Day 1 and Beyond for Multi-satellite Retrievals in GPM (Invited)

Merged multi-satellite estimates of precipitation constitute one of the key goals of the Global Precipitation Measurement (GPM) mission. These allow users access to quasi-global precipitation estimates at relatively fine time/space scales without detailed knowledge of satellites, sensors, or algorithms. The Integrated Multi-satellitE Retrievals for GPM (IMERG) algorithm will provide the at-launch combined-satellite precipitation dataset being produced by the U.S. GPM Science Team. This talk will review IMERG's development as a unified U.S. algorithm that takes advantage of strengths in three current U.S. algorithms, namely the Tropical Rainfall Measuring Mission (TRMM) Multi-satellite Precipitation Analysis (TMPA), CPC Morphing algorithm with Kalman Filtering (KF-CMORPH), and Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks using a Cloud Classification System (PERSIANN-CCS). The goal is to provide a long-term, fine-scale record of global precipitation from the entire constellation of precipitation-relevant satellite sensors, with input from surface precipitation gauges as feasible. The record will begin January 1998 corresponding with the start of the TRMM and extend as GPM records additional data. Although homogeneity is considered desirable, the use of diverse and evolving data sources works against the strict long-term homogeneity that characterizes a Climate Data Record (CDR). We plan to compute multiple runs at different latencies (most likely around 4 hours, 12 hours, and 2 months after observation time) to address the needs of different groups of users. We will describe the current focus of bringing up the Day-1 IMERG in the Precipitation Processing System using TRMM-based calibration until the GPM sensor algorithms finish check-out in 2014, and then transition to GPM-based calibration after that. However, at the same time we are looking ahead to the next challenges for Day 2 and beyond. This talk will briefly outline extending the estimation scheme to higher latitudes, refining error estimates, revising the precipitation gauge wind-loss corrections, addressing orographic enhancement, introducing sub-monthly precipitation gauge analyses, and revisiting calibration schemes when the sensors involved have differing capabilities. We also describe approaches to cloud-growth algorithms (likely based on geosynchronous data) and joint observation-model datasets.