A Framework for Wildfire Prediction and Loss Assessment: Potential Application in the Financial Sector—Case Study in Yunnan
This working paper introduces a climate risk assessment framework that leverages remote sensing data and climate projections to enhance the understanding and management of wildfire risks for financial institutions under climate change scenarios. Focusing on Yunnan Province's agriculture sector, the study assesses how changes in wildfire severity and frequency impact five key crops under different climate scenarios, revealing significant threats to smallholder farmers and associated risks for financial institutions.
- This paper introduces a framework that combines remote sensing data and climate projections, presenting an alternative to traditional physical models for assessing and managing the wildfire impact on assets under future climate warming projection.
- Through showcasing the agriculture sector in Yunnan, the paper reveals changes in severity and frequency of wildfires in Yunnan and impact on five different crops under different climate scenarios.
- With extreme wildfire events accounting for 15 to 30 percent of annual losses, a single event poses significant threats to smallholder farmers, which can lead to noticeable risk for financial institutions.
- Financial institutions in Yunnan are facing complex climate-induced impacts and the challenges of limited readiness and capacity to manage risk, requiring a comprehensive solution and external empowerment.
Key Findings:
Based on the research conducted, the study draws the following conclusions regarding Yunnan Province:
- During 2011 to 2020, historical loss from wildfires in Yunnan’s five target agricultural crops ranged from US$135.8 million to $254.8 million per year.
- Extreme wildfire events account for 15 to 30 percent of annual loss, posing a significant threat to smallholder farmers.
- The current framework of the wildfire loss simulation model shows good performance in most of the crops through testing, providing reliable results for practical use.
- In Yunnan Province, historical data shows the highest regional mean Fire Weather Index (FWI) in March and April, but future projections shift the peak to February and March. Among the scenarios, Shared Socioeconomic Pathway (SSP)2-45 shows the highest mean FWI, followed by SSP1-26, SSP3-70, and SSP5-85, all exceeding observed values.
- Scenario analysis of wildfire loss demonstrates the complexity of the impact of wildfires on crops under climate change. We did not observe a trend where higher emissions led to greater impacts under different climate scenarios.
- Insurance companies are highly enthusiastic about integrating physical climate risk analysis to develop relevant products like catastrophe and parametric insurance. Banks prioritize using climate risk analysis tools for effective credit risk management. Stress testing for climate risks is crucial but resource intensive.