Stochastic Modelling for Generating Monthly and Annual Rainfall at Rajasthan

Abstract

Rainfall is a phenomenon, which directly or indirectly affects all the sectors like agriculture, insurance, industry and other allied fields. It is well established that the rainfall is changing on both the global and the regional scales due to global warming. So the prediction of actual amount of rainfall is very complex till now. But now a day’s stochastic rainfall models are widely used in hydrologic analysis for rainfall forecasting. These models can provide precipitation input to models whenever data are not available. In this paper, Markov first order process is used to calculate the amount of rainfall for all the districts of Rajasthan. For this purpose rainfall data for 41 years (1970–2010) for each 33 districts of Rajasthan is used. Using first order Markov chain model monthly and yearly magnitude of rainfall is predicted for the year 2011. All the parameters of this model were predicted from the historical rainfall records. Obtained results are compared with the observed rainfall. This model has 18.89% average relative error for all 33 districts of Rajasthan for monthly analysis of annual rainfall where as 24.02% is the observed for annual rainfall. It is observed that monthly total of rainfall for each district is more efficient than yearly predicted rainfall.

Keywords— stochastic, hydrologic analysis, Markov first order,
rainfall

Cite this Article Saurabh Singh, Babita Pandey, Vijay Shankar Dogra. Stochastic Modelling for Generating Monthly and Annual Rainfall at Rajasthan. Journal of Water Resource Engineering and Management. 2017; 4(3): 28–39p.