Skip to content

Acquire IMERG data

About the data

WGS84 - EPSG:4326

  • Bounding box 94,-12,142,7

  • GeoJSON 

    {"type":"FeatureCollection","features":[{"type":"Feature","properties":{},"geometry":{"type":"Polygon","crs":"EPSG:4326","coordinates":[[[94,-12],[94,7],[142,7],[142,-12],[94,-12]]]}}]}

BBox

Historical

Timestep Release Period Unit Link
Daily Final Run 1 Jun 2000 - 30 Jun 2020 mm/day https://gpm1.gesdisc.eosdis.nasa.gov/data/GPM_L3/GPM_3IMERGDF.06/
Daily Late Run 1 Jul 2000 - now mm/day https://gpm1.gesdisc.eosdis.nasa.gov/data/GPM_L3/GPM_3IMERGDL.06/

Near-real time

Timestep Release Period Unit Link
30-minutes Final Run 1 Jan 2020 - 30 Jun 2020 mm/hour https://gpm1.gesdisc.eosdis.nasa.gov/data/GPM_L3/GPM_3IMERGHH.06/
30-minutes Early Run 1 Jul 2020 - now mm/hour https://gpm1.gesdisc.eosdis.nasa.gov/data/GPM_L3/GPM_3IMERGHHE.06/

Download and subset for Indonesia via OpenDAP

Using wget, see https://disc.gsfc.nasa.gov/data-access#mac_linux_wget 

wget --http-user=[USERNAME] --http-password=[PASSWORD] --load-cookies ~/.urs_cookies --save-cookies ~/.urs_cookies --keep-session-cookies --no-check-certificate --auth-no-challenge=on -r --reject "index.html*" -np -e robots=off <insert complete HTTPS OPENDAP URL>

username: gis.wfp.indonesia

password: Send email to benny.istanto@wfp.org to get the password

OPENDAP URL: for 1 Jan 2019, check the data catalog from url on the left column. https://gpm1.gesdisc.eosdis.nasa.gov/opendap/hyrax/GPM_L3/GPM_3IMERGHH.06/2019/001/3B-HHR.MS.MRG.3IMERG.20190101-S003000-E005959.0030.V06B.HDF5.nc4?precipitationCal[0:0][2740:3219][779:969],time,lon[2740:3219],lat[779:969]

GPM_3IMERGHH.06 - Code of product. Notes: NASA frequently update the version of data, current version is 06. If the download failed, change into new version 07 or 08 and so on.

2019 - year in YYYY

001 - Date in Julian Date for 30min data

3B-HHR.MS.MRG.3IMERG.20190101-S003000-E005959.0000.V06B.HDF5.nc4 - filename

  • 3B-HHR.MS.MRG.3IMERG - product code for 30min final run

  • 3B-HHR-E.MS.MRG.3IMERG - product code for 30min early run

  • 20190101-S003000-E005959.0000 - Date and time information, each file has a unique information, check the pattern in data source.

  • Start Date/Time: All files in GPM will be named using the start date/time of the temporal period of the data contained in the product. The field has two subfields separated by a hyphen.

  • Start Date: YYYYMMDD Start Time: Begin with Capital S and follow with HHMMSS

  • End Time: Begin with Capital E and follow with HHMMSS Hours are presented in a 24-hour time format, with 00 indicating midnight. All times in GPM will be in Coordinated Universal Time (UTC). The half-hour sequence starts at 0000, and increments by 30 for each half hour of the day.

  • V06B - version of data. Frequently change

  • HDF5 - Original format of the data

  • nc4 - Data format after subsetting process using OPENDAP

precipitationCal - Selected variable to download

[0:0][2740:3219][779:969],time,lon[2740:3219],lat[779:969] - time and bounding-box from original to subset for Indonesia.

Conversion NC to GeoTIFF, and clip using polygon

All downloaded subset IMERG are in netCDF format. We need to convert it to GeoTIFF, the easiest way is using GDAL: 

gdal_translate -of GTiff -a_srs EPSG:4326 NETCDF:"filename.nc4":precipitationCal fileoutput.tif

But sometimes GDAL couldn’t handle rotated poled grid from subset netCDF, we could try rioxarray based on solution from https://gis.stackexchange.com/a/367651

Then we need to crop the GeoTIFF using Indonesia boundary. GDAL can handle this task: 

gdalwarp --config GDALWARP_IGNORE_BAD_CUTLINE YES -srcnodata -999 -dstnodata NoData -cutline idn_bnd_subset_clip_imerg_grid_a.shp -crop_to_cutline input.tif clip_output.tif

Prepare input data (rainfall accumulation) for model

Historical

Calculate accumulation of 1 to 5 days rainfall, daily rolling windows for both Final and Late run.

