Downloading GPM IMERG Rainfall with Google Earth Engine

This standalone Google Earth Engine (GEE) script automates the extraction of high-resolution satellite rainfall data from NASA's GPM (Global Precipitation Measurement) IMERG dataset for any catchment worldwide.

The workflow generates individual GeoTIFF rasters for each time step, ready for hydrological modeling with HydroPol2D or other distributed models.

What the workflow does

The script downloads precipitation data from GPM IMERG V07 (Integrated Multi-satellitE Retrievals for GPM), NASA's state-of-the-art global precipitation product that combines observations from multiple satellite sensors.

Key Features

Native 30-minute resolution: Captures sub-hourly rainfall variability crucial for flash flood modeling
Global coverage: 60°N-60°S (covers most populated regions worldwide)
0.1° spatial resolution: Approximately 11 km at the equator
Long record: June 2000 to near real-time (updated with approximately 3-month lag)
Calibrated data: Uses gauge-adjusted precipitation estimates
Multiple temporal resolutions: Aggregate to hourly, 3-hourly, 6-hourly, or daily
Flexible catchment input: Works with HydroBASINS or custom shapefiles

Output Specifications

Raster Properties

Property	Value
Resolution	0.1° × 0.1° (approximately 11 km at equator)
Units	mm/h (precipitation rate)
Format	GeoTIFF
Coordinate System	EPSG:4326 (WGS84) by default
NoData Value	-9999
Export Region	Bounding box of catchment

Filename Convention

The script generates descriptive filenames encoding temporal information:

30-minute data:

GPM_30min_2023_01_14_00_00.tif
GPM_30min_2023_01_14_00_30.tif
GPM_30min_2023_01_14_01_00.tif

Hourly aggregation:

GPM_60min_2023_01_14_00_00.tif
GPM_60min_2023_01_14_01_00.tif
GPM_60min_2023_01_14_02_00.tif

3-hourly aggregation (same temporal resolution as MSWEP):

GPM_180min_2023_01_14_00_00.tif
GPM_180min_2023_01_14_03_00.tif
GPM_180min_2023_01_14_06_00.tif

Daily aggregation:

GPM_1440min_2023_01_14_00_00.tif
GPM_1440min_2023_01_15_00_00.tif
GPM_1440min_2023_01_16_00_00.tif

How to Use the Script

Step 1: Define Your Study Area

Option A - Using HydroBASINS (recommended for most users)

Modify these parameters in the script:

Set useHydroBASINS to true
Choose HydroBASINS level (1-12, where higher equals smaller catchments)
Enter your catchment ID(s) in the catchmentIds list

Level recommendations:

Level 5-6: Large river basins (greater than 10,000 km²)
Level 7: Medium catchments (1,000-5,000 km²)
Level 8-9: Small catchments (100-1,000 km²)

Finding your HYBAS_ID:

Visit the HydroSHEDS website or use GEE's Inspector tool
Load HydroBASINS at your desired level in Google Earth Engine
Click on your catchment with the Inspector tool
Copy the HYBAS_ID value

Option B - Using Custom Shapefile

If you have your own catchment shapefile:

Upload shapefile to GEE Assets (Asset → New → Shape files)
Set useHydroBASINS to false in the script
Uncomment and modify the customCatchment line with your asset path

Step 2: Set Date Range

Specify the rainfall event or study period you want to download.

Format: YYYY-MM-DD

Important notes:

GPM IMERG V07 available from June 1, 2000 to near real-time
There is typically a 3-month lag for the final product
For short events (days to weeks): use 30-minute resolution
For long periods (months to years): consider aggregated resolutions

Example configurations:

Flash flood event: 2-3 days at 30-minute resolution
Monthly simulation: 30 days at hourly or 3-hourly resolution
Seasonal analysis: 90 days at 3-hourly or daily resolution
Annual water balance: 365 days at daily resolution

Step 3: Choose Temporal Resolution

Select the time step (in minutes) for your rainfall outputs.

Resolution	Minutes	Files per Day	Typical Use Case
30-min	30	48	Flash floods, urban drainage, short storm events
Hourly	60	24	Event-based modeling, detailed storm analysis
3-hourly	180	8	Multi-day events, medium-term simulations
6-hourly	360	4	Weekly to monthly simulations
12-hourly	720	2	Long-term climate analysis
Daily	1440	1	Water balance, seasonal/annual analysis

File count examples:

2-day event at 30-min: 96 files (approximately 200 MB)
1-month at hourly: 720 files (approximately 1.5 GB)
1-month at 3-hourly: 240 files (approximately 500 MB)
1-year at daily: 365 files (approximately 750 MB)

Aggregation method: For temporal resolutions greater than 30 minutes, the script calculates the mean precipitation rate over the aggregation period, preserving mm/h units.

