Edwin Yoshimar Chávez Pascual, Salvador Lozano Trejo, Martha Patricia Jerez Salas, Gustavo Omar Díaz Zorrilla, Ernesto Castañeda Hidalgo
DOI: 10.59427/rcli/2025/v25.001-008
Precipitation is a key component in the water management of agricultural regions with dry seasons, such as the periurban communities of Oaxaca, where rainfall and temperature regimes show high variability. The objective was to adjust a SARIMA model to forecast the precipitation pattern within the RH20 Basin. Satellite data from NASA POWER (1984-2022) were processed in Rstudio using the Box and Jenkins approach. The data were transformed to logarithms to achieve stationarity. The optimal model was SARIMA (1,1,1) (1,0,2)12, with an AIC value of (1681.8), AICc (1682.0), and BIC (1706.6), along with a non-significant Ljung-Box test (p>0.9798), indicating a good fit. The model allows for estimating monthly precipitation averages for the period 2023-2026, which will enable better water planning invulnerable rural contexts. SARIMA is suitable for periodic time series and can be a supporting tool in climate adaptation strategies.
Pag 001-008
