Measurement Uncertainty of Satellite-based Precipitation Sensors

Uncertainties associated with satellite-based multi-sensor precipitation products are from two sources: the upstream sensors used and the algorithms to merge the sensor retrievals. Several satellite-based precipitation products, generated from disparate merging algorithms, share remarkable similarities in error characteristics, this suggests these errors can be traced back to their upstream sensor inputs. In the post 2013 era of the NASA's Global Precipitation Measurement (GPM) mission, it is critical for both data producers and users to understand the realistic limitations to which these satellite sensor inputs have and propagate to merged products. This paper presents results focusing on quantifying the measurement uncertainty of satellite precipitation sensors. Several Passive Microwave (PMW) precipitation sensors have been studied, including TMI, AMSR-R, SSM/I, MHS, and AMSU-B. A high-resolution ground radar-based dataset, the next generation multi-senor QPE (Q2) data over the contiguous U.S. is used as the ground reference. Additional quality control and quality assurance (QA/QC) have been performed on the reference data. The high spatial and temporal resolution of the reference data, allows rigorous collocation (within 10 minutes) and relatively more precise comparison with the satellite overpasses. From our results, PMW sensor retrievals exhibit fairly systematic bias varying by seasons and rain rates, with overestimates in summer at intermediate rain rates and underestimates in winter at high-end rain rates. This feature is also observed in the merged products, suggesting the dominant contribution of the sensor errors to merged products. In addition, at satellite snapshot scale, the false alarm and missed detection rates become much more significant.