Detection and forecast of the rainy seasons, application over Analamanga

Changes like decrease of the rainfall amount or late rainy seasons are observed over Madagascar nowadays due to the climate changes.To tackle the negative impacts of climate changes, this paper aims to detect rainy seasons period (onset dates and offset dates) in order to forecast rainy season and set calendar.Climate variables, from 1979 to 2021, such as the daily precipitations, the temperatures, the Dipole Index Mode and the Oceanic Nio Indexes) were used.The detection of the rainy seasons was done by the mean of the accumulation of the anomalies of the daily mean of the precipitations.Modelling and forecasting stages were made by using Fuzzy Logic with triangular membership functions.The skill scores were the Root Mean Squared Error (RMSE), the Mean Absolute Error (MAE), the RMSE and MAE scaled by the standard deviation and the average of the onset and demise dates of the rainy seasons.Findings show that the investigated region was subdivided into two regions: A, where the climatological wet spell starts between 25 th October and 12 th November, and B, where the climatological wet spell starts between November 16 and 20.The best forecasting models were observed with values of RMSE (MAE) between 0.52 and 0.80 (0.04 and 0.36).Moreover, scaled by the standard deviation (average), these skill scores are less than 1 (have values between 0.05% and 7.30%).