Step 4: Configure Output Settings

Output folder: Specify the Google Drive folder name where files will be saved. The script will create this folder if it doesn't exist.

Coordinate system: Keep as EPSG:4326 (WGS84) for GPM native projection, or specify a local projection (e.g., UTM zones, State Plane) for your region.

Common projections:

EPSG:4326 - WGS84 (global, recommended for GPM)
EPSG:3310 - California Albers
EPSG:32610 - UTM Zone 10N (California)
EPSG:32643 - UTM Zone 43N (India)

Step 5: Run the Script

Copy the complete script from the code section below
Open Google Earth Engine Code Editor
Paste the script into a new script file
Modify the user inputs (Section 1)
Run the script (click Run button or press Ctrl+Enter)
Wait for processing - The console will show progress
Go to Tasks tab (right panel)
Click RUN on each export task
Wait for exports to complete - Files appear in Google Drive

Processing time:

Task submission: 10-30 seconds
Export per file: 1-5 minutes (depends on catchment size)
Total time: varies with number of files

Understanding the Outputs

Bounding Box vs Exact Catchment

The script exports rainfall data for the rectangular bounding box around your catchment, not the exact catchment shape.

Why?

Faster processing
Maintains regular GPM grid structure
Provides context for adjacent areas
HydroPol2D and most hydrological models can clip to exact boundaries during processing

If you need exact catchment clipping: The exported catchment shapefile can be used to clip rasters in GIS software or during model preprocessing.

Files Included in Export

Rainfall rasters:

One GeoTIFF per time step
Filename includes date and time
Units: mm/h (precipitation rate)

Catchment boundary:

Shapefile (.shp + supporting files)
Same coordinate system as rainfall rasters
Use for clipping or visualization

Quality Assurance

The script includes visualization layers on the map:

Red polygon: Merged catchment boundary
Blue rectangle: Export bounding box
Sample rainfall layer: First time step visualization
Total rainfall layer: Cumulative precipitation over entire period

Console outputs:

Catchment area in square kilometers
Number of 30-minute images available
Date range confirmation
Export task submission status
Total rainfall statistics

Data Considerations

GPM IMERG Strengths

High temporal resolution: 30-minute observations
Global coverage: Consistent data worldwide
Long record: 20+ years of data
Gauge calibration: Improved accuracy in regions with rain gauge networks
Multi-sensor fusion: Combines passive microwave, infrared, and radar data

GPM IMERG Limitations

Spatial resolution: 11 km may be coarse for small catchments (less than 100 square kilometers)
Latency: approximately 3 month lag for final calibrated product (V07)
Uncertainty: Lower accuracy over oceans, mountains, and snow/ice
Bias: Can underestimate or overestimate in certain regions (especially complex terrain)
Coverage gaps: 60°N-60°S only (no polar regions)

Validation Recommendations

Always validate GPM against local observations when available:

Compare with rain gauge data
Check against radar estimates
Verify total accumulations
Assess spatial patterns

For critical applications, consider:

Bias correction using local gauge data
Ensemble approaches with multiple datasets
Uncertainty quantification

Applications

This workflow is designed for:

Hydrological Modeling

HydroPol2D distributed rainfall-runoff simulation
Flash flood forecasting
Urban drainage analysis
Water balance studies

Event Analysis

Storm characterization
Rainfall intensity-duration-frequency analysis
Extreme event reconstruction
Historical flood event analysis

Climate Studies

Long-term precipitation trends
Seasonal rainfall patterns
Drought monitoring
Climate model validation

Water Resources

Reservoir inflow estimation
Irrigation planning
Hydropower assessment
Water availability studies

Troubleshooting

Common Issues

"Memory capacity exceeded" error:

Reduce date range (try smaller time periods)
Increase temporal resolution (e.g., use daily instead of 30-min)
Reduce catchment size (use smaller HydroBASINS level)

No data in exported rasters:

Check date range is within GPM period (2000-06-01 onwards)
Verify catchment is within GPM coverage (60°N-60°S)
Confirm catchment ID is correct

Exports not appearing in Tasks:

Wait 10-30 seconds after running script
Check console for error messages
Refresh Tasks tab

Very large file sizes:

Consider using aggregated temporal resolution
Export smaller date ranges
Use appropriate spatial resolution for your application

Additional Resources

Support

For issues with:

