Live fuel moisture content: variability, predictability and impact on fire behavior and activity

Abstract

Live Fuel Moisture Content (LFMC) – the ratio of water mass to the dry mass of live fuel - is a critical factor of fire behavior and hazard. This parameter is largely controlled by weather conditions and is affected by climate changes. There is therefore an increasing need, to understand its variability, to improve its predictability and its impact on fire behavior and activity. This would enhance the development of tools for operational fire risk management and wildfire research. Here we compile several recent findings regarding these issues, most of them were based on the French LFMC database collected for operational purposes and containing more than 20,000 measurement dates during 22 fire seasons on 30 sites and 25 species distributed over the French Mediterranean. First we evaluated the predictability of LFMC by fitting linear relationships between LFMC and various daily empirical drought indices (Drought Code=DC, Keetch-Byram Drought Index=KBDI, Dead Moisture Code=DMC) and a water balance model describing the Relative Water Content of the soil (RWC). We found a limited explanatory power of the drought indices based on climate only due to large differences between sites and species. This support the view that predictions can be improved by accounting for stand and species specific parameters by developing more process-based approaches. The RWC model is a first step toward this direction, and can account for stand level parameterization of soil properties and leaf area index of the vegetation if data are available. To account for species specific traits involved in plant dehydration, we suggest an adaptation of a plant hydraulic model to simulate LFMC response to water potential according to plant hydraulic traits. This is promising with regard to the development of a fully-mechanistic approach of the prediction of LFMC, which would include the responses of both plant and soil. Second, we used an existing dataset reporting shrub fire experiments to investigate the response function of fire rate of spread (ROS) to LFMC. We found a very significant effect of LFMC below 100 %. However, because most LFMC values in the ROS database are higher than those prevailing in the French database, large uncertainties in ROS estimation were obtained below LFMC of ca. 70%. Finally, we explored the relationship between and fire occurrence using a French database of fire activity and the French LFMC database. We found that the response of the fire occurrence of more than 1ha was very similar to the response of rate of spread. Our global approach drawing on long-term field LFMC data constitutes an important step forward in weaving the mechanistic links between climate, vegetation functioning and fire activities.