Relationship of a Landsat cumulative disturbance index to canopy nitrogen and forest structure

Spatially-explicit knowledge of the timing, frequency, and intensity of forest disturbances is essential for forest management, yet little is known about how disturbances such as forest harvests and insect outbreaks might accumulate in their effects over time. Capturing the many forest harvest and insect defoliation events occurring over twenty-five years, we transformed a series of Landsat images into cumulative disturbance maps covering Green Ridge State Forest (GRSF) and Savage River State Forest (SRSF) in western Maryland. These maps summed yearly ΔDI images, which were defined as the change in a yearly tasseled cap disturbance index (DI), relative to a synthetic reference condition map created by finding the minimum DI value for all years. Intensive field-plot surveys and AVIRIS imagery collected during the summer of 2009 provided measurements of forest structure and canopy nitrogen. With these data, we found that while the most recent year's ΔDI had little relation, increases in the cumulative DI were related to decreased field-measured current canopy cover (R2 = 0.66 at GRSF, 0.67 at SRSF and 0.34 combined) and watershed-averaged AVIRIS canopy N (R2 = 0.40 at GRSF, 0.57 at SRSF and 0.54 combined). The latter relationship was obscured at the field-plot level of analysis, suggesting that fine scale studies will also need to account for other drivers (e.g. species composition) of variability in canopy N. Nevertheless, our study demonstrates that Landsat time series data can be synthesized into cumulative metrics incorporating multiple disturbance types, which help explain important cumulative disturbance-mediated changes in ecosystem functioning.