The GPM Radiometer Algorithm - from Validation to Uncertainties

The Global Precipitation Measurement (GPM) mission was launched in February 2014 as a joint mission between JAXA from Japan and NASA from the United States.  GPM carries a state of the art dual-frequency precipitation radar and a multi-channel passive microwave radiometer that acts not only to enhance the radar’s retrieval capability, but also as a reference for a constellation of existing satellites carrying passive microwave sensors.  In April 2022, GPM approved V 7 of its precipitation products starting with GMI and continuing with the constellation of radiometers.  The precipitation products from these sensors are consistent by design and show relatively minor differences in the mean global sense.  Validation results will be shown for work done over the Continental United States using a Radar/Gauge composite as truth, and Kwajalein atoll to represent truth over tropical oceans.  The validation results are a necessary but not sufficient component to quantify the algorithm’s uncertainties.  Good results for bias, MAR and RMSE are demonstrated.  Validation results, however, are only able to assess errors at their own sites, and systematic errors, in particular, are not actually systematic, but regime dependent errors that vary as a function of how well the algorithm assumptions are captured at the validation sites.   This talk will explore ways of validating not by location, but by precipitation states that, the environment that the precipitation evolves in, as a way of obtaining robust statistics of individual precipitation states that are universal and can be applied with confidence to areas outside the validation domain.