When fire detections occur, how do rainfall, soil moisture, wind, vegetation condition, peat extent, and ENSO context differ from matched non-detections?
Interactive research report · Indonesia and global context
Indonesia Wildfire Evidence Report
Evidence-bounded researchMethods, uncertainty, and claim limits remain visible
Why this research exists, and what it is designed to ask
This opening page contains the research purpose, scope, and reading paths only. Statistical results, maps, and source audits live on their own pages so each kind of evidence can be read in the right context.
A public-interest investigation into evidence—not a dashboard of conclusions
Wildfire narratives often mix climate, vulnerable peat, land-use change, and human intent. This report keeps those questions separate.
The purpose is to test what the available evidence can support while keeping uncertainty, missing data, and alternative explanations visible. Begin with the question that matters to you, then move to the evidence page designed for it.
Research agenda
The questions this report is built to examine
Are fire-positive forest cells more often followed by mapped forest loss or land-cover transition than comparable fire-negative cells?
Do spatial patterns observed in Indonesia resemble patterns in other countries, and where do differences in source coverage prevent direct comparison?
Which public claims can the completed evidence address, and which still require actor-, ownership-, policy-, or intent-specific data?
Choose a reading path
One report, organized by purpose
Each page groups related evidence. Anchors are used only within a page, so browser navigation always matches the direction and destination shown.
01Read the findings
Start with the fitted statistical evidence, uncertainty intervals, robustness checks, and the difference between association and causation.
Open page →02Explore maps and comparisonsInspect Indonesia province context, the global comparison, and Kalimantan reporting layers without treating aggregate polygons as burned footprints.
Open page →03Audit methods and sourcesReview the data inventory, validation gates, missingness rules, provenance ledger, and the claims the design cannot establish.
Open page →