocols & Analytical Methods

Application Notes and Protocols for WAM2layers: A Python Package for Atmospheric Moisture Tracking

For Researchers, Scientists, and Drug Development Professionals

This document provides a comprehensive guide to installing and running the WAM2layers Python package, a powerful tool for atmospheric moisture tracking. These protocols are designed to assist researchers in tracing the origin and fate of atmospheric moisture, a critical component in understanding hydrological cycles, climate dynamics, and their implications across various scientific fields.

Introduction to WAM2layers

WAM2layers is an open-source Eulerian atmospheric moisture tracking model written in Python. It enables users to conduct two primary types of analyses:

• Backward tracking: Identifying the evaporative source regions of precipitation for a specific area and time.

• Forward tracking: Determining where moisture from a particular evaporative source region will eventually precipitate.

The model operates on a two-layer representation of the atmosphere, which has been shown to provide more accurate results than single-layer models, especially in regions with significant wind shear. WAM2layers is designed to work with gridded atmospheric data, such as the widely used ERA5 reanalysis dataset.[1]

Installation Protocol

The recommended method for installing WAM2layers is within a dedicated conda environment to manage dependencies effectively.

Prerequisites

• A command-line interface (e.g., Terminal on macOS/Linux, PowerShell on Windows).

• Miniconda or Anaconda distribution installed on your system.

Step-by-Step Installation

• Create and activate a new conda environment: Open your command-line interface and execute the following commands. This will create a new environment named wamenv with Python 3.11 and Cartopy, a key dependency for visualization, and then activate it.

• Install WAM2layers from PyPI: With the wamenv environment activated, install the latest version of WAM2layers using pip. To include the necessary packages for visualization, it is recommended to install the [viz] extra.

• (Optional) Editable installation from GitHub: For users who wish to view or modify the source code, an editable installation is recommended. This involves cloning the GitHub repository and installing it in editable mode.

Experimental Workflow

A typical WAM2layers experiment consists of three main stages, executed as command-line operations: preprocessing the input data, running the tracking simulation, and visualizing the results.

WAM2layers Experimental Workflow.

Stage 1: Preprocessing

The initial step involves processing the raw atmospheric data (e.g., ERA5) into a format suitable for the tracking model. This is managed by the wam2layers preprocess command.

Methodology:

• Prepare a configuration file: Create a YAML file (e.g., config.yaml) that specifies the parameters for your experiment. This includes the date range, geographical region, and paths to input and output files.

• Download input data: Obtain the necessary ERA5 data for your specified time period and region. The wam2layers package provides a utility to download example data to get started quickly.

• Execute the preprocessing command: Run the following command in your terminal, pointing to your configuration file:

This will generate preprocessed data files in the output directory specified in your configuration.

Stage 2: Tracking

Once the data is preprocessed, the core moisture tracking simulation is performed using the wam2layers track command.

Methodology:

• Configure tracking parameters: In your YAML configuration file, define the tracking direction (forward or backward), the "tagging" region (the area of interest for precipitation or evaporation), and other relevant settings.

• Run the tracking simulation: Execute the tracking command:

The model will then calculate the movement of "tagged" moisture through the atmosphere over the specified time period.

Stage 3: Visualization

The final stage is to visualize the tracking results, which are typically in the form of NetCDF files. The wam2layers visualize command facilitates the creation of maps and plots.

Methodology:

• Specify visualization options: The visualization can be configured through the same YAML file or with command-line arguments to specify the desired output format and variables to plot.

• Generate visualizations: Run the visualization command:

This will produce graphical representations of the moisture sources or sinks, depending on the tracking direction.

Data Presentation: Model Performance

The performance of WAM2layers v3 has been evaluated in various case studies. The following table summarizes key quantitative aspects of the model's application.

Parameter	Value / Description	Reference
Input Data Source	ERA5 Reanalysis	[1]
Horizontal Resolution	0.25° x 0.25° (native ERA5)
Temporal Resolution	Hourly
Vertical Layers	2 (upper and lower)	[1]
Default Layer Boundary	~812 hPa
Tracking Direction	Forward or Backward	[1]

Logical Relationship: The Two-Layer Model

WAM2layers simplifies the complex vertical structure of the atmosphere into two layers: an upper and a lower layer. This approach is a core concept of the model and is crucial for accurately capturing moisture transport, especially in the presence of varying wind directions at different altitudes.

Conceptual Diagram of the WAM2layers Two-Layer Structure.

This diagram illustrates the fundamental components of the WAM2layers model. Moisture can move horizontally within each layer and be exchanged vertically between the layers. Additionally, there is interaction with the Earth's surface through evaporation and precipitation, primarily involving the lower layer. This two-layer simplification is a key strength of the model, balancing computational efficiency with physical realism.[1]

References

• 1. research.tudelft.nl [research.tudelft.nl]

A Beginner's Guide to Atmospheric Moisture Tracking with WAM2layers

An Introduction to the Water Accounting Model - 2 layers (WAM2layers)

The WAM2layers model is a powerful open-source tool written in Python used for atmospheric moisture tracking.[1][2][3] It allows researchers to investigate the origins and fates of atmospheric moisture, providing critical insights into the hydrological cycle, weather patterns, and climate dynamics.[4][5][6] The model can operate in two primary modes:

• Backward tracking: To determine the evaporative sources of precipitation in a specific region.[1][2][4]

• Forward tracking: To identify where moisture from a particular area will eventually precipitate.[1][2][4]

This guide provides a detailed protocol for beginners to install, configure, and run a basic atmospheric moisture tracking experiment using WAM2layers.

Application Notes and Protocols

Installation and Setup

WAM2layers requires a Python environment and can be installed using pip. It is highly recommended to use a virtual environment to manage dependencies.

Experimental Protocol: Installation

• Prerequisites: Ensure you have a Python environment (version 3.11 or higher) and a command-line interface.[7] For Windows users, it is recommended to use Windows Subsystem for Linux (WSL).[7]

• Install Python: If Python is not installed, it is recommended to install it via Miniconda, which simplifies package management.[7]

• Create a Virtual Environment:

• Install WAM2layers: Install the package directly from PyPI. For a complete installation that includes visualization and development packages, use the following command:[7]

• Verify Installation: Check that the WAM2layers command-line interface is accessible:

Data Requirements and Preparation

WAM2layers processes gridded atmospheric data from sources like reanalysis datasets (e.g., ERA5) or climate model outputs.[4] The core of the model consists of a two-step process: preprocessing and tracking.[4] The preprocessing step collapses the input data onto a two-layer vertical grid.[4]

Parameter	Description	Typical Source
q	Specific humidity	ERA5, CMIP
u	Zonal wind component	ERA5, CMIP
v	Meridional wind component	ERA5, CMIP
sp	Surface pressure	ERA5, CMIP
evap	Evaporation	ERA5, CMIP
precip	Precipitation	ERA5, CMIP
tcwv	Total column water vapor	ERA5, CMIP

A summary of essential input data for a WAM2layers experiment.

For convenience, WAM2layers provides a utility to download example datasets, which is ideal for beginners.[4]

Computational Workflow: A Backward Tracking Experiment

This protocol outlines the steps to run a backward tracking experiment to identify the moisture sources for a historical precipitation event.

Experimental Protocol: Running a Simulation

• Download Example Data: Use the wam2layers download command to fetch a pre-packaged example case. This will create a new directory with the necessary input data and configuration file.[8]

• Review the Configuration File: Open the config-eiffel.yaml file. This file contains all the settings for the experiment, including the date range, tracking direction (forward or backward), and paths to input/output files.

• Preprocessing: Run the preprocessing step to prepare the ERA5 data. This command reads the raw data and aggregates it into the two-layer model structure.[4][8]

• Tracking: Execute the core tracking algorithm. For this example, we perform a backward track to find the origin of precipitation.[8][9]

• Visualization: After the tracking is complete, generate a visualization of the results.[8]

Visualizations

Logical Workflow of a WAM2layers Experiment

The following diagram illustrates the logical steps involved in conducting a typical WAM2layers experiment, from initial setup to final analysis.

References

• 1. Welcome to WAM2layers! — WAM2layers documentation [wam2layers.readthedocs.io]
• 2. WAM2layers [zenodo.org]
• 3. GitHub - WAM2layers/WAM2layers: Atmospheric moisture tracking [github.com]
• 4. GMD - Atmospheric moisture tracking with WAM2layers v3 [gmd.copernicus.org]
• 5. researchgate.net [researchgate.net]
• 6. researchgate.net [researchgate.net]
• 7. Installation — WAM2layers documentation [wam2layers.readthedocs.io]
• 8. wam2layers · PyPI [pypi.org]
• 9. youtube.com [youtube.com]

Application Notes and Protocols for WAM2layers ERA5 Input Data

For Researchers, Scientists, and Drug Development Professionals

These application notes provide a detailed guide on the input data requirements for the WAM2layers model when utilizing the ERA5 reanalysis dataset. Adherence to these protocols is crucial for accurate and reproducible atmospheric moisture tracking studies.

Introduction to WAM2layers and ERA5

WAM2layers is a widely used Eulerian moisture tracking model that quantifies the transport of atmospheric water vapor. It is instrumental in understanding precipitation sources and sinks, which has significant implications for hydrology, climatology, and related fields. ERA5 is the fifth generation ECMWF atmospheric reanalysis of the global climate, providing high-resolution hourly data on a vast number of atmospheric, land, and oceanic variables. The combination of WAM2layers and ERA5 allows for detailed and robust analyses of the atmospheric branch of the hydrological cycle.

Input Data Requirements

The WAM2layers model requires specific atmospheric and surface variables from the ERA5 dataset for its calculations. These variables can be broadly categorized into upper-air data and surface-level data. The model can be run using either model-level or pressure-level upper-air data; however, using pressure-level data is often more practical due to its easier accessibility.

Data Presentation: Quantitative Summary

The following tables summarize the essential and recommended ERA5 variables for running WAM2layers.

Table 1: Upper-Air Variables (Pressure Level Data)

Variable Name	ERA5 Short Name	Description	Vertical Levels	Temporal Resolution	Spatial Resolution
U-component of wind	u	Zonal wind speed	All available pressure levels	Hourly	0.25° x 0.25°
V-component of wind	v	Meridional wind speed	All available pressure levels	Hourly	0.25° x 0.25°
Specific humidity	q	Mass of water vapor per unit mass of moist air	All available pressure levels	Hourly	0.25° x 0.25°

Table 2: Surface and Single-Level Variables

Variable Name	ERA5 Short Name	Description	Temporal Resolution	Spatial Resolution
Total precipitation	tp	Accumulated precipitation over a period	Hourly	0.25° x 0.25°
Evaporation	e	Accumulated evaporation over a period	Hourly	0.25° x 0.25°
Surface pressure	sp	The pressure at the surface of the Earth	Hourly	0.25° x 0.25°
10m U-component of wind	u10	Zonal wind speed at 10 meters above the surface	Hourly	0.25° x 0.25°
10m V-component of wind	v10	Meridional wind speed at 10 meters above the surface	Hourly	0.25° x 0.25°
2m dewpoint temperature	d2m	Temperature to which air would have to cool to become saturated	Hourly	0.25° x 0.25°
Total column water vapour	tcwv	Total amount of water vapour in a column of air	Hourly	0.25° x 0.25°

Note: While Total Column Water Vapour (tcwv) is not strictly required, its inclusion is highly recommended to help mitigate potential errors in the model's water balance calculations.[1]

Experimental Protocols: Data Acquisition and Preprocessing

The successful application of WAM2layers hinges on the correct acquisition and preprocessing of ERA5 data. The following protocol outlines the key steps in this process.

Data Download from the Copernicus Climate Data Store (CDS)

• Account Creation: Register for an account on the Copernicus Climate Data Store (CDS) portal.

• API Key Installation: Install the CDS API key on your local machine to enable programmatic data downloads.

• Data Request Script: Utilize a Python script with the cdsapi library to request the required variables. It is recommended to download data in NetCDF format.

  • For upper-air data on pressure levels, specify the variables u, v, and q for all available pressure levels.

  • For surface data, request the variables tp, e, sp, u10, v10, and d2m.

  • Specify the desired date range, time steps (hourly), and geographical extent for your study area.

WAM2layers Preprocessing Workflow

The WAM2layers software includes a command-line interface for preprocessing the raw ERA5 data into a format suitable for the tracking model. The general workflow is as follows:

• Configuration File: Create a YAML configuration file (e.g., config.yaml). This file specifies the paths to your input ERA5 files, the desired output directory for the preprocessed data, the date range, and other relevant parameters.

• Run the Preprocessing Command: Execute the WAM2layers preprocessing script from your terminal, pointing to your configuration file. A typical command would be:

• Output: The preprocessing step will generate a single NetCDF file containing all the necessary variables, vertically integrated and structured for efficient use by the WAM2layers tracking algorithm.

Visualizations

Logical Relationship of Input Data

The following diagram illustrates the relationship between the different categories of ERA5 input data required by the WAM2layers model.

Caption: ERA5 input data categories for the WAM2layers model.

Experimental Workflow for Data Preparation

This diagram outlines the sequential steps involved in preparing ERA5 data for a WAM2layers simulation.

Caption: Workflow for ERA5 data preparation for WAM2layers.

References

• 1. gmd.copernicus.org [gmd.copernicus.org]

Configuring a WAM2layers Backward Tracking Experiment: Application Notes and Protocols

Application Notes

The Water Accounting Model 2-layers (WAM2layers) is a powerful open-source tool written in Python used for tracking the atmospheric moisture transport.[1][2][3] It can determine the origin of precipitation through backward tracking or the destination of evaporated moisture via forward tracking.[1][3][4] This document provides detailed protocols for configuring and executing a backward tracking experiment with WAM2layers, designed for researchers and scientists in atmospheric science, hydrology, and related fields.

A backward tracking analysis with WAM2layers identifies the source regions of evaporation for a given precipitation event or for the total precipitation in a specific area over a certain period.[4] The model operates by solving a prognostic equation for tagged moisture in gridded atmospheric datasets, such as reanalysis data or climate model output.[2] The core of the model is driven by a configuration file in YAML format, which allows for reproducible and transparent experimental setups.[4][5]

Experimental Protocols

Installation

To begin, a working Python environment is required. WAM2layers can be installed using pip, and it is highly recommended to use a virtual environment to manage dependencies.[6]

Protocol for Installation:

• Prerequisites: Ensure you have a command-line interface and a Python environment with pip installed. For Windows users, WSL or PowerShell can be used.[6] If Python is not installed, it is recommended to install it via conda.[6]

• Create a Virtual Environment (Recommended):

• Install WAM2layers:

  For visualization capabilities, install with the [viz] extra:

Data Preparation

A WAM2layers experiment requires meteorological data, typically from reanalysis datasets like ERA5. The necessary data includes variables such as precipitation, evaporation, and moisture fluxes at various vertical levels. [5] Protocol for Data Preparation:

• Define a Case Study: Identify the region of interest (the "tagging region" where precipitation occurs) and the time period for the analysis. [5]2. Acquire Meteorological Data: Download the required atmospheric data for your chosen period and spatial domain. The WAM2layers documentation provides guidance on the specific variables needed.

• Data Preprocessing: The raw data needs to be preprocessed to a format that WAM2layers can ingest. This involves ensuring the data is properly structured and all necessary variables are present. The model includes preprocessing scripts to facilitate this step. [5]

Configuration

The experiment is controlled by a YAML configuration file. [4][5]This file specifies the tracking direction, date range, data paths, and other parameters. An example configuration file can be found on the WAM2layers GitHub repository. [5] Protocol for Configuration:

• Create a Configuration File: Create a new file named config.yaml.

• Specify General Settings:

  • log_level: Set the verbosity of the output messages (e.g., INFO, DEBUG).

  • preprocess_path: Path to the preprocessed input data.

  • output_path: Directory where the output files will be saved.

• Define the Tracking Period:

  • track_start_date: The beginning of the tracking period.

  • track_end_date: The end of the tracking period.

• Set Tracking Direction:

  • direction: For a backward tracking experiment, set this to backward.

• Specify the Tagging Region:

  • Define the geographical boundaries of the area where precipitation is being tracked from.

• Configure Preprocessing and Tracking Settings:

  • Adjust parameters related to the model's numerical scheme and output frequency as needed. [4]

Data Presentation

The following table summarizes the key parameters in the config.yaml file for a backward tracking experiment.

