Abstract:
The forest area in Indonesia is decreasing, caused by forest fires either intentionally
through land clearing or unintentionally due to climate change. Hotspots are an indicator
used in detecting fires on a land. Efforts can be made to overcome the fire problem by
forecasting the number of hotspots in Indonesia using intervention analysis. The
intervention method is used to examine data that has increased in an extreme manner.
There are two functions in the intervention method, namely step and pulse functions. This
study uses pulse function intervention on hotspot occurrence data in Indonesia because
the intervention is temporary and only occurs at a certain time. The results of this study
showed that the intervention SARIMA model with order 𝑏 = 0, 𝑠 = 0, 𝑟 = 1 is a good
model in forecasting the occurrence of hotspots in Indonesia with a MAPE value of
8.06%.
