Modelling weekly ranifall using gamma probablity distribution and Markov chain for crop planing in sum humid (dry) climate of central Bihar

Abstract

A study was conducted during 2000-03 to analyse standard meteorological weekly rainfall data of Patna for 42 years (1960-200 I) using gamma probability distribution, which has been identified as the best-fit model out ofseven competing distributions using Anderson-Darling goodness-of-fit test. The best-fit model has been employed for obtaining the assuredquantun1 ofrainfallpertaining to standard meteorological weeks (23-40) at different probability levels. The probability of occurrence ofdry weeks has been found low (0.13 to 0.38) with the corresponding high probability (0,65 to 0.88) ofwet weeks using the first order Markov chain model during the standard meteorological weeks 26-37. The distribution and amount of rainfall indicated that the transplanting of kharif rice has to be completed during the standard meteorological week 26 (25 June- 1July) for maximum utilization of rainwater of 330 mmand 465 rom during the standard meteorological weeks 26-39 at 70% and 60% probability levels respectively. The additional water required can be met out from groundwater resources for enhancing the productivity of rice-wheat cropping system in the region.