Intercalibrating the Disparate Precipitation Estimates Input to IMERG

The Integrated Multi-satellitE Retrievals for GPM (IMERG) IMERG is a satellite-based precipitation algorithm with surface gauge input that merges multiple satellite precipitation retrievals to compute a single best precipitation estimate every half hour at the 0.1° spatial resolution for the current record, April 2014 - present. Though anchored by GPM, IMERG takes advantage of all precipitation-related satellites of opportunity and as a result the input satellite constellation is constantly evolving over time. Despite the use of the same algorithm (GPROF) to compute all the individual input satellite passive microwave precipitation estimates, and the efforts of the XVAL team to intercalibrate background radiances, each satellite still "sees" something different due to differences in channel complement, footprint resolution, scan strategy, and orbit. To minimize the possible inhomogeneities in the IMERG data record, all input satellite estimates must be calibrated to a single, best standard - the GPM combined GMI/DPR product. It is believed the state-of-the-art microwave imager (GMI) combined with the first spaceborne dual-frequency radar (DPR) provides the highest-quality, most consistent precipitation estimates available. Since the combined GMI/DPR estimates are provided on the (narrow) Ku swath, the IMERG intercalibration must take place in two steps: (1) calibrate all other satellite estimates to the GMI using seasonal climatological histograms, and (2) calibrate GMI and estimates in (1) to the combined GMI/DPR using monthly regional histograms.

The talk will focus on the approach developed for obtaining stable intercalibrations of precipitation for the various satellites included in IMERG, and the level of success in replicating the histograms of the calibrators in each step. Analysis and validation of the complete IMERG precipitation estimates will also be provided.