Development of Methodology to Classify Historical Panchromatic Aerial Photography. Analysis of Landscape Features on Point Au Fer Island, Louisiana - from 1956 to 2009: A Case Study

Recent studies of coastal Louisiana landscapes have shown an increased connection between historical episodic events and current landscape condition. Therefore, the importance of historical landscape reconstruction through the interpretation of panchromatic aerial photo-graphy has increased because it provides synoptic views of hydrology, vegetation, and ecosystems for time periods when data options are limited. Though panchromatic aerial photographs provide a valuable historical record of past landscape conditions, their use is limited in current landscape analyses due to issues with established automated techniques to classify these data (e.g. only one gray level band, and illumination inconsistencies), and the subjectivity and time-intensive nature of human-derived photo-interpretation products. This report documents a method that was developed to improve panchromatic aerial photography classification by increasing accuracy and control and reducing the time-intensive nature of this technique. This method provides a novel approach to selecting landscape features based on a specific range of pixel values (color), contrast, texture, and pattern within a single gray level band of source photography. The resulting techniques were evaluated and used to classify and assess historical land and shoreline change at Point Au Fer Island (PAFI), Louisiana. Assessments show that though this method is more time-intensive than the automated classification approaches used with color-infrared and multispectral data, it provides many advantages over previous panchromatic aerial photography classification methods. These advantages include the use of image level, contrast, and color adjustments; tools to rapidly select features of similar characteristics; and adjustment layers to enhance the visual identification of the land-water features and interface.