How does forest disturbance impact albedo as revealed from historical Landsat and MODIS data?

Forests store about 45% of terrestrial carbon and affect global climate through complex biogeochemical and biogeophysical processes. Past research has tended to focus on the biogeochemical processes, such as how afforestation and deforestation affect global warming through carbon cycling. Recent studies have started to consider how changes of radiative forcing due to disturbance (biogeophysical feedback) affect the climate. Several recent publications have reached a similar conclusion that the climate impacts of forest disturbance and regrowth differs for tropical, temperate and boreal forests in large part due to differing albedo responses. However, additional analysis of observations is needed to better quantify these differences, especially considering the large-scale implications for climate mitigation through land management. In this presentation, we will examine albedo changes caused by forest disturbance and regrowth from historical Landsat data and the MODIS albedo product from different regions and times. The forest disturbance maps from Landsat data are used to determine the disturbed area, type and ages. The MODIS albedo product from 2001 to 2007 is used to extract albedo from "pure" disturbed pixels. The changes of albedo are independently analyzed by time of year, for snowy or non-snowy conditions, and along an axes of years since disturbance, and the type of disturbance. Our initial results show an increasing trend of forest albedo in burned areas in the first several years (1-15 years) under snow-free conditions. Under snow-covered conditions, albedo of burned areas increases in the first several years (~7 years) and then becomes stable. In clear-cut regions, albedo is stable under snow-free conditions but decreases under snow-covered conditions, and returns to the pre-clear-cut value in about 23 years. Future work will focus on direct retrieval of albedo changes from the Landsat record, as well as using land surface and biogeochemical models to assess the "whole climate" impact of future land use/disturbance scenarios.