Parameter	Type	Description	Example Value
log_level	string	Verbosity of logging	INFO
preprocess_path	string	Path to preprocessed data	/path/to/preprocessed_data
output_path	string	Path to save output files	/path/to/output
track_start_date	string	Start date of tracking (YYYY-MM-DD)	2023-01-01
track_end_date	string	End date of tracking (YYYY-MM-DD)	2023-01-31
direction	string	Tracking direction	backward
tagging_region	dictionary	Geographical boundaries of the source/sink region	{ "latitude_min": 30, "latitude_max": 40, "longitude_min": -10, "longitude_max": 0 }
output_frequency	integer	Frequency of writing output in days	1

Mandatory Visualization

WAM2layers Backward Tracking Workflow

The following diagram illustrates the general workflow for setting up and running a WAM2layers backward tracking experiment.

Caption: A high-level overview of the WAM2layers backward tracking experimental workflow.

Logic of Backward Moisture Tracking

This diagram illustrates the conceptual logic of the backward tracking process in WAM2layers.

Caption: Conceptual flow of a WAM2layers backward tracking simulation.

References

• 1. Welcome to WAM2layers! — WAM2layers documentation [wam2layers.readthedocs.io]
• 2. GMD - Atmospheric moisture tracking with WAM2layers v3 [gmd.copernicus.org]
• 3. GitHub - WAM2layers/WAM2layers: Atmospheric moisture tracking [github.com]
• 4. Tracking — WAM2layers documentation [wam2layers.readthedocs.io]
• 5. m.youtube.com [m.youtube.com]
• 6. Installation — WAM2layers documentation [wam2layers.readthedocs.io]

Forward Tracking of Evaporation Sources with WAM2layers: Application Notes and Protocols

For Researchers, Scientists, and Drug Development Professionals

These application notes provide a comprehensive guide to utilizing the Water Accounting Model 2-layers (WAM2layers) for forward tracking of evaporation sources. This powerful tool enables researchers to identify the downwind precipitation sinks of moisture originating from specific regions of interest. The following sections detail the underlying principles, experimental protocols, data presentation, and visualization of the WAM2layers model.

Introduction to WAM2layers Forward Tracking

The WAM2layers model is an Eulerian atmospheric moisture tracking tool that quantifies the transport of water vapor in the atmosphere.[1][2][3] In its forward tracking mode, the model tags evaporation from a predefined source region and tracks this tagged moisture as it travels with the atmospheric flow.[2][4][5] By accounting for processes such as advection, diffusion, precipitation, and changes in atmospheric storage, WAM2layers can determine where evaporated water eventually precipitates.[2][3][6]

This methodology is particularly useful for:

• Understanding the role of specific land or water bodies as moisture sources for downwind precipitation.

• Investigating the impact of land-use change on regional and continental rainfall patterns.

• Identifying the origins of moisture contributing to extreme precipitation events.

• Studying the atmospheric branch of the hydrological cycle.

Data Presentation: Quantitative Insights from Forward Tracking Studies

The quantitative outputs of WAM2layers forward tracking experiments provide valuable insights into atmospheric moisture transport. The following tables summarize key findings from published studies, showcasing the model's application across different geographical regions and scientific questions.

Table 1: Forward Tracking of Evaporation from Lake Volta

This table is based on the forward-tracking example case described in the WAM2layers v3 documentation, tracking evaporation from the Lake Volta region in Ghana.

Parameter	Value	Source
Source Region	Lake Volta, Ghana	Kalverla et al., 2025
Tracking Period	July 1998	Kalverla et al., 2025
Total Tagged Evaporation
- Precipitated within Domain	86%	Kalverla et al., 2025
- Transported out of Domain	14%	Kalverla et al., 2025
Primary Precipitation Sinks
- Sahel Region
- Guinea Coast

Table 2: Contribution of Continental Evaporation to Precipitation

This table presents hypothetical data based on the types of results WAM2layers can generate, illustrating the contribution of evaporation from a continental source region to precipitation in various downwind areas.

Source Region	Sink Region	Mean Annual Contribution (%)	Seasonal Peak Contribution (%)
Amazon Basin	Andes Foothills	15	25 (Wet Season)
La Plata Basin	8	12 (Summer)
North American Monsoon Region	Central United States	10	18 (July-August)
Tibetan Plateau	Yangtze River Basin	12	20 (Summer)

Experimental Protocols: A Step-by-Step Guide to Forward Tracking

This section provides a detailed methodology for conducting a forward tracking experiment using WAM2layers. The protocol is divided into three main stages: installation and setup, data preprocessing, and running the forward tracking simulation.

Installation and Setup

• Install WAM2layers: The model is a Python package and can be installed using pip. It is recommended to create a dedicated conda environment to manage dependencies.

• Download Example Case: WAM2layers provides an example dataset to familiarize users with the workflow.

Data Preprocessing

The first step in a WAM2layers experiment is to process the input atmospheric data into the two-layer vertical structure used by the model. WAM2layers is designed to work with ERA5 reanalysis data from the Copernicus Climate Data Store.

• Obtain Input Data: Download the necessary ERA5 data (e.g., specific humidity, u and v components of wind, surface pressure, evaporation, and precipitation) for your region and period of interest. The WAM2layers documentation provides guidance on acquiring the correct data.

• Run the Preprocessing Step: Use the wam2layers preprocess command, pointing to your configuration file. This command reads the raw atmospheric data and calculates the vertically integrated moisture fluxes and storage for the two atmospheric layers.

Forward Tracking Simulation

Once the data is preprocessed, you can run the forward tracking simulation.

• Configure the Tracking Experiment: Create a YAML configuration file (.yaml) to specify the parameters for your experiment. Key parameters for forward tracking include:

  • direction: "forward"

  • tagging_region: (specify the latitude and longitude bounds of your evaporation source)

  • preprocess_dir: (path to the preprocessed data)

  • output_dir: (path to store the tracking results)

  • start_date: and end_date: (the period for the tracking simulation)

• Execute the Tracking Run: Use the wam2layers track command with your configuration file.

  This command will read the preprocessed data, "tag" the evaporation from your specified source region at each time step, and track the movement of this tagged moisture through the atmosphere, calculating where it precipitates.

Visualization and Analysis

WAM2layers includes a basic visualization script to generate a map of the tracked precipitation.

• Generate Output Visualization: Use the wam2layers visualize output command.

  This will create a plot showing the spatial distribution of precipitation that originated from your tagged evaporation source. Further analysis can be performed on the NetCDF output files using standard data analysis software (e.g., Python with xarray, R, Panoply).

Mandatory Visualizations: Diagrams of Workflows and Concepts

The following diagrams illustrate the key workflows and conceptual underpinnings of a WAM2layers forward tracking experiment.

Caption: High-level workflow for a WAM2layers forward tracking experiment.

Caption: Conceptual diagram of moisture movement in WAM2layers.

References

• 1. WAM2layers - WUR [wur.nl]
• 2. GMD - Atmospheric moisture tracking with WAM2layers v3 [gmd.copernicus.org]
• 3. researchgate.net [researchgate.net]
• 4. Tracking — WAM2layers documentation [wam2layers.readthedocs.io]
• 5. Welcome to WAM2layers! — WAM2layers documentation [wam2layers.readthedocs.io]
• 6. EGUsphere - Atmospheric moisture tracking with WAM2layers v3 [egusphere.copernicus.org]

Analyzing WAM2layers Output Data in Python: An Application Note

For Researchers, Scientists, and Drug Development Professionals

This application note provides a detailed protocol for analyzing output data from the WAM2layers atmospheric moisture tracking model using Python. WAM2layers is a powerful tool for understanding atmospheric dynamics and studying rainfall patterns by mapping their sources and sinks.[1] This guide will walk you through the necessary steps to install required software, load, and analyze the model's output, and visualize the results.

Data Presentation

WAM2layers output typically consists of gridded atmospheric data that quantifies the movement of moisture. The primary output variables will differ depending on whether a forward or backward tracking experiment was performed. Key quantitative data from a WAM2layers analysis can be summarized as follows:

Data Category	Potential Variable Names (in NetCDF output)	Description	Dimensions	Units
Tracked Moisture Fluxes	ft_n, ft_e	Vertically integrated northward and eastward moisture fluxes of tagged water vapor.	time, latitude, longitude	kg m⁻¹ s⁻¹
Tracked Precipitable Water	s_track	The amount of tagged water vapor in the atmospheric column.	time, latitude, longitude	kg m⁻²
Source/Sink Contributions	e_track, p_track	Tracked evaporation (source) or precipitation (sink) at the surface.	time, latitude, longitude	kg m⁻² s⁻¹
Moisture Balance	moisture_budget_error	The error in the moisture budget calculation, useful for assessing model performance.	time, latitude, longitude	kg m⁻²

Note: The exact variable names may vary based on the specific version of WAM2layers and the configuration of the experiment. It is essential to inspect the output file to determine the precise names of the variables.

Experimental Protocols

This section details the methodology for loading, analyzing, and visualizing WAM2layers output data using Python. The recommended library for handling this type of gridded data is xarray, which provides a powerful and intuitive interface for working with labeled multi-dimensional arrays, such as those found in NetCDF files.

Software Installation

First, ensure you have Python 3.9 or higher installed. The wam2layers package and its dependencies can be installed using pip. It is highly recommended to use a virtual environment to manage project dependencies.

This installation includes xarray for data handling and cartopy and matplotlib for plotting.

Loading WAM2layers Output Data

WAM2layers output is typically saved in NetCDF format (.nc files). The xarray library is the standard tool for reading and writing these files in the atmospheric and climate sciences.

The following Python script demonstrates how to open a WAM2layers output file:

Data Inspection and Exploration

After loading the dataset, the first step is to inspect its contents to understand the variables, dimensions, and metadata.

Basic Analysis: Calculating Regional Moisture Sources

A common analysis is to calculate the total moisture contribution from a specific region over time. This involves selecting a spatial area and summing the tracked evaporation or precipitation.

Mandatory Visualization

Clear visualization is crucial for interpreting the spatial and temporal patterns in WAM2layers output. The following diagrams, generated using Graphviz, illustrate key workflows and relationships.

References

• 1. Tracking — WAM2layers documentation [wam2layers.readthedocs.io]

Application Notes and Protocols for Mapping Precipitation Sources of Extreme Weather Events

Audience: Researchers, scientists, and professionals in atmospheric science, hydrology, and climate science.

Introduction: The identification of moisture sources contributing to extreme precipitation events is critical for understanding hydroclimatic processes, improving weather forecasting models, and assessing the impacts of climate change. This document provides an overview and detailed protocols for two primary methodologies used to map these precipitation sources: Lagrangian trajectory modeling and stable isotope analysis of water. These methods can be employed independently or in conjunction to provide a more robust characterization of moisture origins.

Methodology 1: Lagrangian Trajectory Modeling

Lagrangian models are computational tools used to determine the pathways of air parcels. For precipitation source analysis, these models are run backward in time from a location of interest (the "receptor" site) to identify the regions from which the air masses, and therefore the moisture, originated. The Hybrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT) model, developed by NOAA, is a widely used tool for this purpose.[1][2][3]

Quantitative Data Presentation

The output of Lagrangian models is typically a set of trajectories that can be analyzed to determine the statistical contribution of different source regions. While specific quantitative outputs are highly dependent on the event and location, the following table outlines key input parameters for a HYSPLIT analysis of an extreme precipitation event.

Parameter	Description	Typical Setting/Value for Extreme Events	Rationale
Meteorological Data	Gridded meteorological data used to drive the model (e.g., wind, pressure, temperature fields).[4]	Global Data Assimilation System (GDAS) at 0.5 or 1.0-degree resolution. For higher resolution, North American Mesoscale (NAM) or High-Resolution Rapid Refresh (HRRR) can be used.	GDAS provides global coverage for long-range transport. Higher-resolution datasets are better for shorter-range, more detailed analysis of convective systems.
Trajectory Direction	Forward or backward in time.[5]	Backward	To identify the source regions of air parcels arriving at the event location.
Total Run Time	The duration for which the back-trajectory is calculated.	120-240 hours (5-10 days)	This duration is typically sufficient to trace moisture back to its oceanic or major terrestrial source region.
Starting Location(s)	The latitude, longitude, and altitude(s) from which to start the back-trajectories.	Centered on the location of extreme precipitation. Multiple starting altitudes (e.g., 500m, 1500m, 3000m AGL) should be used.	Precipitation is a result of processes throughout the atmospheric column. Multiple altitudes capture moisture transport at different levels.
Vertical Motion Model	The method used to calculate the vertical movement of the air parcel.[5]	Model vertical velocity (Omega)	This uses the vertical velocity fields from the meteorological input data, providing a more realistic representation of atmospheric motion.
Ensemble Options	Running multiple trajectories from slightly perturbed starting points to account for uncertainty.	Trajectory ensemble using meteorological variations.[5]	This provides a probabilistic assessment of the source region, which is more robust than a single trajectory.

Experimental Protocol: HYSPLIT Back-Trajectory Analysis

This protocol outlines the steps to identify moisture sources for an extreme precipitation event using the HYSPLIT model.

• Define the Event:

  • Identify the date, time, and geographic location (latitude and longitude) of the extreme precipitation event.

• Acquire Meteorological Data:

  • Download the appropriate meteorological data files (e.g., GDAS, NAM) that cover the time period of the event and the preceding 10 days. This can often be done directly through the HYSPLIT interface.[5]

• Set Up the HYSPLIT Model:

  • Open the HYSPLIT graphical user interface (GUI).

  • Input the starting location(s) of the event. It is recommended to use multiple starting points both horizontally (to cover the area of precipitation) and vertically (e.g., at different altitudes above ground level).

  • Select the "backward" trajectory direction.

  • Set the total run time (e.g., 120 hours).

  • Choose the meteorological data file(s).

  • Select the vertical motion model (e.g., "Model vertical velocity").

  • Configure ensemble options if desired.

• Run the Model:

  • Execute the model run. HYSPLIT will generate trajectory output files.

• Analyze the Output:

  • Visualize the back-trajectories on a map.

  • Perform a cluster analysis on the trajectories to group them into distinct pathways. This helps in identifying the primary source regions.

  • Couple the trajectory data with meteorological variables (e.g., specific humidity) along the trajectory path to identify where the air parcels gained moisture.

  • Use statistical methods like Potential Source Contribution Function (PSCF) or Concentration Weighted Trajectory (CWT) to create maps showing the geographic areas with the highest probability of being a moisture source.[3]

Workflow Visualization

Caption: Workflow for HYSPLIT back-trajectory analysis.

Methodology 2: Stable Isotope Analysis

The stable isotopes of water, specifically Oxygen-18 (¹⁸O) and Deuterium (²H or D), are powerful tracers of the water cycle. The isotopic composition of precipitation is influenced by the source of the water vapor and the processes of evaporation and condensation it undergoes. Lighter isotopes (¹H and ¹⁶O) evaporate more readily, while heavier isotopes (²H and ¹⁸O) condense more readily.[6][7] This fractionation process leads to distinct isotopic "fingerprints" for water from different sources.

Quantitative Data Presentation

The isotopic composition is expressed in delta (δ) notation in parts per thousand (‰) relative to the Vienna Standard Mean Ocean Water (VSMOW). The following table provides example δ¹⁸O and δD values for major moisture sources. Note that these values can vary significantly based on season, temperature, and specific location.

Moisture Source	Typical δ¹⁸O (‰)	Typical δD (‰)	Key Characteristics
Tropical Atlantic Ocean	-2.0 to 0.0	-10.0 to 0.0	Relatively enriched due to warm sea surface temperatures.
North Atlantic Ocean	-5.0 to -2.0	-30.0 to -15.0	More depleted than tropical sources due to cooler temperatures.
Tropical Pacific Ocean	-4.5 to -2.0	-25.0 to -10.0	Isotopic signatures can be more negative than the Indian Ocean.[8]
Indian Ocean	-2.0 to 0.0	-10.0 to 0.0	Similar to the Tropical Atlantic, influenced by monsoon dynamics.
Arctic Ocean	-20.0 to -10.0	-150.0 to -70.0	Highly depleted due to cold temperatures and long transport paths.
Terrestrial Recycled Moisture	Highly variable	Highly variable	Tends to be more depleted than oceanic sources in the same latitude, influenced by evapotranspiration.

