Abstract:

Wildfires are increasing in severity and intensity in the Western United States, and preparing for and mitigating the effects of wildfires is needed. This includes the need for accurate wildfire-outlooks, pre-fire fuel treatments and prescribed burning, and responding to events post-fire. Wildfires can often occur in remote areas and in complex terrain. Remote sensing offers a path forward by providing heterogeneous and dynamic observations of fuel load, type, and flammability – and these measurements can be useful for pre-fire mitigation activities, outlooks and post-fire assessments. In this talk I will present on the application of high-resolution thermal remote sensing observations of plant water stress from the Ecosystem Spaceborne Thermal Radiometer on Space Station (ECOSTRESS) (70 m, every 3 – 5 days) and other remote sensing data sets to address wildfire hazard prediction, planning for prescribed burns, assessing fuel treatments and post-fire impacts. Using ECOSTRESS data we trained random forest models to predict 1-week outlooks of fire severity hazard for the state of New Mexico. These models can be useful for identifying which regions are likely to experience more severe effects of burning, and opportunities to plan prescribed burns.  We also used plant stress observations derived from ECOSTRESS to assess the effectiveness of fuel treatments for a case study in Northern California. We examined a region that had fuel treatments before the Caldor wildfire took place and compared the recovery of vegetation in the treated and untreated regions. In addition, wildfires are known to have profound impacts on watershed hydrology by altering land cover and soils. We also explore the impacts of these watershed-scale changes for coastal basins in Southern California with implications on coastal water quality.