Resumen

DAS, Prabir Kumar et al. Spatial analysis of wet spell probability over India (1971-2005) towards agricultural planning. Atmósfera [online]. 2020, vol.33, n.1, pp.19-31.  Epub 04-Dic-2020. ISSN 0187-6236.  https://doi.org/10.20937/atm.52499.

The spatial analysis of the wet spell probability over the Indian region has been carried out using daily gridded (0.5º × 0.5º) rainfall data of 1971-2005 during the summer monsoon period, i.e. June-September. A threshold was applied to the weekly cumulative rainfall to convert the rainfall data into wet spell information. A Markov chain model was employed to estimate the initial and conditional probabilities of the wet spell for each grid and the spatio-temporal distribution of the wet spells probabilities was analyzed. The probability maps were able to capture the summer monsoon scenario over the Indian region, representing the onset, progression and withdrawal of monsoon rainfall. Higher wet spell probability was observed over the west coast and northeastern parts of India, i.e., the initial probability was maximum over these regions. However, lower probability values were observed in West Rajasthan, Gujarat and southern India. A threshold of 80% of the maximum initial probability was used to standardize the spatially-variable probability information, and a week with more than the threshold values was considered as a probable wet week. The duration of the longest probable wet spell was highest along the west coast and in northeastern India, whereas it was lowest in western and southern India. The start and duration of the longest spell of the probable wet week can be used for rainfed-agricultural planning, i.e., the start of sowing/planting, selection of crops and varieties based on their length of growing period, optimum harvesting period to avoid wet spell, etc.

Palabras llave : Markov chain; wet spell probability; probable wet week; Indian summer monsoon; rainfall.