Interannual variability of tropical rainfall as seen from TRMM V6

Considerable uncertainty surrounds the issue of whether precipitation over the tropical oceans (30o N/S) systematically changes with interannual sea-surface temperature (SST) anomalies that accompany El Nino (warm) and La Nina (cold) events. Although it is well documented that El Nino-Southern Oscillation (ENSO) events with marked SST changes over the tropical oceans produce significant regional changes in precipitation, water vapor, and radiative fluxes in the tropics, we still cannot yet adequately quantify the associated net integrated changes to water and heat balance over the entire tropical oceanic or land sectors. Resolving this uncertainty is important since precipitation and latent heat release variations over land and ocean sectors are key components of the tropical heat balance in its most aggregated form.

Rainfall estimates from the Version 5 Tropical Rainfall Measuring Mission (TRMM) Precipitation Radar (PR) averaged over the tropical oceans have not solved this issue and, in fact, show marked differences with estimates from two TRMM Microwave Imager (TMI) passive microwave algorithms. Recent work has shown that path-integrated attenuation derived from the effects of precipitation on the radar return from the ocean surface exhibits interannual variability that agrees closely with the TMI time series, yet the PR rainfall interannual variability (and attenuation derived predominantly from reflectivity) differs even in sign. We will explore these apparent inconsistencies and examine changes in these relationships for the new TRMM Version 6 retrievals. The focus is on uncertainties in microphysical assumptions necessitated by the single-frequency PR measurement and the consequent difficulties posed for detecting climate-related precipitation signals.

Another complication of the TRMM record for climate studies is the effect of a 50 Km orbit boost performed in Aug 2001 to enable a longer mission lifetime. Due to associated 1/ r-squared effects on the reflectivity signal and beamfilling, TRMM reflectivity and rainfall estimates are lowered. We use the existing SSM/I record as well as precipitation estimates from the 12.6 GHz QuikSCAT radar to understand this offset and isolate the interannual precipitation signal in the tropics since the beginning of the TRMM mission.