Abstract

The aim of this study is to predict climate data in order to develop agricultural plans for crops in Western Thailand. This study analyzed the time series characteristics of mean monthly temperature data in Western Thailand collected from Information Services and Climate Statistics, and the Meteorological Departments over consecutive days in Tak, Kanchanaburi, Prachuap Khiri Khan, Phetchaburi, and Ratchaburi provinces. From the trend analysis, the linear trend of mean monthly temperature in Tak, Prachuap Khiri Khan, Phetchaburi, and Ratchaburi showed a slight increasing trend, and a slight decreasing trend in Kanchanaburi. According to the time series analysis result, (Seasonal Autoregressive Integrated Moving Average) the SARIMA (p, 0, q) (P, 1, Q)12 model was a proper model to predict the data set of mean monthly temperature in Western Thailand. The forecast values in next five years (2019–2023) by this model showed that the lowest mean temperature in Tak, Kanchanaburi, Phetchaburi, and Ratchaburi would be in December (25.27–26.64 C) and January (26.73–26.78 C) for Prachuap Khiri Khan. The highest mean temperature in Tak, Kanchanaburi, and Ratchaburi would be in April (31.79–32.05 C) and May (30.37–30.67 C) for Prachuap Khiri Khan and Phetchaburi. Furthermore, the study found that mean temperature is expected to increase in the future. The forecasting values can be applied in crop cultivation planning in accordance to the low and high temperature periods and for adaptation to increasing temperatures.