Experimental Protocol: Precipitation Isotope Analysis

This protocol describes the collection and analysis of precipitation samples to determine their isotopic composition.

• Sample Collection:

  • During an extreme weather event, collect precipitation samples at regular intervals (e.g., hourly or every few hours).

  • Use a clean collection vessel, such as a high-density polyethylene (HDPE) bottle with a funnel.

  • Ensure the collector is placed in an open area, away from buildings and trees, to avoid contamination.

  • Immediately after collection, transfer the sample to a smaller, airtight vial with no headspace to prevent evaporation.

  • Record the date, time, and location of each sample.

• Sample Storage:

  • Store the samples in a cool, dark place (refrigeration at 4°C is ideal) until analysis.

• Sample Analysis:

  • The analysis of δ¹⁸O and δD is performed using an isotope ratio mass spectrometer (IRMS) or a laser absorption spectrometer.

  • Prepare the samples and standards according to the instrument manufacturer's instructions.

  • Run the samples on the spectrometer. The instrument will measure the ratio of the heavy to light isotopes (¹⁸O/¹⁶O and ²H/¹H).

  • The results are reported in delta notation (δ¹⁸O and δD) in per mil (‰) relative to VSMOW.

• Data Interpretation:

  • Plot the δ¹⁸O and δD values on a dual-isotope plot.

  • Compare the isotopic composition of the precipitation samples to the known values of different moisture sources (see table above).

  • Analyze the temporal evolution of the isotopic values during the event. A shift in isotopic composition can indicate a change in moisture source.

  • For a more advanced analysis, use the isotopic data in conjunction with back-trajectory modeling to link specific isotopic signatures to air parcel pathways.

Logical Relationship Visualization

Caption: Logical flow from moisture source to isotopic identification.

References

• 1. essd.copernicus.org [essd.copernicus.org]
• 2. researchgate.net [researchgate.net]
• 3. HYSPLIT - Wikipedia [en.wikipedia.org]
• 4. ntrs.nasa.gov [ntrs.nasa.gov]
• 5. HYSPLIT – Air Resources Laboratory [arl.noaa.gov]
• 6. youtube.com [youtube.com]
• 7. researchgate.net [researchgate.net]
• 8. hess.copernicus.org [hess.copernicus.org]

Technical Notes & Optimization

common errors in WAM2layers configuration

Welcome to the WAM2layers technical support center. This guide provides answers to common questions and solutions for issues you may encounter during the configuration and execution of your moisture tracking experiments.

Frequently Asked Questions (FAQs)

Here are answers to some of the most common questions about WAM2layers.

Q1: What is WAM2layers and what are its main applications?

A1: WAM2layers is an atmospheric moisture tracking model written in Python. It is designed to identify the origin of precipitation (backtracking) or determine the destination of evaporated moisture (forward tracking). Its primary applications are in hydrology, meteorology, and climate science to study moisture recycling, atmospheric rivers, and the sources of rainfall for specific regions or extreme weather events.

Q2: What kind of input data do I need to run WAM2layers?

A2: WAM2layers typically requires gridded atmospheric data from reanalysis datasets like ERA5 or output from climate models. Key input variables include wind fields (u and v components), specific humidity, and surface fluxes like evaporation and precipitation. The model is designed to work with data on different vertical levels (model levels or pressure levels) and aggregates them into a two-layer system.

Q3: How is a WAM2layers experiment configured?

A3: Experiments are configured using a YAML file (commonly config.yaml). This file contains all the settings for your experiment, including the time period to analyze, the geographical domain, paths to input data, the tracking direction (forward or backward), and the location of the output files. Using a configuration file makes experiments easily reproducible and shareable.

Q4: What is the difference between the "preprocessing" and "tracking" steps?

A4: The WAM2layers workflow is a two-step process:

• Preprocessing: In this first step, the raw input data (e.g., from ERA5) is read and converted into a format that the tracking model can use. This involves calculating the necessary moisture fluxes and collapsing the atmospheric data into the two-layer vertical grid used by the model.

• Tracking: This is the core step where the model integrates tagged moisture through the atmosphere over time, either forward or backward, based on the preprocessed data and the settings in your configuration file.

Troubleshooting Guide

This guide provides solutions to specific errors and problems you might encounter.

Installation & Environment

Q: I'm having trouble installing WAM2layers. What is the recommended method?

A: The most reliable way to install WAM2layers is by using a dedicated Conda environment. This ensures that all required dependencies are installed without conflicting with other Python packages on your system.

Experimental Protocol: Recommended Installation

• Install Conda: If you don't have it, install Miniconda or Anaconda.

• Create a new environment: Open your terminal and create a new environment. It is recommended to use Python 3.11 or newer.

• Activate the environment:

• Install WAM2layers: The package is available on PyPI. For visualization capabilities, install with the [viz] extra.

This process isolates your installation and is the first step in resolving many common issues.

Configuration Errors (config.yaml)

Q: My script fails immediately with a FileNotFoundError. What should I check?

A: A FileNotFoundError at the start of a run usually points to an incorrect path in your config.yaml file.

• Check data_dir: Ensure the data_dir path points to the correct location of your preprocessed input data.

• Check tagging_region_mask: If you are using a NetCDF file to define your tagging region, verify that the path to this file is correct.

• Absolute vs. Relative Paths: Try using absolute paths to your data directories to avoid ambiguity.

Q: I'm getting a ValueError or TypeError related to dates.

A: Date and time formats in the configuration file are critical. Ensure that the track_start_date and track_end_date follow the YYYY-MM-DD HH:MM format precisely. For example: 2022-01-01 00:00. Any deviation can lead to parsing errors.

Q: The model output seems strange and the tracked moisture is not where I expect it. What could be wrong in my configuration?

A: This can happen if the tagging_region is incorrectly defined. The tagging region specifies the area where moisture is "tagged" for tracking.

• Bounding Box: If using a bounding box ([lon_min, lon_max, lat_min, lat_max]), ensure the coordinates are in the correct order and within the expected range (-180 to 180 for longitude, -90 to 90 for latitude).

• Mask File: If using a mask file, ensure the NetCDF file contains a variable with values between 0 and 1, where 1 indicates the cells to be included in the tagging region. A common mistake is having a mask with incorrect dimensions or variable names.

The following table summarizes key configuration parameters that are common sources of errors.

Parameter	Common Error	Recommended Check
data_dir	FileNotFoundError	Verify the path is correct and absolute.
track_start_date	ValueError, TypeError	Use YYYY-MM-DD HH:MM format.
track_end_date	ValueError, TypeError	Use YYYY-MM-DD HH:MM format.
tagging_region	Incorrect tracking results	Check coordinate order and ranges.
output_frequency	Inconsistent output times	Ensure the value is a valid pandas frequency string (e.g., '6H', '1D').

Input Data & Preprocessing

Q: The preprocessing step fails with an error about missing variables in my input files.

A: This indicates that your source data (e.g., ERA5 NetCDF files) is incomplete. The preprocessing step expects specific variables to be present. For a standard ERA5-based experiment, you will need files containing:

• Surface level data (evaporation, precipitation, 2m dewpoint temperature, 10m u/v wind components).

• Pressure or model level data (specific humidity, u/v wind components for all vertical levels).

Double-check that you have downloaded all the required variables for the entire time period of your experiment.

Runtime & Tracking Errors

Q: My experiment crashes with a MemoryError during the tracking step.

A: MemoryError is common when tracking over a large domain or a long time period, as WAM2layers needs to hold significant amounts of data in memory. To mitigate this:

• Reduce the Tracking Domain: In your config.yaml, reduce the geographical area (preprocess_bounds) to the smallest region necessary for your research question.

• Shorten the Time Period: Run the experiment for a shorter duration (track_start_date to track_end_date). You can often run experiments in sequential chunks.

• Increase System Memory: If possible, run the experiment on a machine with more RAM.

Q: The model output contains NaN (Not a Number) values or the results look numerically unstable.

A: This can be caused by violating the Courant-Friedrichs-Lewy (CFL) condition, which relates to numerical stability. The CFL condition requires that the model's time step is short enough that moisture does not travel more than one grid cell in a single step.

• Adjust target_frequency: In your config.yaml, try decreasing the target_frequency (i.e., making the time step shorter). For example, change it from '6H' (6 hours) to '3H' (3 hours). This increases computational time but improves stability.

Troubleshooting Workflow

When you encounter an error, it is recommended to follow a systematic approach to identify the cause. The diagram below illustrates a logical workflow for troubleshooting a failing WAM2layers experiment.

WAM2layers Technical Support Center: Optimizing Computational Performance

This technical support center provides troubleshooting guidance and frequently asked questions (FAQs) to assist researchers, scientists, and drug development professionals in optimizing the computational performance of their WAM2layers experiments.

Frequently Asked Questions (FAQs)

Q1: What is WAM2layers and what is it used for?

WAM2layers is an atmospheric moisture tracking model written in Python.[1][2] It is designed to determine the origin of precipitation (backtracking) or the destination of evaporated moisture (forward tracking).[1][2][3] This tool is valuable for understanding the hydrological cycle, meteorology, and the impact of climate and land-use changes on rainfall patterns.[4][5]

Q2: What are the main steps in a WAM2layers experiment?

A typical WAM2layers experiment involves the following core steps:

• Installation and Configuration: Setting up the Python environment and creating a YAML configuration file with experiment-specific settings.[3][6]

• Data Preprocessing: Ingesting and preparing atmospheric data (e.g., from ERA5) for the tracking simulation.[4][7][8]

• Tracking: Running the core backtracking or forward tracking simulation.[9]

• Visualization: Generating plots and visualizations of the tracking results.[1]

Q3: What factors can influence the computational performance of my WAM2layers experiment?

Several factors can impact the runtime and memory usage of your simulations:

• Input Data Resolution: Higher spatial and temporal resolution data will increase computational load.[4][5]

• Tracking Period: Longer simulation periods require more processing time.

• Spatial Domain: Larger geographical areas for tracking will increase the amount of data to be processed.[4][7]

• Hardware: The number of available CPU cores and the amount of RAM will directly affect performance, especially when running parallel computations.

• Software Configuration: The chosen settings in the configuration file, such as input_frequency and target_frequency, can impact performance.[3]

Troubleshooting Guides

Issue 1: Slow Computation Times

Long processing times are a common challenge, especially with high-resolution data and long tracking periods. Here are strategies to improve performance:

Solution:

• Parallelize with Dask: WAM2layers is designed to work with Dask, a parallel computing library in Python, to distribute computations across multiple processor cores. While the official documentation does not provide a specific tutorial on Dask integration, the general principles of Dask can be applied. For embarrassingly parallel workloads, where the same operation is applied to different chunks of data, Dask can significantly speed up processing.

• Reduce Data Resolution or Domain: If feasible for your research question, consider reducing the spatial or temporal resolution of your input data or limiting the geographical tracking domain.[4][7] This will decrease the amount of data that needs to be processed.

• Hardware Upgrade: If you consistently experience slow performance, consider running your experiments on a machine with more CPU cores and a larger amount of RAM.

Issue 2: Memory Errors

Memory errors, often indicated by messages like MemoryError, can occur when the size of the data being processed exceeds the available RAM.

Solution:

• Optimize Data Loading: Instead of loading the entire dataset into memory at once, process it in smaller chunks. The upgraded version of WAM2layers has improved memory management by processing data in a loop and only loading neighboring time steps into memory.[11]

• Adjust Chunk Sizes: In the context of Dask, the size of the data chunks can affect memory usage. Experiment with smaller chunk sizes to reduce the memory footprint of each parallel task.

• Close Unused Processes: Ensure that other memory-intensive applications are closed before running a large WAM2layers experiment.

• Increase System Memory: If memory errors persist, you may need to run your experiment on a machine with more RAM.

Issue 3: Configuration and Input Data Errors

Incorrect settings in the config.yaml file or issues with the input data can lead to errors during the preprocessing or tracking stages.

Solution:

• Verify Configuration File: Carefully review your YAML configuration file for correct formatting and valid parameter values. The official WAM2layers documentation and the example configuration file on the GitHub page are excellent resources.[3]

• Check Input Data: Ensure that your input data is correctly formatted and accessible. WAM2layers v3 has built-in support for ERA5 and CMIP data.[4] For other datasets, a custom data loader script may be required.[4]

• Review Log Files: WAM2layers generates log files that can provide detailed error messages.[7] Inspect these logs to identify the specific cause of the error.

• Consult the Community: If you are unable to resolve the issue, consider seeking help from the WAM2layers community through the GitHub Discussions forum.[1]

Experimental Protocols

Protocol 1: Benchmarking Computational Performance

To assess the impact of different configurations on performance, you can conduct a series of benchmark experiments.

Methodology:

• Establish a Baseline: Run your experiment with a standard configuration (e.g., default settings, single-core processing) and record the total execution time and peak memory usage.

• Vary a Single Parameter: In subsequent runs, modify one parameter at a time. For example:

  • Parallelization: Run the experiment with an increasing number of Dask workers (CPU cores) and record the execution time for each run.

  • Chunk Size: If using Dask, experiment with different data chunk sizes and measure the impact on performance and memory usage.

  • Data Resolution: If applicable, run the experiment with different input data resolutions.

• Record and Analyze: Systematically record the execution time and memory usage for each experimental run.

• Summarize Results: Present the results in a table for easy comparison.

Table 1: Example Performance Benchmark Data

Number of CPU Cores	Execution Time (minutes)	Peak Memory Usage (GB)
1	120	8
2	65	9
4	35	10
8	20	12

Note: The values in this table are for illustrative purposes only. Actual performance will vary depending on the specific experiment and hardware.

References

• 1. wam2layers · PyPI [pypi.org]
• 2. Welcome to WAM2layers! — WAM2layers documentation [wam2layers.readthedocs.io]
• 3. m.youtube.com [m.youtube.com]
• 4. GMD - Atmospheric moisture tracking with WAM2layers v3 [gmd.copernicus.org]
• 5. doaj.org [doaj.org]
• 6. Installation — WAM2layers documentation [wam2layers.readthedocs.io]
• 7. research.tudelft.nl [research.tudelft.nl]
• 8. researchgate.net [researchgate.net]
• 9. Tracking — WAM2layers documentation [wam2layers.readthedocs.io]
• 10. Finding the Best Chunking Strategy for Accurate AI Responses | NVIDIA Technical Blog [developer.nvidia.com]
• 11. Cracking the (moisture) tracking code | by Peter Kalverla | Netherlands eScience Center [blog.esciencecenter.nl]

WAM-2layers Preprocessing: Technical Support Center

This technical support center provides troubleshooting guidance and answers to frequently asked questions related to the preprocessing scripts for WAM-2layers, a widely used atmospheric moisture tracking model. It is intended for researchers, scientists, and drug development professionals who are utilizing this tool for their experiments.

Frequently Asked Questions (FAQs)

Q1: What is the purpose of the WAM-2layers preprocessing step?

A1: The preprocessing step in WAM-2layers is crucial for transforming raw atmospheric data, typically from reanalysis datasets like ERA5, into a format that the main tracking algorithm can utilize.[1][2] This involves collapsing the multi-level atmospheric data into a two-layer model and calculating the necessary variables for moisture tracking.[1][2]

Q2: What are the essential input variables required for the preprocessing scripts?

A2: To successfully run the preprocessing scripts, you will need the following variables from your input dataset: horizontal wind components (u and v), specific humidity (q), precipitation (P), evaporation (E), and surface pressure (ps). Additionally, total column water can be included to improve accuracy.

Q3: Can I use atmospheric data other than ERA5 with WAM-2layers?

A3: Yes, while WAM-2layers is optimized for ERA5 data, it is possible to use other gridded atmospheric datasets. However, this may require creating a custom data loader script to ensure the data is converted into the standard internal format that WAM-2layers expects.

Q4: Where can I find the official documentation and community support for WAM-2layers?

A4: The primary resources for WAM-2layers are its official documentation and the GitHub repository. The documentation provides a quickstart guide and theoretical background, while the GitHub repository contains the source code, an issue tracker for reporting bugs, and community discussions.

Troubleshooting Guide

This guide addresses common issues that may arise during the WAM-2layers preprocessing stage.

