Wet and dry spell analysis for crop planning using Markov's chain model

Abstract

The Indian monsoon is characterised by a series of wet and dry spells. In semi-arid regions the number and length of dry spells decide the success or failure of the kharif crops. Hence, study of onset of monsoon and wet- and dry- spells assumes much importance in rainfed cropping systems in the country. In the present study, the Markov chain probability model was used to calculate the chances of occurrences of dry- and wet- spells for Rahuri using 44-year (1975 to 2018) weekly rainfall data. The average annual rainfall is found to be 564.55 mm and the coefficient of variation (CV) as 32.88% which is higher than the threshold limit (25) indicating the uneven distribution of rainfall. Monsoon starts on the 25th SMW (18th – 24th June) and remains active up to 43rd SMW (22–28th October) with a total length of 18 weeks (126 days). Initial, conditional and consecutive dry- and wet-week probabilities showed that chances of occurrence of a week getting wet is high at the beginning and ending of kharif season excluding 25th to 33rd weeks. Chances of occurrence of the wet week of more than 35% at the beginning of kharif season in the 23rd and 24th SMW indicates that summer ploughing and initial seed bed preparations and sowing operations could be taken up since 25th SMW with supplemental irrigations. Lower values of CV in kharif season showed that variability of rainfall during this period is less which supports the growing of rainfed crops. The significant contribution of weekly rainfall (>20 mm) during 34th – 40th SMW with high consecutive wet week probability indicates the potential of harvesting excess runoff water for future supplemental irrigations and also soil erosion measures to be taken up for soil erosion control.