Development of a Historical Multi-Year Land Cover Classification Incorporating Wildfire Effects

Abstract

Land cover change impacts ecosystem function across the globe. The use of land cover data is vital in the detection of these changes over time; however, most available land cover products, such as the National Land Cover Dataset (NLCD), are produced relatively infrequently. The most recent NLCD at the time of this research was produced in 2006 and does not adequately reflect the impact of land cover changes that have occurred since, including the occurrence of two large wildfires in 2008 in our study area. Therefore, there is a need for the classification of historical remotely sensed data, such as Landsat scenes, through replicable methods. While it is possible to collect field data coinciding with current or future Landsat acquisitions, it is impossible to retrospectively collect data for previous years; thus, fewer studies have focused on the classification of historical scenes. Using a single year of field reference and multi-year aerial photography data, we applied a simple decision tree classifier to accurately classify historic satellite data and produced maps of land cover to incorporate the effects of 2008 wildfires occurring between NLCD production dates. Overall accuracy ranged from 76 to 90 percent and was assessed using conventional error matrices.

Fire and the Distribution and Uncertainty of Carbon Sequestered as Aboveground Tree Biomass in Yosemite and Sequoia & Kings Canyon National Parks

“…Although the present study used a sufficient number of forest plots (1646), studies with too few plots to represent the landscape heterogeneity also may have errors related to the time since disturbance, a principal contributor to heterogeneity [24]. (5) Vegetation type mismatch-when plot level data are scaled to the landscape, errors in vegetation cover assignment can yield errors in scaled biomass.…”