Statistical forecasting of soil dryness index in the southwest of Western Australia

The soil dryness index (SDI) is a numerical value reflecting the dryness of soils, deep forest litter, logs and living vegetation. It is used to plan the timing of spring and autumn prescribed burning in the southwest of Western Australia, and to plan the level of preparedness for wildfire emergencies. This paper examines three different methods for forecasting the annual cycle of the weekly SDI series. We analyze the use of the previous 5 years’ weekly average (P5YWA), seasonal decomposition (SD) and Box–Jenkins multiplicative seasonal ARIMA (SARIMA) models. Results show that the more complex SARIMA model performs better than the simpler P5YWA and SD models in forecasting the annual cycle of SDI data. As the SARIMA model provides good forecasting information regarding the behavior of the annual cycle of SDI series, we suggest building an operational SARIMA forecasting program to inform bushfire control operations such as spring burning.