2007 GSA Denver Annual Meeting (28–31 October 2007)

Paper No. 11
Presentation Time: 4:15 PM

ACCURACY OF DIGITAL ELEVATION FLOW ACCUMULATION FOR LOCATING OVERLAND RUNOFF FLOWPATHS: A FIELD STUDY IN WESTERN, NY


RICHARDS, Paul L. and NOLL, Mark R., Earth Sciences, SUNY Brockport, 350 New Campus Drive, Brockport, NY 14420-2914, prichard@brockport.edu

A flow accumulation model, weighted by theoretical phosphorous loads, was evaluated in the field to determine if it could identify and rank overland runoff flowpaths. Twelve flowpaths were instrumented with runoff collectors and sampled to determine their potential for contributing phosphorous. Aerial photographs were also used to verify the model. An analysis of flow centerlines predicted from the model suggests there is a 95% probability that the actual flow line will be within 300 feet of the model flowpath. This probability drops to 80% and 50% within 200 feet and 60 feet respectively. Tributary junctions located by the model showed a directional bias downstream of the actual junction. A probability analysis of the distance between model and observed junctions suggest that there is a 95% chance that the modeled tributary is within 885 feet of the observed junction. These errors are larger than what they should be, even if the effects of rasterization are accounted for. Results suggest that while flow accumulation techniques correctly identified some important sites of concern, they cannot be used to precisely locate or rank these features without additional field verification. One explanation for the poor performance of the model is from anthropogenic modifications to topography which occurred after the DEM was created or topographic modifications that are too small to be represented by the resolution of the DEM (10 meters). It is important to note that the more advanced hydrologic models that utilize topography for routing overland ruoff are also subject to the same problems that made our model inaccurate. Field observations suggest that subtle features (road berms, tile drains, channel alteration, landform modification due to development) can have a tremendous impact on surface drainage. In some cases these changes lead to hydrologic connectivity to the drainage network, while in other cases once-connected flowpaths are shut off.