Resumen
The occurance of forest fire indonesia especially in Kalimantan is a potential threat to sustainable development. The purpose of this research is a early warning system in forest fire in Kalimantan, by estimating the hotspot as indicators based on visibility and climate data. This research using F test, T test, Multiple Linear Regression analysis, Principle Component Analysis (PCA) and Principle Component Regression Analysis (PCR) Vvisibility, hotspot and temperature data have releated, meaning the very big effect with forest fire incident. Test result of T test and ANOVA P-Value less than 0.05, there is influence between independent variables in this visibility and climate factor against dependent variables in this is the number of hotspots. Relation of climate variables to 10 days forest fire in Central Kalimantan R2 adjusted is 0.4699 with F calculate larger from F table is 160.0940. Relation of climate variables to dasarian forest fire in central kalimantan as early warning system has R2 adjusted that is 0.4176 with f calculate larger from table F of 129.3551. Conclusion forest fires following monsoon character and being affected by el nino events, visibility has a closer and can be used as a indicator of forest fire and land intensity, hotspot in a relationship has a close connection with visibility and climate condition at the same decade period, used equations for early warning system for predicted fire genesis indicates with hotspot amount, compiled from climate condition 10 days.