Effects of a catastrophic blow down event in northern Minnesota, USA were assessed using field inventory data, aerial sketch maps and satellite image data processed through the North American Forest Dynamics programme. Estimates were produced for forest area and net volume per unit area of live trees pre- and post-disturbance, and for changes in volume per unit area and total volume resulting from disturbance. Satellite image-based estimates of blow-down area were similar to estimates derived from inventory plots and aerial sketch maps. Overall accuracy of the image-based damage classification was over 90%. Compared to field inventory estimates, image-based estimates of post-blow down mean volume per unit area were similar, but estimates of total volume loss were substantially larger, although inaccessibility of the most severely damaged inventory plots may have depressed the inventory-based estimate. This represents the first application of state model differencing to storm damage assessment. The image-based procedure can be applied to historical archives of satellite imagery and does not require pre-disturbance field inventory data.
Bibliographical noteFunding Information:
The authors thank the US Department of Agriculture Forest Service, Superior National Forest for funding the intensification of the FIA plot sample within the BWCAW during the inventory years immediately following the blowdown; NASA Applied Sciences and Terrestrial Ecology Programmes for funding Landsat image acquisition, processing and applications development; and Quinn Chavez of US Department of Agriculture Forest Service Forest Health Protection programme for providing aerial sketch map geospatial data of Minnesota 1999 wind damage. The authors also thank the anonymous reviewers and the associate editor for prescribing substantial improvements to the manuscript.