ENSO Precipitation Variations as Seen by GPM and TRMM Radar and Passive Microwave Observations

Tropical precipitation variations related to ENSO are the largest-scale such variations both spatially and in magnitude and are also the main driver of surface temperature-surface rainfall relationships on the inter-annual scale. GPM (and TRMM before it) provide a unique capability to examine these relations with both the passive and active microwave approaches. Documenting the phase and magnitudes of these relationships are important to understand these large-scale processes and to validate climate models. However, as past research by the authors have shown, the results of these relations have been different for passive vs. radar retrievals. In this study we re-examine these relations with the new GPM Version 5 products, focusing on the 2015-2016 El Nino event. The recent El Nino peaked in Dec. 2015 through Feb. 2016 with the usual patterns of precipitation anomalies across the Tropics as evident in both the GPM GMI and the Near Surface (NS) DPR (single frequency) retrievals. Integrating both the rainfall anomalies and the SST anomalies over the entire tropical ocean area (25N-25S) and comparing how they vary as a function of time on a monthly scale during the GPM era (2014-2017), the radar-based results show contrasting results to those from the GMI-based (and GPCP) results. The passive microwave data (GMI and GPCP) indicates a slope of 17%/C for the precipitation variations, while the radar NS indicates about half that ( 8%/C). This NS slope is somewhat less than calculated before with GPM's V4 data, but is larger than obtained with TRMM PR data ( 0%/C) for an earlier period during the TRMM era. Very similar results as to the DPR NS calculations are also obtained for rainfall at 2km and 4km altitude and for the Combined (DPR + GMI) product. However, at 6km altitude, although the reflectivity and rainfall magnitudes are much less than at lower altitudes, the slope of the rainfall/SST relation is 17%/C, the same as calculated with the passive microwave data. The reasons for these differences are explored and lead to conclusions that the radar-based estimates of surface rainfall with GPM have limitations (and are negatively biased) in relatively intense rainfall and this leads to an underestimation of large-scale rainfall under El Nino conditions, where more oceanic rainfall, and more intense rainfall are prevalent.