A machine learning model can assess the effectiveness of different management strategies

Wildfires are a growing threat in a world shaped by climate change. Now, researchers from Aalto University have developed a neural network model that can accurately predict the occurrence of fires in peatlands. They used the new model to assess the effect of different fire risk management strategies and identified a series of interventions that would reduce the incidence of fires by 50-76%.

The study focused on the Central Kalimantan province of Borneo in Indonesia, which has the highest density of peatland fires in Southeast Asia. Drainage to support agriculture or residential expansion has made peatlands increasingly vulnerable to recurring fires. In addition to threatening lives and livelihoods, peatland fires release significant amounts of carbon dioxide. However, prevention strategies have encountered difficulties due to the lack of clear and quantified links between proposed interventions and fire risk.

The new model uses measurements taken before each fire season from 2002 to 2019 to predict the distribution of peatland fires. Although the results can be broadly applied to peatlands elsewhere, further analysis should be carried out for other contexts. “Our methodology could be used for other contexts, but this specific model would need to be re-trained on new data,” says Alexander Horton, the postdoctoral researcher who conducted the study.

The researchers used a convolutional neural network to analyze 31 variables, such as land cover type and pre-fire vegetation and drought indices. Once trained, the network predicted the probability of a bog fire at each point on the map, producing a predicted distribution of fires for the year.

Overall, the neural network predictions were correct 80-95% of the time. However, while the model was generally correct in predicting a fire, it also missed many fires that actually occurred. About half of the observed fires were not predicted by the model, meaning it is not suitable as a predictive early warning system. Large groups of fires tended to predict well, while isolated fires were often missed by the network. Along with further work, the researchers hope to improve the performance of the network so that it can also serve as an early warning system.

The team took advantage of the fact that fire predictions were generally correct to test the effect of different land management strategies. By simulating different interventions, they found that the most plausible plausible strategy would be to convert shrubs and brush to swamp forests, which would reduce the incidence of fires by 50%. If this was combined with the blocking of all but the main drainage channels, the fires would decrease by 70% in total.

However, such a strategy would have obvious economic disadvantages. “The local community desperately needs a stable, long-term culture to boost the local economy,” says Horton.

Another strategy would be to establish more plantations, as good management greatly reduces the risk of fire. However, plantations are among the main drivers of forest loss, and Horton points out that “plantations are mostly owned by large corporations, often based outside Borneo, which means profits are not directly fed back into the environment. local economy beyond the provision of labor for the local workforce.

Ultimately, fire prevention strategies need to balance risk, benefit and cost, and this research provides the insights to do just that, says Professor Matti Kummu, who led the study team. “We tried to quantify how the different strategies work. It is more about informing policy makers than providing straightforward solutions.

Source of the story:

Materials provided by Aalto University. Note: Content may be edited for style and length.

Leave a Comment

Your email address will not be published.