The script: Contact your support team
GPM data: Visit GES DISC or GPM help desk
Google Earth Engine: Check GEE Developers forum
HydroPol2D: Visit model documentation website

Version History

v1.0 (2026): Initial release with GPM IMERG V07 support
- 30-minute to daily temporal resolutions
- HydroBASINS and custom shapefile support
- Automatic task generation
- Quality assurance visualizations

GEE Code

//// ============================================================================
//// GPM IMERG V07 RAINFALL DOWNLOADER - STANDALONE
//// Author........: Marcus Nobrega
//// Date..........: 2026
//// Purpose.......: Download GPM IMERG V07 rainfall data for any catchment
//// Resolution....: 0.1° × 0.1° (native, ~11 km at equator)
//// Units.........: mm/h (precipitation rate)
//// Data Range....: June 2000 - Present (near real-time)
//// ============================================================================


//// ============================================================================
//// SECTION 1: USER INPUTS - MODIFY THESE PARAMETERS
//// ============================================================================

// =============================================================================
// OPTION 1: Use HydroBASINS Catchment
// =============================================================================
// HydroBASINS level: 1-12 (higher = smaller catchments)
// Level 7 = medium-sized catchments (~1000-5000 km²)
// Level 8 = smaller catchments (~200-1000 km²)
var useHydroBASINS = true;  // Set to true to use HydroBASINS
var hybasLevel = 7;         // HydroBASINS level (1-12)

// List of HYBAS_ID values to download rainfall for
// You can find HYBAS_IDs by:
// 1. Going to: https://www.hydrosheds.org/hydrobasins
// 2. Or using the Inspector tool in GEE to click on catchments
var catchmentIds = [
  7071457620  // Example: San Francisco Bay Area catchment
];

// =============================================================================
// OPTION 2: Use Your Own Shapefile (uploaded to GEE Assets)
// =============================================================================
// Uncomment the line below and comment out the HydroBASINS section if using custom shapefile
// var useHydroBASINS = false;
// var customCatchment = ee.FeatureCollection('users/yourusername/your_catchment_shapefile');

// =============================================================================
// Date Range for Rainfall Download
// =============================================================================
// Format: 'YYYY-MM-DD'
// GPM IMERG V07 available from: 2000-06-01 to near real-time
var gpmStartDate = '2023-01-01';  // Start date
var gpmEndDate   = '2023-01-31';  // End date

// =============================================================================
// Temporal Resolution (in minutes)
// =============================================================================
// Choose ONE of the following options:
//   30    = Native 30-minute resolution (best for short periods: days/weeks)
//   60    = Hourly (good for weeks/months)
//   180   = 3-hourly (same as MSWEP, good for months)
//   360   = 6-hourly (good for long periods)
//   720   = 12-hourly
//   1440  = Daily (good for years)
var gpmTemporalResolution = 360;  // Minutes

// =============================================================================
// Output Settings
// =============================================================================
// Google Drive folder name where files will be saved
var outputFolder = 'GPM_SanFrancisco';

// Coordinate Reference System (CRS)
// Recommended: Keep as 'EPSG:4326' (WGS84 lat/lon) for GPM data
// For local projections, use appropriate EPSG code (e.g., 'EPSG:32610' for UTM Zone 10N California)
var outputCRS = 'EPSG:4326';

// No data value for exported rasters
var noDataValue = -9999;

// =============================================================================
// Visualization Settings
// =============================================================================
var mapZoom = 9;  // Initial map zoom level


//// ============================================================================
//// SECTION 2: LOAD CATCHMENT GEOMETRY
//// ============================================================================

var catchmentUnion;
var geometry;

if (useHydroBASINS) {
  // Build HydroBASINS asset path
  var hybasPath = 'WWF/HydroSHEDS/v1/Basins/hybas_' + hybasLevel;
  
  // Load HydroBASINS and filter by HYBAS_ID
  var hybas = ee.FeatureCollection(hybasPath);
  var catchments = hybas.filter(ee.Filter.inList('HYBAS_ID', catchmentIds));
  
  // Dissolve into single geometry
  catchmentUnion = catchments.union(1);
  geometry = catchmentUnion.geometry();
  
  print('Using HydroBASINS Level:', hybasLevel);
  print('Selected HYBAS_IDs:', catchmentIds);
  print('Number of catchments:', catchments.size());
  
} else {
  // Use custom shapefile
  catchmentUnion = customCatchment;
  geometry = catchmentUnion.geometry();
  
  print('Using custom catchment shapefile');
}

// Create bounding box for export (not clipped - for HydroPol2D compatibility)
var bounds = geometry.bounds();

