Precipitation Extremes in the New TMPA and GPCP 1DD Data Sets

Two data sets routinely computed by the authors are long enough, each in excess of a decade, that it is reasonable to examine the behavior of global extreme precipitation. Furthermore, each has been significantly updated in new versions within the last few months. The TRMM Multi-satellite Precipitation Analysis (TMPA) provides 0.25°x0.25° 3-hourly estimates of precipitation in the latitude band 50°N-50°S for the years 1998-present, while the GEWEX/Global Precipitation Climatology Project (GPCP) One-Degree Daily (1DD) precipitation product provides 1°x1° daily global estimates of precipitation for 1997-present. The TMPA incorporates many (intercalibrated) microwave estimates of precipitation, augmented by microwave-calibrated infrared (IR) estimates, while the 1DD consists of microwave-calibrated IR estimates in the band 40°N-40°S and TOVS (or AIRS) sounding-based estimates at higher latitudes. Both datasets incorporate monthly raingauge analyses, but it should be emphasized that the day-to-day occurrence of precipitation is entirely based on the satellite data. The new TMPA is referred to as Version 7, and the new 1DD is Version 1.2.

The analysis presented here focuses on basic parameters that are stable and well-suited to comparison with station data or model estimates. These include means, frequency of precipitation, and 95th percentile values. The frequency of no precipitation is used as a measure of dryness. Overall, there is fair consistency between the 1DD and TMPA datasets. One result of the comparison confirms that several of the parameters, including frequency of precipitation and 95th percentile values are sensitive to the spatial scale. In addition to enhancing our confidence in the results previously reported, this comparison allows us to examine contrasts in behavior for the same region across the seasonal cycle. For example, the TMPA tends to have drier estimates than the 1DD at higher latitudes, ~40-50°, particularly in the winter hemisphere where the microwave algorithms currently lack sensitivity to the reduced precipitation signals. These results will be compared to the previous Version 6 TMPA and Version 1.1 1DD comparison results. Many of the differences between versions can be explained by the changes in input data for each of the data sets, including upgrades to the same consistent monthly gauge analysis in both data sets and corrections to deficient AMSU precipitation estimates in the TMPA.