Presentation

View presentation here

Abstract

Live fuel moisture content (LFMC; mass of plant water per unit dry biomass) has been shown to be a key determinant of fire ignition and spread, but this influence has been studied relatively little at ecosystem-scale, likely because of a lack of data on LFMC at large scales. We recently developed wall-to-wall maps of LFMC for the western US at 15-day intervals and 250 m resolution from 2016 - present based on microwave remote sensing. Since LFMC’s response to climate can vary widely with plant and soil traits, such remotely-sensed maps can help us analyse LFMC’s effects on wildfires. Here, we used these maps to quantify the effect of LFMC on wildfire occurrence and burned area. In a first analysis, we used historic wildfire records from 2016 - 2020 using the Moderate Resolution Imaging Spectroradiometer (MODIS) burned area product to reveal that wildfire occurrence patterns could be predicted with greater accuracy if LFMC, rather than only meteorological factors, was accounted for. Including LFMC as an explanatory varirable improved prediction accuracy by 2-9% depending on land cover, indicating that meteorological factors alone were insufficient to capture landscape-scale heterogeneity in LFMC. This implies that patterns of LFMC are not fully explained by meteorological factors alone and that accounting for LFMC’s different responses to weather in different locations ( for e.g., due to differences in species cover and plant hydraulic traits) can improve ignition risk estimation. We further tested the hypothesis that LFMC’s sensitivity to drought-like conditions regulates the increase in wildfire burned area due to climate change-driven rise in vapor pressure deficit (VPD; an indicator of atmospheric aridity). We show that the temporal slope between annual burned area and mean fire-season VPD is strongly linked to the sensitivity of LFMC to climate-derived moisture balance (termed, plant-water sensitivity) with R2=0.71. For the same rise in VPD, burned area increased by more than twice as much in ecosystems with high plant-water sensitivity compared to those with low plant-water sensitivity. This has led to rapid increases in human wildfire risk, both because the population living in areas with high plant-water sensitivity grew 50% faster during 1990-2010 than in other wildland-urban interfaces and because VPD has risen more rapidly than average in these vulnerable areas.