In brief
Under President Prabowo Subianto’s directive, the National Research and Innovation Agency (BRIN) is validating fire‑prediction models using satellite imagery. The models aim to give authorities months’ notice and enable better prevention and response. BRIN will also collaborate with universities and share daily hotspot data with the National Disaster Mitigation Agency and the Forestry Ministry.
Indonesia’s National Research and Innovation Agency (BRIN) is advancing a new set of fire‑prediction models that aim to forecast the risk of forest and land fires three to six months in advance. The initiative follows an instruction from President Prabowo Subianto delivered during a limited meeting at Jakarta’s presidential palace on 27 October 2026, and the research is already underway, according to BRIN chief Arif Satria.
BRIN’s chief noted that the agency is currently validating the models to forecast fire conditions in the three‑ and six‑month windows. The models will draw on satellite imagery and other remote‑sensing data to assess the likelihood and potential spread of fires across Indonesia’s tropical landscapes.
Methodology and Data Sources
Under the project, BRIN will use satellite‑based observations to monitor surface temperature and detect potential fire hotspots – denser, brighter spots of heat that can indicate incipient fires. A hotspot, in plain language, is a small area of intense heat detected by satellite sensors and can serve as an early indicator of fire risk.
The agency plans to combine these observations with climatological variables such as precipitation, temperature, and vegetation moisture to train predictive algorithms. The models will be calibrated against historical fire records to ensure that they can accurately anticipate where and when fires are likely to ignite.
ARIF SATRIA: "The research is already underway, and we are now validating the model to forecast fire conditions three and six months in advance."
"The research is already underway, and we are now validating the model to forecast fire conditions three and six months in advance," said Arif Satria.
BRIN will also support research into how fires spread in areas affected by forest and land fires, enabling stronger modeling of fire dynamics for future periods.
Collaboration and Knowledge Sharing
To strengthen the modeling effort, BRIN intends to partner with universities across Indonesia, inviting specialists in atmospheric science, ecology, and data science to contribute expertise on fire forecasting and fire‑spread patterns.
In addition, BRIN will continue its daily monitoring of fire hotspots via satellite imagery and will share that data with the National Disaster Mitigation Agency (BNPB) and the Forestry Ministry. The data exchange is intended to help these agencies respond to fires earlier and more effectively.
The agency is mandated to use satellites and satellite imagery for multiple purposes, including hotspot detection, and is working to integrate those observations into a national decision‑support system.
Significance for Prevention and Response
ARIF SATRIA: "Predicting fire risks is crucial to strengthening prevention and response efforts, rather than waiting until fires have already broken out."
"Predicting fire risks is crucial to strengthening prevention and response efforts, rather than waiting until fires have already broken out," said Arif Satria.
By providing longer lead times, the models could give local authorities and emergency services the additional hours, days, or months needed to mobilize resources, set up firebreaks, and issue timely evacuation orders. In practice, a six‑month forecast could identify high‑risk periods in advance of the monsoon season when dry conditions often exacerbate fire potential.
Challenges and Future Directions
The initiative faces scientific challenges such as accounting for the complex interplay of weather anomalies, human activity, and land‑use changes that can drive fire ignition. BRIN will have to refine its algorithms to account for such variables and improve the accuracy of spatial predictions at fine scales.
Once validated, the system will likely evolve into an operational forecast service that integrates with existing early‑warning frameworks. The Spanish and Canadian research communities have implemented similar approaches, and BRIN plans to benchmark its models against those methods to ensure global competitiveness.
Because Indonesia is the world’s most fire‑prone tropical country, successful deployment of a medium‑range forecast model could save lives, protect biodiversity, and reduce carbon emissions from uncontrolled fires.
In the coming months, BRIN will release more detailed technical specifications and engage its university partners in the next phases of model refinement. The ultimate goal is a fully operational system that delivers reliable fire‑risk outlooks six months ahead, fully integrated with national disaster‑management protocols.
