HR: 15:10h
AN: H53J-07 INVITED     [Abstracts]
TI: Statistical Issues in Validating Rainfall Measurements
AU: * Bell, T L
EM: Thomas.L.Bell@nasa.gov
AF: NASA/Goddard Space Flight Center, Mail Code 613.2, Greenbelt, MD 20771 United States
AU: Kundu, P K
EM: kundu@climate.gsfc.nasa.gov
AF: Joint Center for Earth Systems Technology, Univ. of Maryland Baltimore County, NASA/Goddard SFC Mail Code 613.2, Greenbelt, MD 20771 United States
AB: No method of measuring rainfall over large areas is completely satisfactory because of the extreme variability in rain rates both in space and in time and the lack of practical remote sensing devices with a simple linear response to rain rate. Rain gauges measure rain in the most direct fashion possible, but individual gauges cover only small areas. Radar and passive microwave methods cover large areas, but their estimates of rainfall depend on a host of factors causing errors that are generally understood but not always predictable. These problems are made even more difficult in complex terrains. Determining areal rainfall with rain-gauge arrays requires assessing error levels in the rain-gauge measurements themselves and providing statistical descriptions of the errors introduced in the areal estimates by the imperfect spatial coverage of the gauge array. Gauge data are often used to assess error levels in radar and satellite remote sensing estimates, but in this case the antennas receive information from much larger areas than the gauges cover, and this mismatch must be taken into account in comparing gauge data to remote sensing estimates for validation purposes. Statistical studies of data from well-controlled field programs and statistical models can help with estimating errors in areal averages from rain-gauge arrays and in validating remote sensing measurements. Results and insights from some of these studies will be discussed.
DE: 1853 Precipitation-radar
DE: 1854 Precipitation (3354)
DE: 1855 Remote sensing (1640)
DE: 1873 Uncertainty assessment (3275)
SC: Hydrology [H]
MN: Fall Meeting 2005