32nd Conference on Broadcast Meteorology/31st Conference on Radar Meteorology/Fifth Conference on Coastal Atmospheric and Oceanic Prediction and Processes

Friday, 8 August 2003: 9:00 AM
Comparison of TRMM Satellite-Based Rainfall with Surface Radar and Gauge Information
T. Keenan, BMRC, Melbourne, Vic., Australia; and E. Ebert, V. Chandrasekar, V. Bringi, and M. Whimpey
Poster PDF (552.2 kB)
A multi-faceted approach to validation Tropical Rainfall Measuring Mission (TRMM) products is examined employing data from tropical Australia. Consideration is given to validation on the scale of direct satellite measurement and for end products used to imply national and regional scale rainfall distributions.

The TRMM precipitation radar (PR) products (2A25) and near surface rainfall from the TRMM Microwave Imager (TMI) (2A12) and combined PR and TMI (2B31) TRMM data (direct satellite measurements in the form of surface rainfall and reflectivity measurements) are compared with measurements from the C-Band Polarimetric (CPOL) radar in the tropical monsoon environment of Darwin, Australia (12S 131 E). Two seasons of monsoon rainfall (pre-boast 1999/2000 and post boast 2001/2002) are employed enabling polarimetric and traditional power-based rainfall products from CPOL coincident with TRMM satellite over-passes to be compared. In this case, the polarimetric-based self-consistency method of attenuation correction, as developed by Bringi et al (2002), is applied to the CPOL data and the method of Bolen and Chandrasekar (2000) is used to account for the different view angles, beam widths and reflectivity values of the two platforms.

NASA produces gridded level 3 TRMM products on a variety of temporal and spatial scales including the 3B31 monthly 0.50 resolution products; the daily 3B42 10 merged infrared estimates calibrated against TRMM data; the 3B43 TRMM 10 combination with GPCC gauge observations and other experimental real-time products at e.g. 3B40RT. 3B41RT and 3B42RT. Validation of these products is also necessary to define error characteristics. In this case the Australian national rainfall analysis, based on a 0.250 grid at 24 h temporal resolution is compared with these level 3 products. This rainfall analysis is based on a network of 5000 Australian gauges. Time series of regional rainfall, for the Australian tropics and subtropics from these two sources will be compared. Errors will be discussed as a function of rain rate and the ability of the satellite estimates to get regional patterns correct evaluated in terms of the contiguous rain area technique of Ebert and McBride (2000).

Supplementary URL: