Satellite-Based Precipitation Measurement Using PERSIANN System
PERSIANN (Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks) is a satellite-based rainfall estimation algorithm. It uses local cloud textures from longwave infrared images of the geostationary environmental satelli
- PDF / 1,878,961 Bytes
- 22 Pages / 441 x 666 pts Page_size
- 77 Downloads / 160 Views
Abstract PERSIANN (Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks) is a satellite-based rainfall estimation algorithm. It uses local cloud textures from longwave infrared images of the geostationary environmental satellites to estimate surface rainfall rates based on an artificial neural network algorithm. Model parameters are frequently updated from rainfall estimates provided by low-orbital passive microwave rainfall estimates. The PERSIANN algorithm has been evolving since 2000, and has generated near real-time rainfall estimates continuously for global water and energy studies. This paper presents the development of the PERSIANN algorithm in the past 10 years. In addition, the validation and merging PERSIANN rainfall with ground-based rainfall measurements for hydrologic applications are also discussed. Keywords PERSIANN · Artificial neural network · Precipitation · Precipitation data merging
1 Introduction Realistic precipitation estimation is crucial to the global climate and land surface hydrologic studies. However, while rain gauges and radar can provide relatively continuous measurements with high temporal frequency, the gauges are sparsely located and provide only point-scale measurements and the radar coverage is limited by topography. The limitation of in-situ precipitation observation over the remote regions makes global climatic and hydrological studies rely mainly on satellite observations and the numerical weather prediction modeling analysis. Since the 1970s, the satellite information has been used to analyze precipitation. Since then, a large number of rain retrieval algorithms have been developed. As visible (VIS) and infrared (IR) images provide excellent temporal resolutions of
K.-L. Hsu Center of Hydrometeorology and Remote Sensing (CHRS), Department of Civil and Environmental Engineering, University of California, Irvine
S. Sorooshian et al. (eds.), Hydrological Modelling and the Water Cycle, 27–48 C Springer Science+Business Media B.V. 2009
27
28
K.-L. Hsu, S. Sorooshian
cloud albedo and top temperatures less than an hour from geosynchronous orbit (GEO) satellites; they are frequently used to monitor cloud motion. These image channels, however, do not provide direct information to infer the actual rainfall at the ground surface. The indirect relationship gives rise to the retrieval of surface rainfall with high uncertainty. Passive Microwave (PMW) sensors carried by satellites in low earth orbits (LEO) sense rainfall clouds more directly. The rain retrieval algorithms based on PMW sensors provide better instantaneous rainfall estimate, however the hind side is that, each LEO satellite only provides limited (1–2) samples in a day for a specified study area. In 1997, the launch of the Tropical Rainfall Measurement Mission (TRMM) provided the first satellite to measure precipitation with an orbital radar sensor to calibrate the other passive microwave sensors (Kummerow et al., 1998; Kummerow et al., 2000; Simpson et al., 1988). Because of it
Data Loading...
