Examining Drivers of Post-Wildfire Vegetation Dynamics Across Multiple Scales Using Time-Series Remote Sensing

Abstract

Ecosystem response to disturbance is a function of environmental factors interacting at a number of spatio-temporal scales. This research explored ecosystem response to wildfire as a function of local and broad-scale environmental factors using satellite based time-series remote sensing data. This topic was explored as a series of three independent but related studies. The first study focused on the evaluation of techniques for the analysis of time-series satellite data for describing post-fire vegetation trends at sites in the US, Spain, and Israel. Time-series data effectively described post-fire trends, and reference sites were valuable for differentiating between post-fire effects and other environmental factors. The use of phenological indicators derived from the time-series shows promise as a monitoring tool, but requires further investigation. The next study evaluated the influence of broad-scale climate factors on rates of post-fire vegetation regeneration across the western US. Rates of post-fire regeneration were higher with increased precipitation and higher minimum temperatures. Changes in climate are likely to result in shifts in post-fire vegetation dynamics, leading to important feedbacks into the climate system. The use of time-series data was a valuable tool in measuring trends in post-fire vegetation across a large area and over an extended period. The final study used time-series vegetation data to measure variations in post-fire vegetation response across an extensive 2002 wildfire. Regression tree analysis related post-fire regeneration to local environmental factors such as burn severity, soil properties, vegetation, and topography. Residuals from modeled rates of post-fire regeneration were evaluated in the context of management activities and site characteristics using expert knowledge. Post-fire rates of regeneration were a function of water availability, pre-burn vegetation, and burn severity. Management activities, soil differences, and shifts in vegetation community composition resulted in deviations from the modeled post-fire regeneration rates. The results of these three research studies indicate that remotely sensed time-series vegetation data provide a useful tool for measuring post-fire vegetation dynamics. Both broad-scale and local environmental factors play important roles in defining post-fire vegetation response, and the use of remote sensing and geospatial data sets can be useful in integrating these factors and enhancing management decisions.