Updating Spatial Datasets Using Geographic Information Systems and Remote Sensing Techniques

With an increase in the availability and application of remote sensing and Geographic Information Systems (GIS), the development of new methods to accurately and efficiently update spatially derived dataset has become important. Large and small- scaled landcover datasets are costly and time-consuming ventures. Landscape altering events often occur before a mapping project has been completed. Such was the case with the existing vegetation and fire fuel load spatial datasets of Petersburg National Battlefield (PETE) and Shenandoah National Park (SHEN), where Hurricane Isabel and the Rocky Top fire, respectfully, caused extensive damage to the natural environment. In both instances a GIS coupled with imagery and Remote Sensing techniques were used to update the pre-existing datasets. An object-oriented classifier; VLS's Feature Analyst was used to map areas of downed trees at PETE. Various classification methods were used to map the Rocky Top fire perimeter and the severity of it upon vegetation. The differenced Normalized Burn Ratio (dNBR) was successful at not only mapping the fire's perimeter but in mapping the different severity levels as well. Validated with field work, the techniques used were able to accurately document and quantify changes that occurred to the natural environment, allowing geospatial datasets to be updated in a time efficient and cost effective manner. Introduction and Study Areas Having recently completed an extensive mapping project for the North East Region of the US National Park Service, North Carolina State University's Center for Earth Observation has begun exploring techniques that will allow small-scaled landcover datasets to be updated in a time efficient and cost effective manner. This is especially important because most parks in the North East Region fall within the Wildland Urban Interface where accurate and current geospatial information is a necessity to effectively implement proper management of available resources. The objective of this research project is to explore several GIS applications to determine which are appropriate and feasible for two types of damage associated with natural resources typically seen in the North Eastern US. Study sites for this project include: Shenandoah National Park (SHEN), where in June of 2002 the Rocky Top Fire burned nearly 1500 acres, and Petersburg National Battlefield (PETE), where in September of 2003 Hurricane Isabel leveled extensive stands of deciduous and coniferous forests.