Common Preprocessing Errors

Error Message/Symptom	Potential Cause	Suggested Solution
FileNotFoundError or No such file or directory	The path to the input data in the configuration file is incorrect, or the data files are not present in the specified directory.	Verify the input_folder path in your configuration file. Ensure that your atmospheric data files are correctly named and located in the specified directory.
KeyError or missing variable error	The input NetCDF files are missing one or more of the required variables (e.g., 'u', 'v', 'q', 'sp', 'e', 'tp').	Check the contents of your input data files using a tool like ncdump or xarray to confirm that all necessary variables are present and correctly named. Refer to the WAM-2layers documentation for the expected variable names.
Incorrect timestamp or datetime errors	The time coordinate in the input data is not correctly formatted or is inconsistent across different input files.	Ensure that the time variable in your NetCDF files is in a recognized datetime format. The WAM-2layers GitHub issues page suggests that consistent time coordinates are crucial.
Slow preprocessing performance or memory errors	The input dataset is very large (e.g., high resolution or long time series), leading to excessive memory usage.	The WAM-2layers model has been updated from Python 2 to Python 3 to improve performance and memory management. Ensure you are using the latest version. For very large datasets, consider processing the data in smaller chunks or using a machine with more available RAM.
Inaccurate or unexpected preprocessed data values	The units of the input variables are incorrect, or the vertical coordinate system is not properly handled.	Verify that the units of your input data match the expectations of the preprocessing scripts (e.g., wind speed in m/s, pressure in Pa). If using model-level data, ensure the vertical coordinate information is correctly interpreted.

Experimental Protocols & Methodologies

The core of the WAM-2layers preprocessing methodology involves the vertical integration of atmospheric data into two distinct layers. The horizontal wind and humidity fields are vertically integrated to compute the total water vapor fluxes. Surface fluxes of precipitation and evaporation are also processed to be used as source and sink terms in the moisture tracking calculations.

For a detailed mathematical description of the preprocessing steps, users are encouraged to consult the official WAM-2layers v3 reference paper, which provides the governing equations and a thorough explanation of how input data is collapsed onto the two-layer grid.

WAM-2layers Preprocessing Debugging Workflow

The following diagram illustrates a logical workflow for troubleshooting issues with the WAM-2layers preprocessing scripts.

Caption: A flowchart for debugging WAM-2layers preprocessing scripts.

References

• 1. GMD - Atmospheric moisture tracking with WAM2layers v3 [gmd.copernicus.org]
• 2. research.tudelft.nl [research.tudelft.nl]

dealing with high Courant numbers in WAM2layers

This technical support center provides troubleshooting guides and frequently asked questions (FAQs) to assist researchers, scientists, and drug development professionals in their experiments with WAM2layers.

Troubleshooting Guides

Issue: My WAM2layers simulation is unstable or producing spurious results.

High Courant numbers are a common cause of instability in atmospheric models like WAM2layers. The Courant-Friedrichs-Lewy (CFL) condition is a critical consideration for the stability of the explicit solver used in WAM2layers.[1] This guide will walk you through the process of identifying and resolving issues related to high Courant numbers.

Step 1: Understanding the Courant Number in WAM2layers

The Courant number (C) is a dimensionless quantity that indicates how far a particle of water (or any tracer) moves in one time step relative to the size of the grid cell. For an explicit solver, as used in WAM2layers, the Courant number should not exceed 1 for the simulation to remain numerically stable.[1] If C > 1, it means that moisture is skipping over grid cells in a single time step, which can lead to inaccurate and unstable results.

Step 2: Identifying the Cause of High Courant Numbers

A high Courant number in WAM2layers is primarily caused by a combination of three factors:

• High Moisture Velocity: Strong winds will transport moisture more rapidly, increasing the Courant number.

• Small Grid Spacing: Finer resolution grids mean that moisture has less distance to travel to cross a grid cell. This is particularly problematic near the poles on a latitude-longitude grid due to the convergence of meridians.[1]

• Large Time Step: A larger time step allows more time for moisture to travel within a single iteration of the model.

Step 3: Resolving High Courant Number Issues

Follow these steps to address high Courant numbers in your WAM2layers experiments:

• Check Your Domain: WAM2layers is not recommended for use with input data on a latitude-longitude grid that extends beyond approximately 75° N/S.[1] The decreasing grid cell size towards the poles leads to prohibitively small time steps. If your region of interest is far from the poles, it is still advisable to exclude the polar regions from your computational domain to prevent spurious results.[1]

• Adjust the Time Step: The most direct way to reduce the Courant number is to decrease the time step of the simulation. While a very small time step can increase computation time and may lead to numerical diffusion, it is essential for stability.[1] WAM2layers does provide some internal limits on fluxes for extreme cases where the CFL criterion might be violated, but it is best practice to choose a time step that satisfies the condition for the vast majority of your simulation.[1]

• Evaluate Your Grid Resolution: If you are using a very high-resolution grid, you will need a correspondingly small time step. Consider if the grid resolution is appropriate for your research question and computational resources.

The following diagram illustrates the troubleshooting workflow for dealing with high Courant numbers in WAM2layers.

Troubleshooting workflow for high Courant numbers.

Frequently Asked Questions (FAQs)

Q1: What is the Courant number and why is it important in WAM2layers?

The Courant number (also known as the Courant-Friedrichs-Lewy or CFL number) is a dimensionless value that is crucial for the stability of numerical solutions to certain differential equations. In the context of WAM2layers, which uses an explicit time-stepping scheme, the Courant number should generally be kept below 1.[1] This ensures that the moisture being tracked does not travel more than one grid cell per time step, which is a prerequisite for a stable and accurate simulation.

Q2: How can I estimate an appropriate time step for my WAM2layers simulation?

To maintain a Courant number at or below 1, the time step should be chosen based on the moisture velocities and the grid spacing of your input data. The table below provides the maximum time step for a Courant number of 1, based on a 0.25° grid spacing as used with ERA5 input data.[1]

Moisture Velocity (m/s)	Maximum Time Step (hours) for 0.25° Grid Spacing
5	1.5
10	0.8
15	0.5
20	0.4
25	0.3
30	0.25

This table is derived from data presented in Kalverla et al. (2025) for the WAM2layers v3 model with ERA5 input data.[1]

Q3: What happens if I choose a time step that is too small?

While a smaller time step is generally safer for stability, an excessively small time step can have some drawbacks. It will significantly increase the computational time required to complete the simulation. Additionally, very small time steps can exacerbate numerical diffusion, which can artificially spread out the tracked moisture.[1] Therefore, there is a trade-off between ensuring stability and managing computational cost and numerical diffusion.

Q4: Can I use WAM2layers for polar regions?

It is strongly advised not to use WAM2layers with input data on a latitude-longitude grid that extends to the poles (beyond ~75° N/S).[1] The convergence of meridians at high latitudes leads to very small grid cell sizes in the zonal direction. To maintain a Courant number below 1, this would require an extremely small time step, making the simulation computationally impractical. Even if your primary area of interest is not at the poles, including them in the computational domain can introduce instabilities that may affect the results in your region of interest.[1]

Q5: Does WAM2layers have any built-in mechanisms to handle high Courant numbers?

Yes, WAM2layers does impose some limits on the fluxes to handle extreme circumstances where the CFL criterion might be violated.[1] However, it is not recommended to rely solely on these internal limits. The best practice is to choose a time step that is small enough to satisfy the CFL condition for the vast majority of the simulation period to ensure the most accurate and stable results.[1]

Experimental Protocols

The methodologies for running WAM2layers experiments are detailed in the official documentation and the reference publication for WAM2layers v3. The general workflow consists of two main steps:

• Preprocessing: In this step, the input data (such as ERA5 reanalysis data) is processed and collapsed onto a two-layer grid. The necessary terms for the moisture balance equations are reconstructed from the input data.[1]

• Tracking: This is the core of the WAM2layers model where the numerical integration is performed. The user can run the model in either a forward or backward tracking mode. This step requires a configuration file that specifies the tracking direction, date range, and paths to the preprocessed data and output directories.

For detailed instructions on setting up and running a WAM2layers experiment, users should consult the official documentation and the supplementary materials of the WAM2layers v3 publication.

References

• 1. GMD - Atmospheric moisture tracking with WAM2layers v3 [gmd.copernicus.org]

WAM2layers Technical Support Center: High-Resolution Climate Data Experiments

Welcome to the technical support center for WAM2layers. This guide provides troubleshooting advice and frequently asked questions (FAQs) to assist researchers, scientists, and drug development professionals in optimizing WAM2layers for high-resolution climate data experiments.

Frequently Asked Questions (FAQs)

Q1: What is WAM2layers?

A1: WAM2layers (Water Accounting Model - 2 layers) is an atmospheric moisture tracking model written in Python. It is used to determine the origin of precipitation (backtracking) or the destination of evaporated moisture (forward tracking). The model is particularly useful for understanding atmospheric dynamics and studying rainfall patterns in the context of climate change.[1][2][3] Version 3 of the model has been significantly rewritten to include performance optimizations for handling high-resolution input data, such as ERA5 reanalysis datasets.[3][4]

Q2: What are the main processing steps in a WAM2layers experiment?

A2: A WAM2layers experiment consists of two main stages: the preprocessing step and the tracking step.[5]

• Preprocessing: In this stage, the input data (e.g., from ERA5) is collapsed onto a two-layer vertical grid, and the necessary atmospheric variables are reconstructed for the model.[5]

• Tracking: This is the core of the model where the movement of "tagged" moisture is calculated. This can be done in two directions:

  • Forward tracking: Tracks where evaporated moisture will precipitate.

  • Backward tracking: Determines where precipitation originally evaporated.[6]

Q3: What are the primary input data requirements for WAM2layers with ERA5 data?

A3: To run WAM2layers with ERA5 data, you will need horizontal wind components (u and v), specific humidity (q), precipitation (P), evaporation (E), and surface pressure (ps). For pressure-level data, you will also need 10m wind components (u10 and v10) and 2m dew point temperature.

Experimental Protocols

Standard Methodology for a Backward Tracking Experiment

This protocol outlines the standard procedure for setting up and running a backward tracking experiment with WAM2layers using high-resolution ERA5 data.

• Installation: Install WAM2layers and its dependencies, preferably in a dedicated conda environment. You can install the package with visualization tools using pip:

• Data Acquisition: Download the required ERA5 data for your region and time period of interest. WAM2layers includes a utility to download example datasets to get started quickly.

• Configuration: Create a config.yaml file to define the parameters of your experiment. This file controls everything from file paths to the tracking period and region.

• Preprocessing: Run the preprocessing step to prepare your input data. This is a data-intensive and time-consuming part of the process.

• Tracking: Execute the backward tracking run. The model will read the preprocessed data and calculate the origins of the precipitation in your defined region.

• Visualization: Once the tracking is complete, you can visualize the results.

Below is a workflow diagram illustrating this experimental protocol.

Caption: Standard experimental workflow for WAM2layers.

Troubleshooting Guides

Issue: MemoryError during Preprocessing or Tracking

High-resolution climate data can be very memory intensive. A MemoryError indicates that your system is running out of RAM.

Logical Flow for Troubleshooting Memory Errors:

Caption: Decision tree for troubleshooting memory errors.

Solutions:

• Reduce the Domain: The most straightforward solution is to reduce the size of your tracking domain or the time period of your analysis in the config.yaml file. This will decrease the amount of data that needs to be loaded into memory at once.

• Check System Resources: Ensure that your machine has sufficient RAM for the resolution of the data you are processing. High-resolution datasets like ERA5 require significantly more memory than older datasets like ERA-Interim.

• Memory Profiling: For more advanced users, you can use Python's memory profiling tools to get a line-by-line analysis of memory consumption.[6][7][8][9][10] This can help identify if a specific function in your workflow is causing a memory leak. You can use the memory_profiler package by decorating the function you want to analyze with @profile.

Issue: Model Instability or Inaccurate Results with High-Resolution Data

When working with high-resolution data, you may encounter issues with the stability of the numerical solver, leading to unrealistic or inaccurate results.

Key Parameter to Check:

• Courant Number: For the numerical solver to be stable, the Courant number should not exceed one.[1] This means that moisture should not travel more than one grid cell per time step. With higher spatial resolution, the grid cells are smaller, and thus a smaller time step is often required to maintain stability.[1]

Optimization Parameters in config.yaml:

The following table summarizes key parameters in your config.yaml that you should consider adjusting when working with high-resolution data.

Parameter	Recommended Setting for High-Resolution Data	Rationale
target_frequency	A smaller value (e.g., '10min' for ERA5)	Higher spatial resolution requires a smaller time step to keep the Courant number below 1 and ensure model stability.[1]
preprocess_path	Path to a fast storage device (e.g., SSD)	The preprocessing step is I/O intensive. Using a fast disk can significantly reduce processing time.
region	As small as necessary for the research question	Reducing the geographical bounding box directly reduces the amount of data processed, leading to lower memory usage and faster run times.
start_date / end_date	Shorter time periods for initial tests	Before running a multi-year analysis, test your configuration on a shorter period to identify potential issues without waiting for a long processing time.

Note: It is highly recommended to perform a sensitivity analysis on the temporal and spatial resolutions to find the optimal settings for your specific research question and computational resources.[11]

References

• 1. GMD - Atmospheric moisture tracking with WAM2layers v3 [gmd.copernicus.org]
• 2. data.4tu.nl [data.4tu.nl]
• 3. EGUsphere - Atmospheric moisture tracking with WAM2layers v3 [egusphere.copernicus.org]
• 4. research.tudelft.nl [research.tudelft.nl]
• 5. forum.access-hive.org.au [forum.access-hive.org.au]
• 6. youtube.com [youtube.com]
• 7. Introduction to Memory Profiling in Python | DataCamp [datacamp.com]
• 8. analyticsvidhya.com [analyticsvidhya.com]
• 9. Memory profiling in Python using memory_profiler - GeeksforGeeks [geeksforgeeks.org]
• 10. memory-profiler · PyPI [pypi.org]
• 11. Theory — WAM2layers documentation [wam2layers.readthedocs.io]

WAM2layers Technical Support Center: Input Data Troubleshooting

This guide provides answers to common questions and solutions for issues related to input data formats for WAM2layers experiments.

Frequently Asked Questions (FAQs)

Q1: What are the fundamental input data requirements for a WAM2layers experiment?

A1: WAM2layers is primarily designed to work with gridded atmospheric datasets, such as reanalysis data (e.g., ERA5) or climate model output, in NetCDF format.[1][2] The preprocessing step requires specific physical variables to compute the moisture fluxes and balances.

The essential input variables required at the surface are:

• Precipitation (P)

• Evaporation (E)

• Surface Pressure (p_s)

For the atmospheric column (on either model or pressure levels), the following are required:

• Horizontal wind components (u and v)

• Specific humidity (q)

While not mandatory, providing the total column water (S_tc) is highly recommended as it helps estimate and mitigate errors.[1][2] All input data must be on the same horizontal grid.

Q2: My experiment is failing with errors suggesting a "grid mismatch." What does this mean and how can I fix it?

A2: A "grid mismatch" error typically means that the NetCDF files you are providing as input do not share the same horizontal spatial grid. WAM2layers requires that all input variables (e.g., precipitation, wind components, humidity) are defined on identical latitude and longitude coordinates for the preprocessing step to work correctly.

Common causes include:

• Using variables from different datasets or model runs that have different resolutions.

• Accidentally re-gridding only a subset of your input files.

• Using inconsistent longitude conventions (e.g., 0 to 360 vs. -180 to 180).

To resolve this, you must standardize all input files to a single target grid. This process is often called "regridding" or "remapping." See the protocol below for a detailed guide on how to perform regridding using Climate Data Operators (CDO).

Q3: I'm receiving a Coordinate variable not found or similar error. What is the cause?

A3: This error indicates that WAM2layers cannot find the expected coordinate variables (typically latitude, longitude, time) in your NetCDF file, or they are not in a recognizable format. WAM2layers, like many climate data tools, expects input files to adhere to certain conventions (such as the Climate and Forecast, CF, conventions).[3]

To troubleshoot this issue:

• Inspect the File: Use a command-line tool like ncdump -h your_file.nc or a Python library like xarray to inspect the header of your NetCDF file. Look for the names of the dimension and coordinate variables.

