Submitted to: Annual American Geophysical Union Hydrology Days
Publication Type: Proceedings
Publication Acceptance Date: 8/1/1999
Publication Date: N/A
Citation: Interpretive Summary: Surface runoff and soil erosion rates are often estimated by use of computer program that require a description of land surface characteristics. The measurement of such characteristics in the field or from maps is time consuming, but can be automated by using a digital representation of the land surface. This paper describes the algorithms and procedures for a computer to automatically extract land surface slope and length value from the digital data. In addition, the performance of the procedures was evaluated by comparing automatically- and map-derived characteristics. Automated and map-derived properties can produce different results. Differences were attributed to underlying assumptions in the methods, and no one method was found to be superior to another. The characteristics obtained by the automated procedure were provided in tabular and digital format for easy interface with runoff and erosion estimation programs. The automated procedure was particularly effective for large watersheds that contain numerous subcatchments for which the traditional evaluation is not practical.
Technical Abstract: Topographic land surface properties, such as elevation, slope or aspect, generally vary in space, and a unique value for a particular subcatchment area is generally not available. In this study several methods are presented to estimate a representative subcatchment length and slope value directly from Digital Elevation Models. Data-Reduction (DR) models are used to reduce the spatially varying length and slope values into a single representative value for the subcatchment. The representative values obtained from different DR models are compared to map-derived measurements. Significant differences between calculated and map-derived length and slope values are found. The differences are not approximation errors, but resulted from the different definition of representative length and slope and different DR models applied to the property under consideration. No one method is found to be superior to another, including the map-based method. It is recommended that the selection of a particular representative value must be made as a function of its particular application. For example, representative slope values based on local terrain slopes are believed to be more appropriate for vertical energy and water vapor flux applications, whereas representative slope and length values based on flow paths are believed to be more suitable for runoff, erosion and contaminant transport applications.