In early 2014, the NASA Precipitation Measurement Missions (PMM) launched the Global Precipitation Measurements (GPM) mission. Since then, great efforts have been directed to evaluate the state-of-the-art precipitation products from the GPM during the first 3 years and now we expect that those will be extensively utilized over the next years for data applications in relevant areas such as hydrological research and climate modeling. Among many GPM products, the Multi-satellitE Retrievals for GPM (IMERG) is most interesting to the users since it delivers the ‘best’ precipitation estimates by combining data obtained from all available microwave and infrared (IR) platforms of the GPM satellite constellation. While waiting for the next version (V04) of the IMERG, I’ve decided to post a short manual on the data especially designed for first-time users not familiar with the data system.

What is GPM IMERG?

The GPM is next-generation of the Tropical Rainfall Measuring Mission (TRMM).  Like the TRMM, the GPM mission aims at providing uniformly calibrated precipitation estimates at a quasi-global scale (60°N-60°S) by merging the measurements from its constellation of microwave and IR satellites. All GPM data sets including measurements obtained from each platform (Level 2) are available on the PMM site. Here, I will only refer to the IMERG (Level 3, merged data) since these 0.1° x 0.1° gridded data are a final product, designed to use in further research and data applications. Although IMERG data are currently available only from 2014, the data will be extended over the TRMM-era (from 1998) during this year.

How can I access IMERG?

Just visit the GPM Data Downloads. It asks for a quick registration using your e-mail; do not forget to tick ‘Near-Realtime (NRT) Products’ if you want to download them as well.

blog_fig02
Mark “NRT” to access IMERG Late/Early data

NRT products refer to data from Late (~18 hours latency) or Early (~6 hours latency) runs of IMERG, those data are generated earlier than IMERG Final run. IMERG produces the data with multiple latencies to address different user needs. IMERG Final run takes longer time to be processed (~3 months) since it corrects biases in satellite estimates, using gauge analysis, to produce high-quality data suitable for research purposes.

Do you want to plot the IMERG data?

Since the data are provided in a HDF format (or NetCDF), Panoply will be the easiest way to plot them. Download the program and just ‘open’ any IMERG file. If you want to work further with the data (as is usually the case), you should write a script so to extract, manipulate, and analyze data as you want. Here it is a sample Python script to open a file and extract ‘precipitationCal’ (gauge-corrected estimates) variables over the whole domain.

import h5py
import numpy as np

#one file for each time step
gpmf = h5py.File(yourIMERGfile.hdf, ‘r’)

#extract data over the whole domain [:,:]
precipitation=gpmf[‘/Grid/precipitationCal’][:,:].astype(np.float32)

The last option is using Giovanni to browse/plot data directly on the site. It is a nice tool not only to have a look at the data for a certain time period and a region but also to check if you extract data and plot them correctly.

Part of my PhD work is to validate IMERG precipitation data using ground reference such as gauge or radar based measurement data. Such ground validation is intended to assess whether satellite estimates meet their accuracy requirements and objectives. The research results will be touched upon in the following blog posts. Stay posted!

Advertisements