Long-term post-wildfire monitoring of phenology and recovery using a MODIS time series

Disturbance severity in forests or grasslands is generally conceptualized as the difference between the state before and immediately after a disturbance event. This approach fails to capture slow-acting disturbances that take years to materialize, secondary disturbance effects such as delayed mortality, or variable rates of recovery. Remotely sensed data can provide a multi-seasonal baseline and long-term post-disturbance record of recovery that also captures associated disturbances, such as post event salvage logging or restoration efforts. Here we track the MODIS satellite-based Normalized Difference Vegetation Index or NDVI of several large wildfires that occurred early in the last decade to measure fire and associated disturbance severity with multi-seasonal and multi-year contexts. Large fires analyzed included Oregon's Biscuit Fire, Colorado's Hayman Fire, Arizona's Rodeo-Chediski Fire, and Georgia's Okefenokee Fire among others. Short-term results were generally consistent with prior post-fire estimates of short-term wildfire severity, but long-term fire effects diversified. Both short and long-term severity varied by topography and vegetation types, as measured by changes in seasonal NDVI, not just single-season NDVI. This broadened monitoring technique provides a moderate resolution record of recovery that can inform questions related to cumulative impacts and ecological resilience.