Abstract
The study examines three satellite-based data sets to estimate long-term precipitation for the Thailand region: the Tropical Rainfall Mapping Mission (TRMM) 3B43, the Climate Prediction Centre morphing technique (CMORPH), and a locally developed regression model using the geostationary meteorological satellite (GMS) covering the Thailand region. Data for the first two sets were available at a spatial resolution of 0.25° × 0.25°, while the local regression model used data from the GMS at a resolution of 5 km × 5 km. The statistical regression model was developed by relating long-term monthly average precipitation from 27 rain gauge stations with concurrent satellite data in the visible and thermal infrared bands. The model was then tested against independent data from 27 rain gauge stations. Satellite/rain gauge ratios were estimated, and a smooth spline surface was used to correct the error from the model. Data from the three approaches were compared with a rain gauge network. The TRMM relation performed better than CMORPH, and the performance for GMS was comparable to TRMM with root mean square different and mean bias difference of 33.6 and 4.2%, respectively. The locally developed regression model was used to produce monthly and yearly total rainfall maps from the GMS data for the entire country.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call