Wildfire Burn Severity Modeling (under construction)

Since August 2022 I have been working with Dr. Weimin Xi at Texas A&M University - Kingsville to study the 2021 Dixie wildfire in northern California. The goal of this project is to use the Dixie fire's large and heterogenous footprint to better understand the effect that forest stand structure and composition have on wildfire severity.

Skills and software:

Google Earth Engine
R
Forest Vegetation Simulator

Remotely sensing burn severity

To quantify burn severity within the Dixie fire's 980,000 ha footprint, I used Google Earth Engine to calculate relative differenced normalized burn ration (RdNBR) as described in Miller & Thode (2009) from two cloud-free Sentinel-2 composites.

Deriving forest stand structure from FIA plots

We were able to secure permission from the U.S. Forest Service to use the exact coordinates of 385 Forest Inventory and Analysis (FIA) plots in and around the burn footprint. From these plots, I calculated metrics such as:

Trees per acre (stratified by DBH)
Basal area per acre
Percent hardwood
Mean bark thickness
Woody debris loading per acre

In addition, I used Forest Vegetation Simulator (FVS) to estimate canopy base height, number of canopy strata, and crown density.