• Check Names: Ensure the coordinate variables are named correctly (e.g., latitude, not lat; longitude, not lon).

• Check Attributes: Verify that the coordinate variables have the appropriate units and axis attributes. For example, the time variable should have units like "hours since 1900-01-01" and a calendar attribute.

• Rename if Necessary: If the coordinate variables are incorrectly named, you can rename them using tools like ncrename from the NetCDF Operators (NCO) toolkit. For example: ncrename -v lat,latitude -v lon,longitude your_file.nc.

Q4: Can I use input data other than ERA5?

A4: Yes, while WAM2layers has been extensively used with ERA5, it is designed to be adaptable to other gridded atmospheric datasets, including output from various climate models.[4] However, doing so requires running the preprocessing step yourself.

To use a custom dataset, you will need to:

• Ensure your data contains all the required variables listed in Q1.

• Format your data into NetCDF files that follow standard conventions (see Q3).

• Confirm that all your input files are on the exact same spatial grid (see Q2).

• Be mindful of the model's limitations, especially regarding high latitudes. WAM2layers is not recommended for use with global latitude-longitude grids that extend beyond approximately 75° N/S due to numerical issues caused by the convergence of meridians near the poles.[1]

Data Presentation

The following table summarizes the standard data format for ERA5 inputs, which serves as a best-practice template for custom datasets.

Parameter	Data Type	Typical Dimensions	Required Variables	Notes
Atmospheric Levels	NetCDF	(time, level, latitude, longitude)	u, v, q	Data can be on model or pressure levels.[1]
Surface Data	NetCDF	(time, latitude, longitude)	P, E, p_s	Must be on the same lat/lon grid as the level data.
Coordinates	float	(latitude), (longitude)	latitude, longitude	Typically a regular latitude-longitude grid.
Time	double	(time)	time	Must include a units attribute (e.g., "hours since YYYY-MM-DD").

Experimental Protocols

Protocol: Resolving a Grid Mismatch Using CDO

This protocol provides a step-by-step method to unify the spatial grid of your input NetCDF files using Climate Data Operators (CDO).

Objective: To remap multiple input files (file_A.nc, file_B.nc) to match the grid of a reference file (reference_grid.nc).

Methodology:

• Identify the Target Grid: Choose one of your input files to serve as the reference for the target grid. Alternatively, you can define a target grid description file. For this example, we will use reference_grid.nc.

• Inspect the Grids (Optional but Recommended):

  • Verify that the grids are indeed different. Use the griddes operator in CDO.

  • cdo griddes file_A.nc > grid_A.txt

  • cdo griddes reference_grid.nc > grid_ref.txt

  • Compare the contents of grid_A.txt and grid_ref.txt. You will likely see differences in the xsize, ysize, xvals, or yvals.

• Perform the Remapping:

  • Use a conservative remapping method (remapcon) for flux variables and a bilinear interpolation (remapbil) for continuous variables like specific humidity. Conservative remapping is generally preferred for variables that represent quantities over a grid cell area (like precipitation) because it conserves the total amount of the quantity.

  • For a flux variable (e.g., precipitation in file_A.nc):

  • For a continuous variable (e.g., specific humidity in file_B.nc):

• Verification:

  • Inspect the grid of the newly created file to ensure it matches the reference grid.

  • cdo griddes file_A_regridded.nc > grid_A_new.txt

  • Confirm that grid_A_new.txt is now identical to grid_ref.txt.

• Repeat: Repeat the remapping step for all input files that do not match the target grid before passing them to the WAM2layers preprocessing script.

Visualizations

Logical Workflow: Troubleshooting Input Data Errors

The following diagram illustrates a logical workflow for diagnosing and resolving common input data format issues during the WAM2layers preprocessing stage.

A flowchart for diagnosing WAM2layers input data errors.

References

• 1. GMD - Atmospheric moisture tracking with WAM2layers v3 [gmd.copernicus.org]
• 2. research.tudelft.nl [research.tudelft.nl]
• 3. research.wur.nl [research.wur.nl]
• 4. researchgate.net [researchgate.net]

WAM2layers numerical stability considerations

This technical support center provides troubleshooting guides and frequently asked questions (FAQs) to assist researchers, scientists, and drug development professionals in their experiments with the WAM2layers model.

Troubleshooting Guides

This section addresses specific issues that may arise during WAM2layers experiments, providing step-by-step guidance to identify and resolve them.

Issue: Model run fails with an error during the preprocessing stage.

Answer:

Errors during the wam2layers preprocess stage are often related to the input data. Follow these steps to diagnose the issue:

• Verify Input Data Integrity: Ensure that your input data files (e.g., from ERA5 or other climate models) are not corrupted and contain all the necessary variables.[1] WAM2layers v3 is primarily focused on working with ERA5 reanalysis data.[1][2]

• Check Configuration File: Open your config.yaml file and verify the following:

  • The file paths to your input data are correct.

  • The variable names in the configuration file match the variable names in your input data files.

  • The date range specified for preprocessing is covered by your input data.

• Examine Data Consistency: Inconsistencies in the input data can lead to errors.[1] The model includes an error term to account for inaccuracies and numerical errors in the input data.[1] Check for and handle any missing values or temporal gaps in your dataset.

• Consult Logs: The WAM2layers model creates a log file for each experiment. This log file will contain detailed error messages that can help pinpoint the exact cause of the failure.

Issue: The output moisture fluxes from my experiment seem physically unrealistic or show unexpected patterns.

Answer:

Unrealistic output can stem from several factors, from the model configuration to the specific atmospheric conditions of your study area.

• Review Tracking Direction: Confirm that you have selected the correct tracking direction for your research question.[3]

  • Forward tracking: Tracks where evaporated moisture precipitates.[3][4]

  • Backward tracking: Tracks the evaporative source of precipitation.[3][4]

• Assess the Impact of Wind Shear: The WAM2layers model uses a two-layer vertical structure to better handle conditions with strong directional wind shear, a known source of large errors in single-layer models.[1] If your study region and period are characterized by strong wind shear, results might still be sensitive to how the vertical layers are defined and integrated.

• Check for Water Balance Closure: While WAM2layers is a water accounting model, discrepancies can arise. The model's numerical implementation includes an error term to account for this.[1] Large, persistent imbalances might indicate an issue with the input forcing data (precipitation and evaporation).

• Visualize Intermediate Fields: If possible, examine the intermediate preprocessed fields, such as the vertically integrated moisture and fluxes, to ensure they are reasonable before running the full tracking simulation.

Frequently Asked Questions (FAQs)

Q1: What is WAM2layers and what are its main applications?

A1: WAM2layers (Water Accounting Model - 2 layers) is an atmospheric moisture tracking model.[5][6] It is used to understand atmospheric dynamics and to study rainfall patterns and extremes by identifying their sources (backward tracking) or their ultimate fate (forward tracking).[1][2][5][7]

Q2: What is the difference between forward and backward tracking?

A2:

• Forward tracking follows moisture from a source region (evaporation) to see where it subsequently precipitates.[3]

• Backward tracking starts from a sink region (precipitation) to identify the original evaporative sources of that moisture.[3]

Q3: What input data is required to run WAM2layers?

A3: WAM2layers is designed to work with gridded atmospheric datasets, such as reanalysis data (e.g., ERA5) or output from climate models.[1][2][5] The necessary input variables typically include wind components, humidity, and surface fluxes of precipitation and evaporation. For pressure-level data, surface values of wind and humidity are also required.[1]

Q4: Can I use my own climate model output with WAM2layers?

A4: Yes, while WAM2layers v3 has a primary focus on ERA5 data, it can be adapted for other datasets.[1][2] This requires modifications to the dataset-dependent preprocessing module to correctly map your model's output to the two-layer structure used by WAM2layers.[1][2]

Q5: How does WAM2layers handle numerical stability?

A5: The numerical implementation of WAM2layers v3 has been rewritten to improve code efficiency, especially for high-resolution input data.[1][2][7] The model's equations include an error term to account for numerical errors and inconsistencies in the input data, contributing to a more robust water balance calculation.[1]

Experimental Protocols & Data

Key Experimental Parameters

The following table summarizes key parameters in the WAM2layers configuration file that influence model behavior and numerical stability.

Parameter	Description	Recommended Setting / Considerations
tracking_direction	Specifies whether to run the model in 'forward' or 'backward' mode.	Choose based on your research question.
preprocess_start_date	The start date for the preprocessing period.	Ensure this is within the bounds of your input dataset.
preprocess_end_date	The end date for the preprocessing period.	Ensure this is within the bounds of your input dataset.
track_start_date	The start date for the tracking simulation.	Must be within the preprocessed date range.
track_end_date	The end date for the tracking simulation.	Must be within the preprocessed date range.
input_folder	Path to the directory containing the raw input data.	Verify the path is correct.
output_folder	Path to the directory where results will be saved.	Ensure you have write permissions to this directory.
tagging_region	A file or definition of the geographical area for tagging moisture.	Ensure the format is compatible and the region is correctly defined.

Visualizations

Logical and Experimental Workflows

The following diagrams illustrate key workflows in WAM2layers.

WAM2layers Forward Tracking Workflow

WAM2layers Backward Tracking Workflow

References

• 1. research.tudelft.nl [research.tudelft.nl]
• 2. Cracking the (moisture) tracking code | by Peter Kalverla | Netherlands eScience Center [blog.esciencecenter.nl]
• 3. Tracking — WAM2layers documentation [wam2layers.readthedocs.io]
• 4. Welcome to WAM2layers! — WAM2layers documentation [wam2layers.readthedocs.io]
• 5. GitHub - WAM2layers/WAM2layers: Atmospheric moisture tracking [github.com]
• 6. GMD - Atmospheric moisture tracking with WAM2layers v3 [gmd.copernicus.org]
• 7. EGUsphere - Atmospheric moisture tracking with WAM2layers v3 [egusphere.copernicus.org]

best practices for long-term WAM2layers simulations

This technical support center provides troubleshooting guides and frequently asked questions (FAQs) to assist researchers, scientists, and drug development professionals in conducting long-term WAM2layers simulations.

Frequently Asked Questions (FAQs)

Q1: What are the key considerations before starting a long-term WAM2layers simulation?

A1: Before initiating a long-term simulation, it is crucial to:

• Ensure Input Data Quality: Use high-quality, consistent, and complete input data, such as ERA5 reanalysis data.[1] Gaps or inconsistencies in the input data can lead to model instability and unrealistic results over long simulation periods.

• Define an Appropriate Tracking Domain: The tracking domain should be large enough to encompass the expected moisture sources or sinks for your research question. WAM2layers logs and reports transport over the domain boundary, which can help you determine if a larger domain is needed for subsequent runs.[2]

• Select a Suitable Time Step: The internal time step of the model is a critical parameter. A smaller time step increases computational cost but can improve numerical stability. It is advisable to perform sensitivity analyses to determine the optimal time step for your specific application and input data resolution.

• Plan for Data Storage: Long-term simulations generate large volumes of output data. Ensure you have sufficient storage capacity and a clear data management plan.

Q2: How can I improve the performance and stability of my long-term WAM2layers simulation?

A2: To enhance performance and stability:

• Use the Latest Version: WAM2layers v3 is a complete rewrite with performance optimizations and improved user-friendliness.[2][3]

• Optimize Your Configuration: The simulation's configuration file allows you to control various settings, including the tracking direction and output frequency.[4] Writing output at less frequent intervals can reduce I/O overhead and improve performance.

• Leverage Pre-processed Data: If available for your region and period of interest, using pre-processed input data can significantly speed up the simulation setup.

• Monitor Your Simulation: WAM2layers v3 includes log messages to monitor the experiment's progress.[1][2] Regularly checking the log files can help you identify and address potential issues early on.

Q3: Does WAM2layers handle leap years in long-term simulations?

A3: While the search results do not provide specific details on how WAM2layers handles leap years, it is a critical consideration for any long-term climate or atmospheric modeling. In general, meteorological and climate models must correctly account for the extra day in a leap year to maintain accurate time series analysis. It is recommended to consult the official WAM2layers documentation or user community for the most up-to-date information on this topic.

Troubleshooting Guide

Problem 1: My long-term simulation is running very slowly.

Question	Answer
Why is my simulation so slow?	Long-term simulations with high-resolution data are computationally intensive. The performance can be affected by the size of your domain, the length of the simulation, the frequency of output, and the efficiency of your storage system.
How can I speed it up?	1. Reduce Output Frequency: In your configuration file, decrease the frequency at which output files are written. 2. Optimize Domain Size: Use the smallest tracking domain that is scientifically sound for your research question. 3. Use a More Powerful Computing Environment: Run your simulation on a machine with more CPU cores, memory, and faster storage. 4. Check for I/O Bottlenecks: Ensure that your storage system can handle the large volume of data being read and written.

Problem 2: My simulation crashed unexpectedly.

Question	Answer
What are the common causes of crashes in long-term simulations?	Crashes can be caused by a variety of factors, including: - Input Data Errors: Corrupted or missing input files. - Numerical Instability: The model solution may diverge due to an inappropriate time step or issues with the input data. - Memory Issues: The simulation may run out of memory, especially with high-resolution data over a large domain. A "signal 9: SIGKILL" error often indicates a memory-related issue.[5] - Configuration Errors: Incorrect settings in the configuration file.
How do I diagnose the cause of the crash?	1. Examine the Log Files: WAM2layers generates log files that contain information about the simulation's progress and any errors encountered.[1][4] 2. Check Input Data: Verify the integrity and completeness of your input data files for the time step at which the simulation crashed. The GitHub issues for WAM2layers suggest that input validation is a common concern.[6] 3. Monitor System Resources: During a rerun, monitor your system's memory and CPU usage to see if you are hitting any limits.
How can I fix the crash?	- Correct Input Data: If the issue is with the input data, you will need to correct or replace the faulty files. - Adjust Time Step: If you suspect numerical instability, try reducing the time step in your configuration file. - Increase Memory: If the crash is due to a lack of memory, try running the simulation on a machine with more RAM or reduce the domain size. - Review Configuration: Carefully check your configuration file for any errors or typos.

Problem 3: The output from my long-term simulation looks unrealistic.

Question	Answer
Why do my results look wrong?	Unrealistic results can stem from: - Incorrect "tagging" region: The region for which moisture sources or sinks are being tracked may be defined incorrectly. The WAM2layers GitHub issues indicate that the tracking script should check for tagging_region values being between 0 and 1.[6] - Boundary Effects: If a significant amount of moisture is transported across the boundaries of your tracking domain, your results may be incomplete.[2] - Model Assumptions: WAM2layers is a two-layer model, which is a simplification of the real atmosphere.[1] For some research questions, this level of simplification may not be appropriate.
How can I validate my results?	1. Analyze Spatial Fields: WAM2layers v3 can output spatial fields for later analysis. Visualizing these fields can help you identify any anomalies.[1] 2. Check Boundary Transport: The log files report on transport across the domain boundaries. If this is a large component of your moisture budget, you may need to expand your domain.[2] 3. Compare with Other Data: Compare your results with other available data sources or previous studies to see if they are consistent. 4. Perform Sensitivity Studies: Run the simulation with different settings (e.g., different time steps, domain sizes) to see how sensitive your results are to these choices.

Experimental Protocols

Protocol: Setting up a Robust Long-Term WAM2layers Simulation

• Input Data Preparation:

  • Acquire a complete and consistent set of input data (e.g., ERA5) for the entire simulation period.

  • Carefully check the data for any gaps or corrupted files.

  • If necessary, preprocess the data to the required format for WAM2layers.

• Configuration File Setup:

  • Create a clear and well-documented configuration file for your experiment.

  • Specify the tracking_direction (forward or backward).

  • Define the date_range for the entire long-term simulation.

  • Set the preprocess_dir to the location of your input data and output_dir for the results.

  • Define the tagging_region for your area of interest.

  • Choose an appropriate output_frequency. For long-term runs, daily or monthly outputs are often sufficient.

  • Start with a conservative time_step and consider performing a sensitivity analysis to find the optimal value.

• Initial Short Test Run:

  • Before launching the full long-term simulation, perform a short test run (e.g., for one month or one year).

  • Verify that the model runs to completion without errors.

  • Analyze the output and log files from the test run to ensure everything looks as expected.

• Full Long-Term Simulation and Monitoring:

  • Launch the full long-term simulation.

