Skip to main content
ARS Home » Research » Publications at this Location » Publication #168056
Title: ISSUES IN TERRAIN ANALYSES: DIGITAL ELEVATION MODEL GENERATION AND PROCESSING TOOLS

Author
item Erskine, Robert - Rob
item Green, Timothy
item RAMIREZ, JORGE - COLORADO STATE UNIVERSITY

Submitted to: Meeting Proceedings
Publication Type: Abstract Only
Publication Acceptance Date: 7/23/2003
Publication Date: 11/2/2003
Citation: Erskine, R.H., Green, T.R., Ramirez, J.A. 2003. Issues in terrain analyses: digital elevation model generation and processing tools. Meeting Proceedings. 2003 American Society of Agronomy/Crop Sci Society of Amer/Soil Science Soiciety of Amer Annual Meeting. Nov. 2-6, 2003. S05-erskine528369-poster (CD-ROM).

Interpretive Summary: Global positioning system (GPS) data, as collected from harvester-mounted yield monitors, provide an efficient means for terrain modeling of agricultural sites. In this study, yield monitor position data are interpolated to grid digital elevation models (DEMs) for three sites (approximately 65 ha each) in northeastern Colorado. Satellite-differentially corrected GPS (DGPS) providing a vertical RMSE of 0.85 m is used with this yield monitor. This error is based on residuals between the DGPS-derived DEMs and DEMs derived from position data collected from a dual-frequency, real-time kinematic GPS. This GPS method produced a vertical RMSE of 3 cm, relative to National Geodetic Survey vertical control. The third data source investigated is the published 30-meter USGS DEMs for these sites, which yielded a vertical RMSE of 1.49 m. Landscape topographic attributes are easily estimated from these grid DEMs, however, as shown in this and other research, these estimates are sensitive to the DEM data source and grid resolution, as well as the flow routing algorithm. The effects of changing the DEM data source and grid resolution on attribute estimation are presented. Furthermore, an assessment of DEMs and flow routing algorithms is performed by correlating a suite of topographic attributes, estimated by the various DEMs and algorithms, to dryland winter wheat crop yields.

Technical Abstract: Global positioning system (GPS) data, as collected from harvester-mounted yield monitors, provide an efficient means for terrain modeling of agricultural sites. In this study, yield monitor position data are interpolated to grid digital elevation models (DEMs) for three sites (approximately 65 ha each) in northeastern Colorado. Satellite-differentially corrected GPS (DGPS) providing a vertical RMSE of 0.85 m is used with this yield monitor. This error is based on residuals between the DGPS-derived DEMs and DEMs derived from position data collected from a dual-frequency, real-time kinematic GPS. This GPS method produced a vertical RMSE of 3 cm, relative to National Geodetic Survey vertical control. The third data source investigated is the published 30-meter USGS DEMs for these sites, which yielded a vertical RMSE of 1.49 m. Landscape topographic attributes are easily estimated from these grid DEMs, however, as shown in this and other research, these estimates are sensitive to the DEM data source and grid resolution, as well as the flow routing algorithm. The effects of changing the DEM data source and grid resolution on attribute estimation are presented. Furthermore, an assessment of DEMs and flow routing algorithms is performed by correlating a suite of topographic attributes, estimated by the various DEMs and algorithms, to dryland winter wheat crop yields.