Modeling Multiple Risks in Forests

Context and challenges

In the past 20 years, European forests have lost 44 million m³ of timber, representing 16% of the average annual harvest (Patacca et al. 2023). Storms account for 46% of the damage, wildfires for 24%, and bark beetles for 17%. These disturbances have increased and are expected to continue intensifying as a result of climate change (Seidl et al. 2014). Climate change increases not only the frequency and intensity of these hazards but also their interactions, such as droughts that facilitate pest infestations or wildfires. These cascading interactions can significantly amplify the scale of damage, highlighting the need for an integrated approach to forest risks (Bastit et al. 2023).

In this context, the socioeconomic challenges primarily involve sustainable forest management, which is currently under threat. Indeed, these recurring and interacting natural risks jeopardize the provisioning of forest-based environmental goods and services, such as timber production and carbon sequestration. As a result, the entire supply chain—from timber producers to processing to consumers—is weakened. Likewise, the government’s environmental commitments regarding carbon sequestration (SNBC, Stratégie Nationale Bas Carbone) are unlikely to be met. The stakes are therefore high, with an urgent need to incorporate this “multiple risks” dimension into forestry models.

Goals

The aim of this thesis is to model the complex interactions between biotic and abiotic natural hazards affecting forests, to assess their economic consequences and their impact on the forest carbon sink, and to identify and characterize effective ex ante and ex post economic measures to limit the negative consequences of these multiple hazards on the forestry sector and carbon sinks. For this, a prospective approach using the French Forest Sector Model (FFSM) will be implemented, with substantial improvements made to the model. 

First, the use of remote sensing data will enable the production of predictive maps of damage caused by multiple hazards (wildfires, storms, bark beetles), which will be integrated exogenously into the FFSM, in line with the current approach. 

Second, an internal module will be developed to endogenously model the complex interactions between multiple hazards and their cascading effects. This module will be calibrated using machine learning applied to historical geospatial data. 

Third, the FFSM will be enhanced by directly integrating functions to simulate ex post market mechanisms (timber storage, transportation, price floor) and ex ante forest management mechanisms (insurance). The methodology will combine the development of probabilistic scenarios involving multiple hazards, the comparative simulation of ex post market mechanisms and ex ante management mechanisms, and cost-benefit analyses. A complementary component will use remote sensing to assess forest carbon losses on a European scale. The ultimate goal is to identify optimal trade-offs between economic stability and maintaining carbon sinks amid increasing forest disturbances.

INRAE structure