Example: start from 1 Jun 2000

  • 1-day: 1 Jun, 2 Jun, 3, Jun, ...

  • 2-days: 1+2 Jun, 2+3 Jun, 3+4 Jun, ...

  • 3-days: 1+2+3 Jun, 2+3+4 Jun, 3+4+5 Jun, ...

  • 4-days: 1+2+3+4 Jun, 2+3+4+5 Jun, 3+4+5+6 Jun, ...

  • 5-days: 1+2+3+4+5 Jun, 2+3+4+5+6 Jun, 3+4+5+6+7 Jun, ...

To make it easier, every 1 - 5 days accumulation data will save as a file with the latest file name: Example, data 2+3+4 Jun 2020, will save as: precip.imergd.20000604.3days.tif

Near-real time

Create rainfall accumulation for every 6-hours from 30-min data

Example: 6-hours rainfall accumulation for every date, for both Final and Early run

Date YYYYMMDD Time 0000
20000601 0000 to 0360
20000601 0390 to 0720
20000601 0750 to 1080
20000601 1110 to 1440
20000602 0000 to 0360
20000602 0390 to 0720
20000602 0750 to 1080
20000602 1110 to 1440

To make it easier, every 6-hours accumulation data will save as a file with the latest file name: Example, data 20000601 for period 0000 to 0360, will save as: precip.imerghh.20000601.0360.6h.tif

Calculate 1 to 5 days rainfall, 6-hours rolling windows for both Final and Late run.

  • 1-day

    precip.20000601.0360.6h to precip.20000601.1440.6h
    recip.20000601.0720.6h to precip.20000602.0360.6h
    recip.20000601.1080.6h to precip.20000602.0720.6h
    precip.20000601.1440.6h to precip.20000602.1080.6h
    precip.20000602.0360.6h to precip.20000602.1440.6h
    …

  • 2-days

    precip.20000601.0360.6h to precip.20000602.1440.6h
    precip.20000601.0720.6h to precip.20000603.0360.6h
    precip.20000601.1080.6h to precip.20000603.0720.6h
    precip.20000601.1440.6h to precip.20000603.1080.6h
    precip.20000602.0360.6h to precip.20000603.1440.6h
    …

  • 3-days

    precip.20000601.0360.6h to precip.20000603.1440.6h
    precip.20000601.0720.6h to precip.20000604.0360.6h
    precip.20000601.1080.6h to precip.20000604.0720.6h
    precip.20000601.1440.6h to precip.20000604.1080.6h
    precip.20000602.0360.6h to precip.20000604.1440.6h
    …

  • 4-days

    precip.20000601.0360.6h to precip.20000604.1440.6h
    precip.20000601.0720.6h to precip.20000605.0360.6h
    precip.20000601.1080.6h to precip.20000605.0720.6h
    precip.20000601.1440.6h to precip.20000605.1080.6h
    precip.20000602.0360.6h to precip.20000605.1440.6h
    …

  • 5-days

    precip.20000601.0360.6h to precip.20000605.1440.6h
    precip.20000601.0720.6h to precip.20000606.0360.6h
    precip.20000601.1080.6h to precip.20000606.0720.6h
    precip.20000601.1440.6h to precip.20000606.1080.6h
    precip.20000602.0360.6h to precip.20000606.1440.6h
    …

Following the standard for 6-hour naming convention, we will adjust name according the time information:

  • 0360 will assign with ID: 00
  • 0720 will assign with ID: 06
  • 1080 will assign with ID: 12
  • 1440 will assign with ID: 18

Then rainfall accumulation data (1 to 5 days) will save as a file with latest filename but substitute with ID: precip.imerghh.YYYYMMDDID.H024.tif

Where

  • precip: precipitation
  • YYYY: year
  • MM: month
  • DD: date
  • ID: time information, see above
  • H: Historical data
  • 024: 1-day rainfall
    • 2-days will be 048
    • 3-days will be 072
    • 4-days will be 096
    • 5-days will be 120

Example:
1-day rainfall from precip.20000601.0360.6h to precip.20000601.1440.6h will save as precip.imerghh.2000060118.H024.tif

5-days rainfall from precip.20000601.1080.6h to precip.20000606.0720.6h will save as precip.imerghh.2000060606.H120.tif