  • Periodically monitor the progress of the simulation by checking the log files.

  • Keep an eye on system resources (CPU, memory, disk space).

• Restart Functionality (if needed):

  • WAM2layers has a restart functionality. If your simulation is interrupted, you may be able to restart it from the last saved state. Be aware that the GitHub issues suggest there can be problems with the restart function, so it is important to verify the integrity of the output after a restart.[6]

Visualizations

Caption: High-level workflow for a WAM2layers experiment.

Caption: A logical flow for troubleshooting WAM2layers simulations.

References

• 1. research.tudelft.nl [research.tudelft.nl]
• 2. GMD - Atmospheric moisture tracking with WAM2layers v3 [gmd.copernicus.org]
• 3. doaj.org [doaj.org]
• 4. egusphere.copernicus.org [egusphere.copernicus.org]
• 5. python - Memory leak when trying to regularize an uneven time series with Pandas - Stack Overflow [stackoverflow.com]
• 6. GitHub · Where software is built [github.com]

Reference Data & Comparative Studies

Validating WAM2layers: A Comparative Guide for Atmospheric Moisture Tracking

An in-depth look at the validation of the WAM2layers model against observational data reveals a scientific field grappling with the inherent complexities of tracking atmospheric moisture. While direct validation is challenging due to a scarcity of ground-truth data, a recent community-wide intercomparison study is beginning to shed light on the model's performance and uncertainties.

The Water Accounting Model - 2 layers (WAM2layers) is an Eulerian atmospheric moisture tracking model widely used to understand the sources and sinks of precipitation. It operates by solving prognostic equations for tagged moisture within gridded atmospheric datasets, such as the widely used ERA5 reanalysis.[1][2] The model can be run in both forward and backward modes to determine where evaporated water will precipitate and where precipitation originated, respectively.[3]

The Challenge of "Ground Truth" in Moisture Tracking

A primary hurdle in validating any atmospheric moisture tracking model is the lack of direct, comprehensive observational data for the pathways of atmospheric water vapor.[4][5][6] Unlike weather models that can be validated against a dense network of surface and satellite observations of temperature, pressure, and precipitation, tracking the origin of a specific rainfall event is a more complex endeavor.

The Moisture Tracking Intercomparison Study: A Step Towards Consensus

To address the uncertainties inherent in different modeling approaches, the moisture tracking community initiated a coordinated intercomparison study.[4][5][7] This effort, with preliminary results presented at the EMS Annual Meeting in 2024, brought together researchers using approximately eight different moisture tracking models, including WAM2layers, to analyze three extreme precipitation events: a monsoon event in Pakistan, a convective precipitation event in Australia, and an atmospheric river-driven event over Scotland.[4]

The goal of this intercomparison is to quantify the uncertainty in moisture source attribution and to provide a reference for future studies.[4][5] While the detailed quantitative results from this study are not yet fully published, the initiative represents the most significant effort to date to benchmark the performance of models like WAM2layers.

Experimental Protocol: A Generalized Validation Workflow

In the absence of direct observational data for moisture pathways, the validation and comparison of WAM2layers typically relies on a multi-faceted approach. The general experimental protocol can be summarized as follows:

• Model Input: WAM2layers is forced with gridded atmospheric data, most commonly the ERA5 reanalysis dataset from the European Centre for Medium-Range Weather Forecasts (ECMWF). This dataset provides hourly estimates of a large number of atmospheric, land, and oceanic climate variables.

• Model Configuration: The model is configured to run in either forward or backward mode for a specific time period and geographical region of interest.

• Inter-model Comparison: The primary method for evaluating WAM2layers is to compare its outputs with those of other moisture tracking models, particularly Lagrangian models like FLEXPART.[3] These comparisons focus on the spatial and temporal distribution of identified moisture sources.

• Process-based Evaluation: The plausibility of the model's results is assessed by examining the physical consistency of the identified moisture sources with known meteorological phenomena, such as atmospheric rivers and monsoonal flows.

• Potential for Isotope Comparison: The use of stable water isotopes (¹⁸O and ²H) in water vapor is an emerging field that holds promise for validating moisture tracking models.[4][8] The distinct isotopic signatures of water from different sources could provide a much-needed observational constraint. However, the application of this technique for the direct validation of WAM2layers has not yet been extensively documented in published literature.

Below is a conceptual diagram of the validation workflow for WAM2layers.

A conceptual workflow for the validation of WAM2layers results.

Quantitative Data Summary

As the detailed quantitative results from the moisture tracking intercomparison study are not yet publicly available, a comprehensive table of performance metrics cannot be provided at this time. The study's findings, once published, are expected to offer the first systematic quantitative comparison of WAM2layers against a suite of other models.

Eulerian vs. Lagrangian Models: A Tale of Two Perspectives

Much of the current understanding of WAM2layers' performance comes from comparisons with Lagrangian models. These two classes of models differ fundamentally in their approach to tracking moisture:

• Eulerian Models (like WAM2layers): These models use a fixed grid and calculate the movement of moisture between grid cells at each time step.

• Lagrangian Models: These models track the movement of a large number of individual air parcels as they are transported through the atmosphere.

These differing methodologies can lead to discrepancies in the identified moisture sources, and understanding these differences is a key area of ongoing research.

The following diagram illustrates the logical relationship between the data source, modeling approaches, and the eventual goal of validating atmospheric moisture tracking.

Logical flow from data to model validation in moisture tracking.

Future Directions

The field of atmospheric moisture tracking is actively working towards more robust validation methods. The results of the ongoing intercomparison study will be a significant step forward. Furthermore, the integration of stable water isotope data into validation workflows holds the potential to provide a much-needed observational benchmark to constrain and improve the accuracy of models like WAM2layers. As these methods mature, a clearer picture of the strengths and limitations of WAM2layers in tracking the intricate pathways of moisture through the Earth's atmosphere will emerge.

References

• 1. research.tudelft.nl [research.tudelft.nl]
• 2. GMD - Atmospheric moisture tracking with WAM2layers v3 [gmd.copernicus.org]
• 3. researchgate.net [researchgate.net]
• 4. Abstract EMS2024-1040 [meetingorganizer.copernicus.org]
• 5. research.wur.nl [research.wur.nl]
• 6. research.wur.nl [research.wur.nl]
• 7. lorentzcenter.nl [lorentzcenter.nl]
• 8. publikationen.bibliothek.kit.edu [publikationen.bibliothek.kit.edu]

A Comparative Guide to WAM2layers and HYSPLIT for Atmospheric Moisture Tracking

For Researchers, Scientists, and Drug Development Professionals

Atmospheric moisture tracking is a critical tool for understanding a wide range of phenomena, from the sources of precipitation that can impact agricultural yields and water resources to the transport of airborne particles relevant in various scientific and industrial contexts. Among the computational models available for these tasks, WAM2layers and HYSPLIT are two prominent tools, each employing distinct methodologies. This guide provides an objective comparison of their performance, methodologies, and data requirements, supported by a proposed experimental framework for direct comparison.

At a Glance: WAM2layers vs. HYSPLIT

The fundamental difference between WAM2layers and HYSPLIT lies in their atmospheric transport calculation approach. WAM2layers is an Eulerian model, which calculates moisture fluxes through a fixed grid over time. In contrast, HYSPLIT is a Lagrangian model that tracks the movement of individual air parcels through the atmosphere. This core distinction influences their applications, data handling, and the interpretation of their results.

Feature	WAM2layers (Water Accounting Model - 2 layers)	HYSPLIT (Hybrid Single-Particle Lagrangian Integrated Trajectory)
Modeling Approach	Eulerian	Lagrangian
Primary Function	Quantifies the contribution of evaporation from a source region to precipitation in a target region (and vice versa).	Computes the trajectories of air parcels to determine their origin or destination.
Moisture Tracking	Solves a prognostic equation for "tagged" moisture within a two-layer atmospheric model.[1][2]	Infers moisture transport by analyzing changes in specific humidity along back trajectories.
Typical Use Cases	Moisture recycling studies, identifying evaporative sources for precipitation events.[2][3]	Source-receptor analysis for precipitation, air pollution, and other atmospheric constituents.[4][5]
Input Data	Gridded meteorological data (e.g., from ERA5 reanalysis) including wind fields, humidity, surface pressure, precipitation, and evaporation.[6]	Gridded meteorological data (e.g., from GDAS, NCEP/NCAR reanalysis) including wind fields, temperature, pressure, and humidity.[7][8]
Output	Gridded data showing the spatial distribution of tracked moisture and its contribution to precipitation or from evaporation.	Trajectory paths of individual air parcels, along with meteorological variables at each point along the trajectory.
Strengths	Provides a quantitative moisture budget for defined regions; conserves water mass within the model domain.	Computationally efficient for tracking from a limited number of locations; versatile for various atmospheric transport applications.[5]
Limitations	Can be computationally intensive for high-resolution, long-duration simulations; results can be sensitive to the vertical resolution of the model.[9]	Trajectory accuracy can decrease with longer simulation times; may overestimate the contribution of local moisture sources.[7]

Conceptual Workflow Comparison

The distinct methodologies of WAM2layers and HYSPLIT result in different analytical workflows for identifying moisture sources for a given precipitation event.

Comparative workflows for moisture source identification.

Experimental Protocol for a Comparative Study

To quantitatively compare the performance of WAM2layers and HYSPLIT, a detailed experimental protocol is necessary. The following methodology outlines a framework for such a comparison, focusing on a historical heavy precipitation event.

1. Case Study Selection:

• Select a well-documented heavy precipitation event with available high-resolution meteorological data. The event should have a clear spatial and temporal extent.

2. Meteorological Data:

• Utilize a consistent, high-resolution reanalysis dataset, such as ERA5, as the input for both models. This ensures that any differences in output are attributable to the models' methodologies and not the input data.

3. WAM2layers Simulation Setup:

• Preprocessing: Prepare the ERA5 data according to the requirements of WAM2layers, including horizontal wind components (u and v), specific humidity (q), surface pressure, precipitation, and evaporation.[6]

• Backward Tracking: Configure and run WAM2layers in backward mode from the defined precipitation region and time. The output will be a gridded dataset showing the contribution of evaporation from each grid cell to the precipitation event.

4. HYSPLIT Simulation Setup:

• Receptor Points: Define a grid of receptor points within the precipitation area.

• Back Trajectory Calculation: For each receptor point, run HYSPLIT to calculate back trajectories for a specified duration (e.g., 7-10 days). Multiple starting altitudes should be used to account for vertical atmospheric structure.

• Moisture Uptake Analysis: Analyze the meteorological variables (specifically, specific humidity) along each trajectory. An increase in specific humidity between two points indicates a region of moisture uptake.

5. Data Analysis and Comparison:

• Spatial Comparison: Spatially compare the identified moisture source regions from both models. This can be done by overlaying the maps of evaporative sources from WAM2layers with the regions of moisture uptake identified by HYSPLIT.

• Quantitative Metrics: Where possible, quantify the contribution from different source regions (e.g., oceanic vs. terrestrial, local vs. remote) as determined by each model.

• Discrepancy Analysis: Analyze any discrepancies in the context of the known differences between Eulerian and Lagrangian models. For instance, investigate if HYSPLIT shows a stronger signal from local sources compared to WAM2layers.

Logical Relationship of Methodologies

The core difference in the logical approach of WAM2layers and HYSPLIT can be visualized as a contrast between a field-based and a particle-based perspective.

Fundamental methodological differences.

Conclusion

Both WAM2layers and HYSPLIT are powerful tools for atmospheric moisture tracking, each with its own set of strengths and limitations rooted in their underlying Eulerian and Lagrangian frameworks. WAM2layers provides a robust, mass-conserving quantification of moisture sources and sinks, making it well-suited for detailed hydrological cycle studies. HYSPLIT offers a versatile and computationally efficient method for tracing atmospheric pathways, with broad applications beyond moisture tracking.

The choice between the two models will depend on the specific research question. For studies requiring a detailed and quantitative moisture budget of a region, WAM2layers is a strong candidate. For exploratory analyses of potential source regions and transport pathways, HYSPLIT provides a valuable and accessible tool. As the field of atmospheric science advances, direct comparative studies, such as the one outlined in the proposed experimental protocol, will be crucial for a more comprehensive understanding of the uncertainties and capabilities of these different modeling approaches.

References

• 1. doaj.org [doaj.org]
• 2. GMD - Atmospheric moisture tracking with WAM2layers v3 [gmd.copernicus.org]
• 3. researchgate.net [researchgate.net]
• 4. m.youtube.com [m.youtube.com]
• 5. HYSPLIT – Air Resources Laboratory [arl.noaa.gov]
• 6. egusphere.copernicus.org [egusphere.copernicus.org]
• 7. acp.copernicus.org [acp.copernicus.org]
• 8. ww2.arb.ca.gov [ww2.arb.ca.gov]
• 9. Theory — WAM2layers documentation [wam2layers.readthedocs.io]

Sensitivity Analysis of WAM2layers Parameters: A Comparative Guide

For Researchers, Scientists, and Drug Development Professionals

This guide provides an objective comparison of the Water Accounting Model 2-layers (WAM2layers), a widely used Eulerian atmospheric moisture tracking model, with a prominent Lagrangian alternative. It focuses on the sensitivity of WAM2layers to key input parameters and presents supporting experimental data to inform model selection and experimental design.

Introduction to WAM2layers and Atmospheric Moisture Tracking

Atmospheric moisture tracking is a crucial technique for understanding the hydrological cycle, including the sources and pathways of precipitation. WAM2layers is a computationally efficient model that tracks atmospheric moisture by solving prognostic equations for tagged moisture in a two-layer representation of the atmosphere.[1][2] It can be run in forward mode to determine where evaporated moisture precipitates, or in backward mode to identify the evaporative sources of precipitation.[1][2] The latest version, WAM2layers v3, includes performance optimizations to handle high-resolution input data and improved user-friendliness.[1]

Alternative approaches, such as Lagrangian models like FLEXPART-WaterSip, track the trajectories of individual air parcels backward or forward in time to determine moisture sources and sinks. The fundamental difference between Eulerian (grid-based) and Lagrangian (parcel-based) models can lead to different estimations of moisture contributions, particularly in regions with complex topography.[3][4]

Sensitivity of WAM2layers to Key Parameters

The performance and output of WAM2layers are sensitive to several key parameters, primarily related to the spatial and temporal resolution of the input meteorological data and the definition of the vertical layers. Understanding this sensitivity is critical for robust experimental design and accurate interpretation of results. The WAM2layers documentation recommends a detailed sensitivity study presented in a Master's thesis, the findings of which are summarized and expanded upon here with data from other relevant studies.

Impact of Input Data Resolution

The resolution of the input data, both spatially and temporally, significantly influences the moisture tracking results.

Key Findings:

• Higher resolution generally improves accuracy: Studies comparing WAM2layers with other models have shown that increasing the spatial and temporal resolution of the forcing data tends to reduce the discrepancies between different models.[3]

• Temporal resolution may be more critical than spatial resolution: For WAM2layers, prioritizing a higher temporal resolution over spatial resolution is recommended for moisture tracking over complex terrains like the Tibetan Plateau.[3]

• Precipitation bias reduction: Increasing the horizontal resolution of the input General Circulation Model (GCM) data has been shown to improve precipitation simulations, which in turn affects the accuracy of moisture source tracking.[5]

Quantitative Data Summary:

Parameter	Variation	Impact on Moisture Contribution Estimation	Source
Spatial Resolution	Increased from N96 to N216 in MetUM GCM	Reduced precipitation biases over the Tibetan Plateau, leading to a more accurate representation of local versus remote moisture sources.[5]	[5]
Temporal Resolution	Not explicitly quantified in available literature	Higher temporal resolution is recommended to better capture the dynamics of moisture transport, especially in regions with complex atmospheric circulation.[3]	[3]

Sensitivity to Vertical Layer Definition

WAM2layers simplifies the atmosphere into two layers. The definition of the boundary between these layers is a critical parameter.

Default Setting:

• The default boundary is set at the interface between IFS (Integrated Forecasting System) levels 111 and 112, which corresponds to approximately 812 hPa in a standard atmosphere.[1] This is typically just above the atmospheric boundary layer.[1]

Impact of Variation:

• The one-layer version of the model was found to produce large errors in cases with strong directional wind shear, highlighting the importance of the two-layer structure.[1]

Comparison with Alternative Models: WAM2layers vs. FLEXPART-WaterSip

FLEXPART-WaterSip is a widely used Lagrangian model that provides a contrasting approach to WAM2layers.

