AUTOMATED VEGETATION MAPPING USING DIGITAL ORTHOPHOTOGRAPHY

We used near-infrared digital orthophotography and three collateral da ta sets to model ecological communities on Block island, Rhode Island. Aerial photography of the island was taken on 19 May 1992 at a scale of 1:40,000. The photography was scanned and processed to remove disto rtions from terrain, aircraft tilt, and optical aberration. The result ing digital orthophotograph was comprised of three spectral bands repr esenting the red, green, and blue colors of the scanned photography an d had a pixel dimension of 1.27 m. Three textural variables were devel oped by calculating the standard deviation within a 10-m radius of eve ry pixel for each of the three spectral bands in the image. The terrai n model that was used to create the orthophoto was also used to derive slope and aspect for each pixel. Soil survey data were used to map th e distribution of soil drainage classes to distinguish wetland from up land vegetation. We used linear discriminant analysis to develop a mod el to distinguish 11 vegetation and cover classes on the island. The f ull model consisted of nine independent variables derived from the ort hophoto, the textural indices, terrain metrics, and soils. Classificat ion accuracies ranged from 60 to 80 percent for an independent validat ion data set. The variable DRAINAGE CLASS dominated the model and expl ained the most variation in vegetation and cover class.