// Display on map
Map.centerObject(geometry, mapZoom);
Map.addLayer(geometry, {color: 'red'}, 'Catchment');
Map.addLayer(bounds, {color: 'blue'}, 'Export Bounds');

// Print catchment info
print('============================================');
print('CATCHMENT INFORMATION');
print('============================================');
print('Catchment area (km²):', geometry.area(1).divide(1e6));
print('Export bounds:', bounds.coordinates());


//// ============================================================================
//// SECTION 3: LOAD GPM IMERG V07 RAINFALL DATA
//// ============================================================================

print('============================================');
print('GPM IMERG V07 CONFIGURATION');
print('============================================');
print('Date range:', gpmStartDate, 'to', gpmEndDate);
print('Temporal resolution:', gpmTemporalResolution, 'minutes');

// Load GPM IMERG V07 collection
// The 'precipitation' band contains calibrated precipitation rate in mm/h
var gpmCollection = ee.ImageCollection('NASA/GPM_L3/IMERG_V07')
  .select('precipitation')
  .filterDate(gpmStartDate, gpmEndDate)
  .filterBounds(bounds);

print('Total 30-min images available:', gpmCollection.size());

// Native GPM resolution: 0.1 degree ≈ 11132 meters at equator
var gpmNativeScale = 11132;


//// ============================================================================
//// SECTION 4: EXPORT RAINFALL RASTERS
//// ============================================================================

// -----------------------------
// OPTION A: Native 30-minute resolution
// -----------------------------
if (gpmTemporalResolution === 30) {
  
  print('============================================');
  print('EXPORTING 30-MINUTE RAINFALL');
  print('============================================');
  print('Units: mm/h (precipitation rate)');
  print('Resolution: 0.1° (~11 km)');
  print('Filename format: GPM_30min_yyyy_mm_dd_hh_mm.tif');
  print('Output folder:', outputFolder);
  
  // Get collection as list
  var gpmList = gpmCollection.toList(gpmCollection.size());
  var numImages = gpmCollection.size();
  
  // Create export tasks for each image
  numImages.evaluate(function(n) {
    print('Creating', n, 'export tasks...');
    
    for (var i = 0; i < n; i++) {
      (function(index) {
        var img = ee.Image(gpmList.get(index));
        var date = ee.Date(img.get('system:time_start'));
        
        // Format date as yyyy_mm_dd_hh_mm
        date.format('yyyy_MM_dd_HH_mm').evaluate(function(dateStr) {
          var exportName = 'GPM_30min_' + dateStr;
          
          Export.image.toDrive({
            image: img.unmask(noDataValue).float(),
            description: exportName,
            folder: outputFolder,
            fileNamePrefix: exportName,
            region: bounds,
            scale: gpmNativeScale,
            crs: outputCRS,
            formatOptions: {noData: noDataValue},
            maxPixels: 1e13
          });
        });
      })(i);
    }
    
    print('✓ All export tasks submitted!');
    print('✓ Go to Tasks tab (right panel) to run exports');
    print('✓ Each file contains precipitation rate in mm/h');
  });
  
}

// -----------------------------
// OPTION B: Aggregated temporal resolution
// -----------------------------
else {
  
  print('============================================');
  print('EXPORTING', gpmTemporalResolution, 'MINUTE RAINFALL');
  print('============================================');
  print('Units: mm/h (mean precipitation rate)');
  print('Resolution: 0.1° (~11 km)');
  
  var resStr = gpmTemporalResolution.toString();
  print('Filename format: GPM_' + resStr + 'min_yyyy_mm_dd_hh_mm.tif');
  print('Output folder:', outputFolder);
  
  // Calculate aggregation parameters
  var intervalsPerPeriod = gpmTemporalResolution / 30;
  print('30-min intervals per output:', intervalsPerPeriod);
  
  // Time period calculations
  var startMillis = ee.Date(gpmStartDate).millis();
  var endMillis = ee.Date(gpmEndDate).millis();
  var millisPerPeriod = gpmTemporalResolution * 60 * 1000;
  var numPeriods = endMillis.subtract(startMillis).divide(millisPerPeriod).floor();
  
  var periods = ee.List.sequence(0, numPeriods.subtract(1));
  
  // Create aggregated collection
  var aggregatedCollection = ee.ImageCollection(periods.map(function(p) {
    var periodStart = ee.Date(startMillis.add(ee.Number(p).multiply(millisPerPeriod)));
    var periodEnd = periodStart.advance(gpmTemporalResolution, 'minute');
    
    var periodImages = gpmCollection.filterDate(periodStart, periodEnd);
    