Key Differences:

• Model Type: WAM2layers is an Eulerian model, while FLEXPART is a Lagrangian model.[3][4]

• Moisture Source Estimation: WAM2layers tends to estimate higher moisture contributions from distant, westerly-dominated sources and lower contributions from local recycling compared to FLEXPART-WaterSip.[3][4]

• Alignment with Evaporation: WAM2layers shows a closer alignment of estimated moisture sources with actual evaporation, particularly in source regions with complex land-sea distributions.[3]

• Computational Cost: While not explicitly quantified in the compared literature, Eulerian models like WAM2layers are generally considered computationally less expensive than Lagrangian models for global, long-term simulations.

Performance Comparison Summary:

Feature	WAM2layers (Eulerian)	FLEXPART-WaterSip (Lagrangian)	Key Considerations
Moisture Source Attribution	Higher contribution from distant sources, lower from local recycling.[3][4]	Higher contribution from local recycling and nearby sources.[3][4]	The choice of model can significantly impact the attribution of precipitation sources.
Accuracy in Complex Terrain	Higher temporal resolution is crucial.[3]	May better capture vertical moisture transport in complex terrain.	Model selection should consider the specific geography of the study area.
Computational Efficiency	Generally more efficient for large-scale, long-term studies.	Can be computationally intensive, especially for high-resolution, global simulations.	The scale of the research question will influence the choice of model.
Evaporation Bias	Estimated moisture sources align well with actual evaporation.[3]	Can have biases in evaporation estimates that may require correction.[3]	WAM2layers may have an advantage in regions with heterogeneous surfaces.

Experimental Protocols

This section outlines a general methodology for conducting a sensitivity analysis of WAM2layers parameters, based on the principles described in the referenced literature.

Objective: To quantify the impact of varying input data resolution and vertical layer definition on the moisture source attribution of a specific precipitation event or region.

Methodology:

• Define a Baseline Experiment:

  • Select a historical precipitation event or a climatological period of interest.

  • Define the geographical domain for the study.

  • Run WAM2layers with a standard, well-documented set of parameters (e.g., ERA5 reanalysis data at a specific resolution, default vertical layer definition). This will serve as the control run.

• Vary Input Data Resolution:

  • Spatial Resolution: Obtain input meteorological data at different spatial resolutions (e.g., 0.25°, 0.5°, 1.0°). Run WAM2layers for each resolution while keeping all other parameters constant.

  • Temporal Resolution: Obtain input data at different temporal frequencies (e.g., 1-hourly, 3-hourly, 6-hourly). Run WAM2layers for each temporal resolution while keeping all other parameters constant.

• Vary Vertical Layer Definition:

  • Modify the pressure level that defines the boundary between the two vertical layers in WAM2layers.

  • Perform a series of runs with systematically varied boundary levels (e.g., shifting the boundary up or down by a certain pressure interval).

• Data Analysis and Comparison:

  • For each experimental run, calculate the moisture source contributions from predefined source regions.

  • Quantify the differences in moisture source attribution between each sensitivity run and the baseline experiment. Express these differences as both absolute and relative changes.

  • Analyze the spatial patterns of the differences to identify regions where the model is most sensitive to parameter changes.

  • If available, compare the results to observational data or outputs from other models (e.g., FLEXPART-WaterSip) to assess the realism of the simulations.

Visualizations

Logical Workflow for WAM2layers Sensitivity Analysis

Caption: Workflow for WAM2layers sensitivity analysis.

Signaling Pathway of Moisture Source Attribution

Caption: Influence of parameters on moisture source attribution.

References

• 1. GMD - Atmospheric moisture tracking with WAM2layers v3 [gmd.copernicus.org]
• 2. researchgate.net [researchgate.net]
• 3. researchgate.net [researchgate.net]
• 4. researchgate.net [researchgate.net]
• 5. research.tudelft.nl [research.tudelft.nl]

Unveiling the Vertical Dimension: A Comparative Guide to Atmospheric Moisture Tracking and the Limitations of the WAM2layers Two-Layer Assumption

For researchers, scientists, and drug development professionals venturing into atmospheric moisture tracking, the choice of model is paramount. This guide provides an objective comparison of the widely-used Water Accounting Model - 2 layers (WAM2layers) with a prominent alternative, the FLEXible PARTicle dispersion model (FLEXPART), highlighting the inherent limitations of WAM2layers' foundational two-layer assumption. By presenting quantitative data from peer-reviewed studies and detailing experimental protocols, this document aims to inform model selection for robust and accurate atmospheric research.

Introduction to Atmospheric Moisture Tracking

Atmospheric moisture tracking is a crucial tool for understanding the hydrological cycle, identifying precipitation sources, and assessing the impact of climate change on water resources.[1] These models can be broadly categorized into two main approaches: Eulerian and Lagrangian.[1]

• Eulerian models , such as WAM2layers, employ a fixed grid system and quantify the movement of moisture between grid cells over time.[2][3]

• Lagrangian models , like FLEXPART, track the movement of individual air parcels through the atmosphere, providing a trajectory-based perspective on moisture transport.[2][3]

WAM2layers, an offline, Eulerian model, simplifies the complex vertical structure of the atmosphere into two distinct layers to address errors observed in single-layer models, particularly in the presence of strong directional wind shear. This simplification, however, introduces limitations that are critical to consider for specific research applications.

Comparison of WAM2layers and FLEXPART-WaterSip

A key alternative to the grid-based approach of WAM2layers is the Lagrangian model FLEXPART, often coupled with the "WaterSip" method for moisture source analysis. The fundamental difference in their approach—fixed grid versus moving parcels—leads to notable variations in their depiction of moisture source-receptor relationships.

Quantitative Performance Evaluation

A study conducted on the Tibetan Plateau provides a quantitative comparison of moisture source contributions to precipitation in two distinct river basins as estimated by WAM2layers and FLEXPART-WaterSip. The results highlight that WAM2layers tends to attribute a larger proportion of moisture to distant sources, whereas FLEXPART-WaterSip emphasizes the role of local recycling.

Moisture Source Region	Yarlung Zangbo Basin (% contribution)	Upper Tarim River Basin (% contribution)
WAM2layers	FLEXPART-WaterSip
Tibetan Plateau (Local)	32%	53%
Southern Indian Ocean (Distant)	30%	11%
Middle East & West Asia (Distant)	Not specified	36% (WAM2layers) vs. 15% (FLEXPART-WaterSip)
A summary of quantitative data comparing moisture source contributions calculated by WAM2layers and FLEXPART-WaterSip for two river basins on the Tibetan Plateau.[1]

Further insight into the performance of Lagrangian models comes from a study that compared various Lagrangian methodologies against the Weather Research and Forecasting (WRF) model with Eulerian Water Vapor Traccers (WRF-WVTs), which is considered a proxy for reality. This study utilized the Mean Absolute Error Skill Score (MAESS) to quantify model performance.

Lagrangian Model Configuration	Mean Absolute Error Skill Score (MAESS)
WaterSip	0.74
Dirmeyer and Brubaker (1999)	0.77
WaterSip (with physics-based adjustments)	0.84
Dirmeyer and Brubaker (1999) (with physics-based adjustments)	0.87
Mean Absolute Error Skill Score (MAESS) for different Lagrangian moisture tracking configurations when compared to WRF-WVTs. A higher score indicates better performance.[4]

Experimental Protocols

The data presented above is grounded in robust experimental methodologies, which are detailed below to ensure transparency and aid in the critical evaluation of the findings.

Tibetan Plateau Moisture Source Comparison

• Models Used:

  • Water Accounting Model - 2 layers (WAM2layers) - an Eulerian model.

  • FLEXible PARTicle dispersion model (FLEXPART) with the "WaterSip" moisture source diagnostic method - a Lagrangian model.

• Study Area:

  • Yarlung Zangbo River basin (monsoon-dominated).

  • Upper Tarim River basin (westerly-dominated).

• Analysis Period: July 2022.[1]

• Input Data: Forcing data for both models was not explicitly specified in the provided search results.

• Key Analysis: The study quantified and compared the relative contributions of eight selected source regions to the precipitation in the two target basins as simulated by each model.[1]

Lagrangian Model Evaluation with WRF-WVTs

• Models Used:

  • FLEXPART-WRF - a Lagrangian particle dispersion model.

  • Two Lagrangian diagnostic tools for moisture source estimation: WaterSip and the Dirmeyer and Brubaker (1999) methodology.

  • Weather Research and Forecasting (WRF) model with Eulerian Water Vapor Tracers (WRF-WVTs) - used as a proxy for reality.

• WRF Model Configuration:

  • Version: 4.3.3.[4]

  • Spatial Resolution: 20 km.[4]

  • Vertical Levels: 38.[4]

• Input Data: Initial and boundary conditions were sourced from the ERA5 reanalysis.[4]

• Study Design: The comparison was conducted for five atmospheric river events.[4]

• Performance Metric: The Mean Absolute Error Skill Score (MAESS) was calculated for contributions from 20 selected source regions to evaluate the performance of the Lagrangian setups against the WRF-WVTs benchmark.[4]

Visualizing the Methodologies and Assumptions

To further elucidate the differences between these modeling approaches and the core assumptions of WAM2layers, the following diagrams are provided.

Figure 1: Eulerian vs. Lagrangian moisture tracking workflows.

Figure 2: Key assumptions of the WAM2layers two-layer model.

Limitations of the WAM2layers Two-Layer Assumption

The primary limitation of the WAM2layers model stems from its fundamental simplification of the atmosphere into two layers. While this approach was a necessary advancement over single-layer models to account for wind shear, it introduces several constraints:

• Inability to Resolve Complex Vertical Profiles: The two-layer structure cannot capture the detailed vertical distribution of moisture and wind. In reality, the atmosphere exhibits complex vertical gradients and layers of moisture transport that are averaged out in the WAM2layers framework. This can be particularly problematic in regions with complex topography or strong convective activity where vertical transport is significant.

• Potential for Misattribution of Moisture Sources: By simplifying vertical transport, WAM2layers may misattribute the origin of precipitation. For instance, moisture transported at higher altitudes may be incorrectly assigned to lower-level flows, leading to inaccuracies in identifying distant versus local moisture sources. The quantitative data from the Tibetan Plateau study, showing a discrepancy in the attribution of distant versus local sources between WAM2layers and the 3D Lagrangian model FLEXPART, supports this limitation.[1]

• Challenges in Convective Environments: In regions dominated by deep convection, moisture can be rapidly transported from the boundary layer to the upper troposphere. The two-layer model struggles to represent these complex vertical exchanges accurately, potentially underestimating the role of convective processes in moisture transport.

• Dependence on the Definition of the Layer Boundary: The boundary between the upper and lower layers in WAM2layers is a critical parameter. The choice of this boundary can influence the model's output, and an arbitrary or poorly defined boundary can introduce biases into the results.

Conclusion

The WAM2layers model is a valuable and computationally efficient tool for atmospheric moisture tracking, particularly for large-scale studies where a simplified representation of the vertical atmosphere is acceptable. However, its two-layer assumption presents significant limitations, especially for research questions that require a detailed understanding of vertical moisture transport, such as studies in mountainous regions or those focused on the role of convection.

In contrast, 3D Lagrangian models like FLEXPART offer a more detailed representation of atmospheric transport pathways, capturing the complexities of vertical motion. While computationally more intensive, they can provide more accurate results in regions with complex terrain and atmospheric dynamics.

The choice between WAM2layers and a 3D alternative ultimately depends on the specific research question, the geographical area of interest, and the computational resources available. For researchers requiring a high degree of accuracy in identifying moisture source-receptor relationships, particularly in vertically complex atmospheric environments, a 3D model like FLEXPART is recommended. WAM2layers remains a suitable option for broader-scale analyses where computational efficiency is a priority and the limitations of its two-layer assumption are acknowledged and considered in the interpretation of the results. The ongoing community-wide intercomparison studies will undoubtedly provide further guidance on the optimal use of different atmospheric moisture tracking models.

References

• 1. doaj.org [doaj.org]
• 2. research.tudelft.nl [research.tudelft.nl]
• 3. gmd.copernicus.org [gmd.copernicus.org]
• 4. Theory — WAM2layers documentation [wam2layers.readthedocs.io]

A Comparative Guide to Atmospheric Moisture Tracking Models for Researchers

An objective analysis of Eulerian and Lagrangian models, supported by experimental data, to guide researchers in selecting the appropriate tool for their specific needs.

Atmospheric moisture tracking is a critical tool for researchers across various disciplines, including atmospheric science, hydrology, and climate change studies. By tracing the movement of water vapor through the atmosphere, scientists can better understand precipitation patterns, the hydrological cycle, and the impact of climate change on water resources. This guide provides a detailed comparison of the two primary methodologies for atmospheric moisture tracking: Eulerian and Lagrangian models, with a focus on the widely used WAM-2layers and FLEXPART-WaterSip models, respectively.

At a Glance: Eulerian vs. Lagrangian Models

Atmospheric moisture tracking models are broadly categorized into two families: Eulerian and Lagrangian. The fundamental difference lies in their frame of reference.

• Eulerian models , such as the Water Accounting Model - 2 layers (WAM-2layers), employ a fixed grid system. They quantify the movement of moisture by calculating the fluxes of water vapor across the boundaries of each grid cell. This approach provides a comprehensive view of moisture transport over large areas.

• Lagrangian models , like the FLEXible PARTicle dispersion model (FLEXPART) coupled with WaterSip, take a different approach by tracking the trajectories of individual air parcels as they move through the atmosphere. This method is adept at identifying the specific pathways of moisture from source to sink.

Feature	Eulerian Models (e.g., WAM-2layers)	Lagrangian Models (e.g., FLEXPART-WaterSip)
Frame of Reference	Fixed grid	Moving air parcels
Primary Output	Moisture fluxes across grid boundaries	Trajectories of individual air parcels
Strengths	- Better at estimating moisture contributions from distant sources.[1] - Aligns well with actual evaporation data, especially over large and varied terrains.[2] - Computationally efficient for large-scale, long-term studies.	- More reliable in regions with smaller-scale convective systems and complex terrain.[1][2] - Accurately depicts the three-dimensional movement of air particles.[1] - Provides detailed source-receptor relationships.
Weaknesses	- May be less accurate in representing moisture transport in regions with strong vertical wind shear if not using a multi-layer model.[2] - Can be less precise in identifying specific moisture pathways.	- Prone to errors when inferring moisture uptake and loss from specific humidity information in particles.[1] - Can be computationally intensive, especially for high-resolution, long-duration simulations.

Quantitative Performance Comparison

Direct validation of atmospheric moisture tracking models is challenging due to the lack of comprehensive observational data.[3][4] However, intercomparison studies provide valuable insights into their relative performance. One common approach is to use a high-resolution Eulerian model with water vapor tracers (like WRF-WVT) as a proxy for reality to evaluate Lagrangian models.

A key performance metric is the Mean Absolute Error Skill Score (MAESS) , which quantifies the accuracy of a model's prediction of moisture source contributions compared to a reference. In a study comparing different Lagrangian diagnostic tools against WRF-WVT for precipitation events associated with atmospheric rivers, the following results were obtained:

Lagrangian Diagnostic Tool	Mean Absolute Error Skill Score (MAESS)	Improved MAESS with physics-based adjustments
Sodemann et al. (2008) methodology	0.74[3]	0.84[4]
Dirmeyer and Brubaker (1999) diagnostic tool	0.77[3]	0.87[4]

These results indicate that while standard Lagrangian methods provide reasonable accuracy, simple physics-based adjustments can significantly improve their performance.[3][4]

Experimental Protocols: A Generalized Workflow

A typical experimental workflow for comparing atmospheric moisture tracking models involves several key steps, from data acquisition to model evaluation. This process is crucial for ensuring a fair and objective comparison.

Figure 1. A generalized workflow for the intercomparison of atmospheric moisture tracking models.

Detailed Methodologies:

• Define Study Area and Period: The geographical region and the timeframe of the analysis are first established based on the research question.

