How to Use Open Data to Predict Forest Fires

Image by Sippakorn Yamkasikorn from pexels
Image by Sippakorn Yamkasikorn from pexels


In 2023, the world witnessed devastating impacts due to wildfires, with record numbers of forests lost across various continents. These fires were driven by a combination of factors, including climate change, which has led to drier and hotter conditions conducive to wildfires. But what if we could use tree data to predict where intervention by humans could prevent mass fires from happening in the first place? It turns out that maybe we can. 

This article by Peeler et al. discusses identifying areas in western US conifer forests where proactive forest management can significantly reduce wildfire-caused carbon loss. It highlights the need for tried and tested techniques, such as forest thinning and prescribed burning, especially in California, New Mexico, and Arizona, which have high vulnerability. Most crucially, it posits that we can use data models to predict where such techniques will be most effective. The authors highlight the importance of collaborative and equitable management strategies to mitigate climate and wildfire risks while maintaining carbon sequestration and supporting community resilience against such disasters. 

The Importance of Intervention

High-severity wildfires in frequent-fire forests disrupt carbon recovery by destroying mature trees, which impedes the regeneration of trees and thus the forest's ability to sequester carbon again. Such wildfires can create a cycle that leads to more frequent and intense fires, further reducing carbon stocks. Climate change exacerbates this issue by creating hotter, drier conditions that hinder tree regeneration even further. Proactive human intervention is necessary to mitigate these effects and prevent ongoing carbon loss.

Studying Current Developments

The study developed a geospatial database that showed indicators for both exposure and sensitivity to wildfire in western US conifer forests, integrating factors like annual burn probability, total carbon, potential carbon loss, and carbon recovery post-wildfire. These indicators were normalized and weighted to identify areas most at risk from wildfires, aiming to inform targeted forest management practices for reducing wildfire-caused carbon loss.

Strategies for Mitigation Using Data

The article advocates for the use of open-source environmental data to enhance forest management accuracy and effectiveness. By refining risk assessments and management approaches with precise data, targeted interventions can be more effectively applied to specific forest conditions. The ultimate goal is to develop online tools that could inform and improve forest management practices to prevent wildfires, relying on contributions from local communities to enrich the data.

Technical Insights

Peeler et al. utilized bivariate maps (maps of two or more variables) and categorized data into quantiles for carbon exposure, sensitivity, and vulnerability to wildfires. It provides a clear framework for identifying areas at different levels of wildfire risk by evaluating the potential for carbon loss and the capacity for carbon recovery. This method allows for a nuanced assessment of regions where carbon stocks are most at risk from wildfires.

Example bivariate map of exposure to fire and carbon loss
Example bivariate map of exposure to fire and carbon loss

Data Accessibility and Tools

Peeler, Jamie et al. (2023). Geospatial data from: Identifying opportunity hot spots for reducing the risk of wildfire-caused carbon loss in western US conifer forests [Dataset]. Dryad. 

Data can be opened using R or ArcGIS Pro.

Subscribe to Ground Truth

Don’t miss out on the latest issues. Sign up now to get access to the library of members-only issues.