Markov Chain Rainfall Probability Model and Rainy Season Length in Kendrapara District, Odisha for Efficient Crop Management Strategies

Abstract

Systematic information on rainfall patterns and distribution is essential for crop selection, adopting improved agronomic practices, designing rainwater harvesting structures, and conserving rainwater. This study analyzed the probabilities of rainfall occurrence to know the rainfall onset, withdrawal, and rainy season length for different blocks of Kendrapara district, Odisha. The rainfall analysis showed that the kharif season shares 75-78% of annual rainfall with July and August being the wettest. Seasonal rainfall showed a significant increase during kharif season while the decreasing trend was observed during the rabi and summer seasons. Markov chain analysis showed that the weekly probability for 20-mm rainfall and consecutive two weeks is more than 70% during SMW 25-38. During the June-October period, assured rainfall at 80% probability is greater than 800 mm in four blocks, while in other blocks it ranged between 610-779 mm. Assured rainfall during the rabi and summer season is uncertain, however, a 105-126 mm rainfall is received in May in many blocks, useful for summer ploughing and green manure crops. Rainy season length analysis at an 80% probability level revealed that only in 14-16 weeks’ rain occurs, thereby, only short-duration paddy can be grown with lower risk. By utilizing October rainfall and conserved soil moisture, short-duration pulses/oilseeds could be grown in all the blocks. However, kharif paddy must be sown/transplanted at the optimum time