Open AccessRainfall estimation using multiple linear regression based statistical downscaling for Piperiya watershed in Chhattisgarh
Author(s) -
Surendra Kumar Chandniha,
Mitthan Lal Kansal
Publication year - 2016
Publication title -
journal of agrometeorology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.295
H-Index - 11
eISSN - 2583-2980
pISSN - 0972-1665
DOI - 10.54386/jam.v18i1.911
Subject(s) - downscaling , hadcm3 , watershed , environmental science , linear regression , climatology , climate change , regression analysis , meteorology , statistics , mathematics , geography , general circulation model , gcm transcription factors , precipitation , computer science , ecology , geology , machine learning , biology
Climatic variability and its behavior is a complex phenomenon that is directly associated with uncertainties. In the climate change study, particularly in hydrological aspects, it is necessary to identify the parameters (predictors) that are directly or indirectly associated with predictands. The forecasted results are directly associated with the selection of predictors. In the present study, the statistical downscaling model (SDSM) has been advocated to downscale the daily rainfall in Piperiya watershed of Chhattisgarh state. SDSM is based on multiple linear regression (MLR) technique. The daily rainfall data (1961-2001) of the Piperiya watershed in Chhattisgarh is considered as input (predictand) to the model. The model has been calibrated and validated on the basis of rainfall period of 1961-1990 and 1991-2001 respectively with large scale predictors of National Centre for Environmental Prediction (NCEP) reanalysis data. Finally, monthly rainfall is predicted on the basis of forecasted future daily rainfall for the periods of 2020s, 2050s and 2080s under the consideration of HadCM3 A2 and B2 emission scenarios.