Characteristics of Amazonian rain measured during TRMM-LBA

The Tropical Rainfall Measuring Mission (TRMM) is a NASA satellite project initiated to address a gap in our ability to accurately observe detailed rainfall patterns over the tropical continents and oceans. To support TRMM, several field campaigns were conducted. The TRMM-LBA (Large-scale Biosphere Atmosphere) experiment was conducted over the southwestern region of the Amazon (state of Rondonia, Brazil) in order to provide detailed information on the precipitation characteristics in the interior of a tropical continent. Information from TRMM-LBA will be used for validation of TRMM satellite products and for initialization/validation of cloud-resolving models and passive microwave retrieval algorithms. TRMM-LBA was conducted in parallel with the WETAMC-LBA campaign aimed at examining the effect of land use change on rainfall in Amazonia.

During the TRMM-LBA field campaign, a variety of instrumentation was deployed during the wet season (January - February 1999) to measure rainfall including several rain gauge networks, disdrometers, the NASA ER-2 aircraft (with radar and passive microwave instruments), a dual-frequency precipitation profiler, and ground-based C-band Doppler (NASA TOGA) and S-band polarimetric (NCAR S-POL) radars. The focus of this study will be on the estimation, validation, and scientific application of rain rate estimates derived from the NCAR S-POL radar.

The S-POL data were carefully corrected for the presence of clear-air echo, ground clutter, anomalous propagation, second-trip echoes, partial beam blocking, precipitation attenuation, and calibration biases by applying polarimetric methods. Using an optimal polarimetric radar technique, maps of rain rate have been calculated from observations of S-POL horizontal reflectivity (Z_h), differential reflectivity (Z_dr), and specific differential phase (K_dp) every ten minutes from 10 January to 28 February 1999. From these rain rate estimates, daily and 30-day rain accumulation maps have been compiled. When validated against the rain gauge totals, preliminary S-POL estimates of monthly rainfall have a bias error in the range of –5% to –11% and a standard error of 14% to 20%.

Using these detailed and accurate estimates of Amazonian rainfall, we have begun to explore several core TRMM and LBA scientific priorities. For example, we have decomposed rainfall statistics into the two predominant regional meteorological regimes that are characterized by low-level easterlies and westerlies. Significant differences in rainfall characteristics exist for the two regimes. Due to a higher frequency of intense rain rate events, convection in the easterly regime typically has a higher mean rate than in the westerly regime. We have also investigated the diurnal cycle of rainfall over Amazonia. As expected, the rainfall is highly diurnally modulated with a rapid increase in area mean rain rate centered on local noon. The overall maximum mean area rain rate occurs around 15L with a secondary maximum in the early morning hours (01L) and local minima at 09L and 22L. We plan to partition rainfall statistics by storm size, morphology, and duration, in addition to meteorological regime. Finally, we plan to investigate the connection between land surface properties in Amazonia (deforested/pasture land vs. forest canopy) to rainfall initiation, amounts, and characteristics.