    // Mean precipitation rate over period (still in mm/h)
    var aggregated = periodImages
      .mean()
      .rename('precipitation')
      .set('system:time_start', periodStart.millis())
      .set('period_start', periodStart.format('yyyy-MM-dd HH:mm'))
      .set('period_end', periodEnd.format('yyyy-MM-dd HH:mm'))
      .set('temporal_resolution_min', gpmTemporalResolution)
      .set('units', 'mm/h');
    
    return aggregated;
  }));
  
  print('Number of aggregated images:', aggregatedCollection.size());
  
  // Get collection as list
  var aggList = aggregatedCollection.toList(aggregatedCollection.size());
  var numAggImages = aggregatedCollection.size();
  
  // Create export tasks for each aggregated image
  numAggImages.evaluate(function(n) {
    print('Creating', n, 'export tasks...');
    
    for (var i = 0; i < n; i++) {
      (function(index) {
        var img = ee.Image(aggList.get(index));
        var date = ee.Date(img.get('system:time_start'));
        
        // Format date as yyyy_mm_dd_hh_mm
        date.format('yyyy_MM_dd_HH_mm').evaluate(function(dateStr) {
          var exportName = 'GPM_' + resStr + 'min_' + dateStr;
          
          Export.image.toDrive({
            image: img.unmask(noDataValue).float(),
            description: exportName,
            folder: outputFolder,
            fileNamePrefix: exportName,
            region: bounds,
            scale: gpmNativeScale,
            crs: outputCRS,
            formatOptions: {noData: noDataValue},
            maxPixels: 1e13
          });
        });
      })(i);
    }
    
    print('✓ All export tasks submitted!');
    print('✓ Go to Tasks tab (right panel) to run exports');
    print('✓ Each file contains mean precipitation rate in mm/h');
  });
}


//// ============================================================================
//// SECTION 5: VISUALIZATION
//// ============================================================================

// Visualize first rainfall image
var sampleGPM = ee.Image(gpmCollection.first());

Map.addLayer(sampleGPM, {
  min: 0,
  max: 10,
  palette: ['white', 'lightblue', 'blue', 'darkblue', 'purple', 'red']
}, 'Sample Rainfall (mm/h)');

// Calculate and visualize total rainfall over period
var totalRainfall = gpmCollection.sum().multiply(0.5);  // mm (30-min * mm/h = mm/0.5hr)

Map.addLayer(totalRainfall.clip(geometry), {
  min: 0,
  max: 200,
  palette: ['white', 'lightblue', 'blue', 'darkblue', 'purple', 'red']
}, 'Total Rainfall (mm)');

// Print statistics
print('============================================');
print('RAINFALL STATISTICS');
print('============================================');
print('Sample image date:', ee.Date(sampleGPM.get('system:time_start')));

var stats = totalRainfall.reduceRegion({
  reducer: ee.Reducer.minMax().combine(ee.Reducer.mean(), '', true),
  geometry: geometry,
  scale: gpmNativeScale,
  maxPixels: 1e13,
  bestEffort: true
});

print('Total rainfall over period (mm):', stats);


//// ============================================================================
//// SECTION 6: EXPORT CATCHMENT BOUNDARY
//// ============================================================================

// Export catchment shapefile for reference
Export.table.toDrive({
  collection: catchmentUnion,
  description: 'Catchment_Boundary',
  folder: outputFolder,
  fileNamePrefix: 'Catchment_Boundary',
  fileFormat: 'SHP'
});

print('============================================');
print('EXPORT SUMMARY');
print('============================================');
print('✓ Rainfall rasters configured for export');
print('✓ Catchment boundary shapefile configured');
print('✓ Go to Tasks tab to start downloads');
print('============================================');

What the workflow does​

Key Features​

Output Specifications​

Raster Properties​

Filename Convention​

How to Use the Script​

Step 1: Define Your Study Area​

Step 2: Set Date Range​

Step 3: Choose Temporal Resolution​

Step 4: Configure Output Settings​

Step 5: Run the Script​

Understanding the Outputs​

Bounding Box vs Exact Catchment​

Files Included in Export​

Quality Assurance​

Data Considerations​

GPM IMERG Strengths​

GPM IMERG Limitations​

Validation Recommendations​

Applications​

Hydrological Modeling​

Event Analysis​

Climate Studies​

Water Resources​

Troubleshooting​

Common Issues​

Additional Resources​

Support​

Version History​

GEE Code​