• Acquire Meteorological Data: High-quality meteorological data is essential for driving the models. Commonly used datasets include the ERA5 reanalysis from the European Centre for Medium-Range Weather Forecasts (ECMWF) and the Global Forecast System (GFS) from the National Centers for Environmental Prediction (NCEP).[5]

• Select and Configure Models: The specific Eulerian and Lagrangian models to be compared are chosen. Configuration involves setting parameters such as grid resolution, number of vertical levels, and the duration of back-trajectories for Lagrangian models.

• Run Eulerian Model: The chosen Eulerian model is run using the prepared meteorological data to simulate moisture fluxes.

• Run Lagrangian Model: The Lagrangian model is executed to generate back-trajectories of air parcels contributing to precipitation in the study area.

• Post-process Model Outputs: The raw outputs from both models are processed to derive comparable quantities, such as the contribution of different source regions to the total precipitation.

• Calculate Performance Metrics: Quantitative metrics are calculated to assess the agreement between the models. This can include direct comparison of moisture source attributions or evaluation against a reference dataset or model.

• Qualitative and Quantitative Comparison: The results are compared both qualitatively (e.g., spatial patterns of moisture sources) and quantitatively (using the calculated performance metrics).

• Sensitivity Analysis: The sensitivity of the model results to key parameters (e.g., trajectory duration, vertical transport schemes) is assessed to understand the robustness of the findings.

Signaling Pathways in Moisture Tracking

The process of atmospheric moisture transport can be conceptualized as a signaling pathway, from the source of evaporation to the sink of precipitation. The tracking models essentially decode this pathway.

Figure 2. A simplified signaling pathway of atmospheric moisture transport.

Conclusion

Both Eulerian and Lagrangian models are powerful tools for atmospheric moisture tracking, each with its own set of strengths and weaknesses. The choice of model should be guided by the specific research question, the characteristics of the study region, and the available computational resources. For large-scale studies focusing on the overall budget and contributions from distant sources, Eulerian models like WAM-2layers are often suitable. For detailed studies of moisture pathways in complex terrain or for specific weather events, Lagrangian models such as FLEXPART-WaterSip may be more appropriate. As demonstrated, ongoing research continues to improve the accuracy of these models, and intercomparison studies are vital for understanding their performance and limitations.

References

• 1. researchgate.net [researchgate.net]
• 2. research.tudelft.nl [research.tudelft.nl]
• 3. GMD - Atmospheric moisture tracking with WAM2layers v3 [gmd.copernicus.org]
• 4. research.tudelft.nl [research.tudelft.nl]
• 5. journals.ametsoc.org [journals.ametsoc.org]

A Comparative Analysis of WAM2layers in Atmospheric Moisture Tracking

For researchers, scientists, and drug development professionals, the accurate tracking of atmospheric moisture is crucial for understanding a wide range of phenomena, from predicting extreme weather events to assessing their impact on ecosystems and infrastructure. The Water Accounting Model - 2 layers (WAM2layers) is a prominent Eulerian model used for this purpose. This guide provides a comparative analysis of WAM2layers, presenting data from key validation studies and comparing its performance against alternative modeling approaches.

This analysis draws upon findings from published case studies that have rigorously evaluated WAM2layers against other atmospheric moisture tracking models, offering insights into its relative strengths and weaknesses.

Performance Comparison of Atmospheric Moisture Tracking Models

Atmospheric moisture tracking models can be broadly categorized into two main types: Eulerian models, which use a fixed grid to track the movement of moisture, and Lagrangian models, which follow the trajectories of individual air parcels. WAM2layers is an Eulerian model, and its performance has been compared with Lagrangian models such as FLEXPART and with more complex online models like Regional Climate Models with tagging capabilities (RCM-tag).

A key study by Van der Ent et al. (2013) provided a foundational comparison of an earlier version of the Water Accounting Model (WAM) with an RCM-tag and a 3D Lagrangian model (3D-T) in a case study tracking moisture from Lake Volta in West Africa.[1][2] This study was instrumental in the development of the two-layer structure of WAM2layers, as it highlighted that strong wind shear was a major source of error in single-layer models.[1]

More recently, a 2024 study by Li et al. provides a direct comparison of WAM2layers with the Lagrangian model FLEXPART-WaterSip in the context of moisture transport over the monsoon- and westerly-dominated basins of the Tibetan Plateau.[3][4] This study underscores the importance of model choice and configuration in different meteorological settings.

The following table summarizes key quantitative findings from these comparative studies.

Parameter	WAM2layers	FLEXPART-WaterSip	RCM-tag (Reference)	Key Findings from Comparative Studies
Moisture Contribution from Westerly-dominated and Distant Sources (Tibetan Plateau)	Generally higher estimates	Lower estimates	Not directly compared in this study	WAM2layers tends to attribute more moisture to distant sources compared to FLEXPART-WaterSip in this region.[5]
Local Moisture Recycling Contribution (Tibetan Plateau)	Lower estimates	Higher estimates	Not directly compared in this study	FLEXPART-WaterSip suggests a greater contribution from local evaporation and nearby sources.[5]
Alignment with Evaporation Data (Tibetan Plateau)	Closer alignment with actual evaporation, especially in regions with complex land-sea distributions.[3]	Evaporation biases can be present but can be partly corrected through calibration.[3]	Not directly compared in this study	WAM2layers shows a notable advantage in aligning estimated moisture sources with observed evaporation patterns.[3]
Performance in Strong Wind Shear Conditions (Lake Volta)	The improved 2-layer version showed significantly better results compared to the single-layer WAM.[1]	Not directly compared in this study	Considered the most accurate in this study.	The study concluded that a 2-layer structure is crucial for accuracy in regions with strong vertical wind shear.[1]

Experimental Protocols

The validation and comparison of atmospheric moisture tracking models rely on carefully designed experimental setups. The following outlines the general methodologies employed in the cited studies.

Case Study 1: Moisture Tracking from Lake Volta (Van der Ent et al., 2013)

• Objective: To evaluate the accuracy of offline moisture tracking models (WAM and 3D-T) against an online model (RCM-tag).

• Reference Model (Ground Truth): A Regional Climate Model (RCM) with an integrated water vapor tagging module (RCM-tag) was used as the benchmark for accuracy.[1]

• Offline Models Compared:

  • WAM (Water Accounting Model): The precursor to WAM2layers, a single-layer Eulerian model.

  • 3D-T: A multi-layer Lagrangian model.

• Experimental Setup:

  • Evaporation from Lake Volta in West Africa was "tagged" within the RCM-tag simulation.

  • The tagged water was tracked to its point of precipitation.

  • The same case study was then simulated using the offline WAM and 3D-T models, driven by the output from the RCM.

  • The precipitation patterns of the tagged moisture from the offline models were compared to the results from the RCM-tag.

• Key Finding: The study identified that the primary cause of errors in the offline models was the presence of strong wind shear in the region. This led to the recommendation and subsequent development of the two-layer structure in WAM2layers to better account for vertical variations in wind.[1]

Case Study 2: Eulerian vs. Lagrangian Models over the Tibetan Plateau (Li et al., 2024)

• Objective: To unravel the discrepancies between the Eulerian WAM2layers model and the Lagrangian FLEXPART-WaterSip model in a region with complex meteorological dynamics.[3][4]

• Models Compared:

  • WAM2layers: A two-layer Eulerian moisture tracking model.

  • FLEXPART-WaterSip: A Lagrangian particle dispersion model coupled with a moisture source diagnostic method.

• Study Area: The Yarlung Zangbo River basin (monsoon-dominated) and the upper Tarim River basin (westerly-dominated) on the Tibetan Plateau.[4]

• Experimental Setup:

  • Both models were used to track the sources of precipitation for the two selected basins.

  • The forcing data for the models were varied in terms of spatial and temporal resolution to assess the impact on the results.

  • The identified moisture source regions and their relative contributions were compared between the two models.

• Key Findings: The study found that WAM2layers attributed a higher contribution from distant, westerly-dominated sources, while FLEXPART-WaterSip indicated a greater role for local recycling and nearby sources.[5] The discrepancies were found to decrease with higher resolution forcing data for WAM2layers.[3] The study also highlighted WAM2layers' better alignment with actual evaporation data.[3]

Visualizing the Methodologies

To better understand the workflows and logical relationships in these comparative studies, the following diagrams are provided.

Workflow of the Van der Ent et al. (2013) model comparison study.

Logical relationships of the findings in the Li et al. (2024) study.

Future Directions and the Community Intercomparison Project

The field of atmospheric moisture tracking is continually evolving, with a recognized need for standardized comparison and validation methodologies. A significant development in this area is the community-wide moisture tracking model intercomparison project, which held a workshop in Leiden in May 2024. This project involves approximately eight different moisture tracking models, including WAM2layers, and focuses on three selected extreme precipitation events. The preliminary results are anticipated to be presented at the European Meteorological Society (EMS) Annual Meeting in September 2024, with a comprehensive paper planned for publication. This initiative is expected to provide a much broader and more detailed quantitative comparison of WAM2layers against a wider array of alternative models, further refining our understanding of the uncertainties and capabilities of different moisture tracking approaches.

References

• 1. HESS - Should we use a simple or complex model for moisture recycling and atmospheric moisture tracking? [hess.copernicus.org]
• 2. hess.copernicus.org [hess.copernicus.org]
• 3. ACP - Unraveling the discrepancies between Eulerian and Lagrangian moisture tracking models in monsoon- and westerly-dominated basins of the Tibetan Plateau [acp.copernicus.org]
• 4. researchgate.net [researchgate.net]
• 5. researchgate.net [researchgate.net]

Safety & Regulatory Compliance

Safe Handling and Disposal of WAM2: A Comprehensive Guide for Laboratory Professionals

Introduction

This document provides essential safety, operational, and disposal protocols for WAM2, a novel compound for research and development. Adherence to these guidelines is critical to ensure personnel safety, environmental protection, and regulatory compliance within the laboratory. This guide is intended for researchers, scientists, and drug development professionals actively using WAM2 in experimental settings.

Immediate Safety and Handling Protocols

Before working with WAM2, it is imperative to consult the Material Safety Data Sheet (MSDS) for specific handling instructions. Always handle WAM2 in a well-ventilated area, preferably within a chemical fume hood.[1] Personal Protective Equipment (PPE), including a lab coat, safety goggles, and chemical-resistant gloves, must be worn at all times.

Waste Segregation: The First Step to Proper Disposal

Proper segregation of waste at the point of generation is crucial to prevent hazardous chemical reactions and to ensure that each waste stream is handled and disposed of correctly.[2] Waste generated from experiments involving WAM2 should be segregated into the following categories: Liquid Chemical Waste, Solid Chemical Waste, and Biologically Contaminated Waste.

Step-by-Step Disposal Procedures for WAM2 and Associated Waste

Liquid Chemical Waste (WAM2-contaminated)

• Aqueous Waste:

  • Collect all aqueous solutions containing WAM2 in a designated, leak-proof, and clearly labeled hazardous waste container.[3]

  • The container label must include "Hazardous Waste," the full chemical name(s) of the contents (including "WAM2"), and the approximate percentage of each component.[1][4]

  • Store the container in a designated satellite accumulation area, such as a fume hood, with secondary containment.[1]

  • Do not mix with organic solvent waste.

• Organic Solvent Waste:

  • Collect all organic solvents contaminated with WAM2 in a separate, designated hazardous waste container.

  • Label the container with "Hazardous Waste," the names of the solvents, and "WAM2," along with their estimated concentrations.

  • Store in the satellite accumulation area, away from incompatible waste streams.

Solid Chemical Waste (Non-Sharp)

• Collection:

  • Place all non-sharp solid waste contaminated with WAM2 (e.g., gloves, pipette tips, weigh boats, contaminated bench paper) into a designated, durable, leak-proof container lined with a chemically resistant bag.

  • This container must be clearly labeled as "Hazardous Chemical Waste" and include "WAM2-contaminated solids."

• Disposal:

  • When the container is three-quarters full, securely close the bag and the container lid.

  • Arrange for pickup by your institution's Environmental Health and Safety (EHS) department.

Biologically Contaminated Waste

For experiments involving cell cultures or other biological materials treated with WAM2:

• Liquid Biohazardous Waste:

  • Decontaminate liquid waste, such as cell culture media containing WAM2, by adding a freshly prepared 10% bleach solution and allowing a contact time of at least 30 minutes.[5][6]

  • After decontamination, the treated liquid can typically be disposed of down the drain with copious amounts of water, provided it does not contain other hazardous chemicals. Always consult your institutional EHS guidelines.[1]

  • Crucially, never autoclave bleach-containing solutions. [5]

• Solid Biohazardous Waste (Non-Sharp):

  • Collect solid biohazardous waste (e.g., cell culture flasks, plates, contaminated gloves) in a biohazard bag.[7]

  • These bags must be placed in a designated biohazard waste container.

  • When three-quarters full, the bag should be loosely sealed to allow for steam penetration and then autoclaved.[5][8]

  • After autoclaving, the now non-hazardous waste can typically be disposed of in the regular trash, but confirm this with local regulations.

Sharps Waste

• Dispose of all sharps (needles, scalpels, serological pipettes, or broken glass) contaminated with either chemical or biological materials into a designated, puncture-resistant sharps container.[5][9]

• Never overfill sharps containers; seal them when they are three-quarters full and arrange for EHS pickup.[5]

Decontamination of Laboratory Equipment

• All non-disposable equipment that has come into contact with WAM2 must be thoroughly decontaminated.

• Wipe surfaces with a 70% ethanol solution. For biological contamination, use a 10% bleach solution followed by a 70% ethanol wipe-down to remove the corrosive bleach residue.[6][10]

Quantitative Data for WAM2 Disposal

Parameter	Guideline	Notes
Bleach Concentration for Decontamination	10% final concentration (1 part bleach to 9 parts liquid waste)	Must be freshly prepared.
Contact Time for Bleach Decontamination	Minimum 30 minutes	Ensure thorough mixing.
Autoclave Sterilization Parameters	121°C, 15 psi	Minimum 30-minute cycle (may vary by load size).
Waste Container Fill Level	Do not exceed 3/4 full	Prevents spills and allows for safe closure.[1][5]

Experimental Protocol: Cell Viability Assay with WAM2

This protocol details a typical workflow for assessing the effect of WAM2 on cancer cell viability, illustrating the generation of multiple waste streams.

• Cell Seeding:

  • Seed MDA-MB-231 cells in a 96-well plate at a density of 5,000 cells per well in 100 µL of DMEM supplemented with 10% FBS.

  • Incubate for 24 hours at 37°C, 5% CO2.

• Compound Treatment:

  • Prepare a 10 mM stock solution of WAM2 in DMSO.

  • Perform serial dilutions in cell culture media to achieve final concentrations ranging from 1 nM to 100 µM.

  • Remove the old media from the cells and add 100 µL of the WAM2-containing media to the respective wells.

  • Waste Generated: Pipette tips (Solid Chemical Waste), old media (Liquid Biohazardous Waste).

• Incubation:

  • Incubate the plate for 72 hours at 37°C, 5% CO2.

• Viability Assessment (MTS Assay):

  • Add 20 µL of MTS reagent to each well.

  • Incubate for 2 hours at 37°C.

  • Read the absorbance at 490 nm using a plate reader.

  • Waste Generated: Pipette tips (Solid Chemical Waste).

• Final Disposal:

  • After the experiment, the entire 96-well plate is considered biohazardous waste. Add bleach solution to each well to decontaminate. After 30 minutes, discard the plate into the solid biohazardous waste container for autoclaving.

Visualizations

Caption: Logical workflow for the segregation and disposal of waste generated from experiments involving WAM2.

Caption: Hypothetical signaling pathway showing WAM2 as an inhibitor of a Receptor Tyrosine Kinase.

References

• 1. google.com [google.com]
• 2. youtube.com [youtube.com]
• 3. youtube.com [youtube.com]
• 4. m.youtube.com [m.youtube.com]
• 5. m.youtube.com [m.youtube.com]
• 6. Appendix Q: Laboratory and Equipment Decontamination Procedures | Office of Research [bu.edu]
• 7. google.com [google.com]
• 8. medsharps.com [medsharps.com]
• 9. 1910.1030 - Bloodborne pathogens. | Occupational Safety and Health Administration [osha.gov]
• 10. unthsc.edu [unthsc.edu]