Usage

This guide walks you through how to run the DUSTY system from start to finish, using your own .json configuration file and selecting between classical and fusion-based dust storm detection modes.

Step-by-Step Execution

To run the DUSTY pipeline:

  1. Open a terminal in your project directory.

  2. Run the main script:

    python DustyMain.py

  3. A file dialog will prompt you to select a configuration .json file. Choose the one you created (see Installation for structure).

  4. You will be prompted:

    Would you like to use Fusion Storm Detector Module
1 or 0

    • Enter 1 for FusionStormDetector (uses both DUSMASS + AODANA)

    • Enter 0 for StormDetector (uses DUSMASS only)

  5. The program will begin downloading and processing data. A progress bar will appear for each stage of the pipeline.

Modules Automatically Run

Based on your inputs, the following modules are executed:

  • MERRA2AODProcessor – Downloads and processes dust mass concentration files (DUSMASS)

  • MERRA2AODANAProcessor – Downloads and processes AODANA files if fusion mode is selected

  • MonthlyDustAnalyzer – Computes time-resolved monthly dust statistics

  • StormDetector or FusionStormDetector – Detects and tracks storm blobs day-by-day

Expected Outputs

All output files are saved in the csv_output_path defined in your .json config.

Key output files:

  • <project_name>_storm_lifecycle.csv A daily record of detected storm blobs with storm ID, size, dust mass, and airport info.

  • <project_name>_monthly_dust_summary.csv A statistical summary of dust activity across the region, highlighting anomalies.

  • <project_name>_monthly_AODANA_summary.csv (Fusion mode only) Similar summary based on aerosol optical depth.

  • Multiple intermediate CSVs for each day’s DUSMASS and AODANA data

Troubleshooting

  • Download errors? Make sure your .netrc file is correctly set up and your NASA Earthdata login is valid.

  • Empty output? Ensure the date and region bounds match days and locations with known dust activity.

  • Slow performance? Try reducing the date range or geographic area in your config.

Best Practices

  • Keep each run in a unique project folder to avoid overwriting CSVs

  • Always check that the output files are being written to the correct directory

  • Use the same time format (%d/%m/%Y %H:%M) in your CSVs and code when debugging