3B.3 A Quick Summary of IMERG Versions and Features

Monday, 8 January 2018: 2:30 PM
Room 18B (ACC) (Austin, Texas)
George J. Huffman, NASA GSFC, Greenbelt, MD; and D. T. Bolvin, D. Braithwaite, K. Hsu, R. J. Joyce, C. Kidd, E. J. Nelkin, S. Sorooshian, J. Tan, and P. Xie

The rapid turnover of versions in various Global Precipitation Measurement (GPM) mission data products, in particular for the Integrated Multi-satellitE Retrievals for GPM (IMERG) merged precipitation product, presents a challenge to dataset users. Throughout, IMERG, a U.S. GPM science team product, has used intercalibrated estimates from the international constellation of precipitation-relevant satellites and other data (including monthly surface precipitation gauge analyses) to compute half hour, 0.1° x 0.1° gridded datasets over 60°N-S in three “Runs”—Early, Late, and Final (~4 hours, ~14 hours, and ~3 months after observation time, respectively). However, the quality of the input datasets and the processing details in IMERG have shifted.

This talk will summarize the shifts in IMERG from Version 03 to 04 in early Spring 2016, and to Version 05 in late Summer 2017. For example, Version 04 replaced approximate pre-launch calibrations with GPM Core Observatory-based calibrations, while Version 05 introduced improved estimates for the primary GPM instrument products (DPR, GMI, and Combined Instrument). In Version 04 the IR estimates were routinely calibrated to the passive microwave estimates. As analysis showed that the Combined Instrument estimates (the IMERG calibration standard) tend to be biased high over land and low over ocean at higher latitudes, in Version 04 we climatologically calibrated IMERG to the Global Precipitation Climatology Project (GPCP) monthly Satellite-Gauge product, except in low- and mid-latitude ocean regions. This calibration leaves the relative time series intact, and only adjusts the mean of the entire series. In Version 05 the primary GPM instrument products have reduced biases, but calibration to GPCP continues to be necessary to achieve the most realistic estimates. Finally, retrospective processing back into the TRMM era is expected in early 2018, after which the legacy TMPA dataset will be retire.

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner