2008 Joint Annual Meeting (5-9 Oct. 2008): Classification of Orthophotos to Identify Vegetative Buffer Strips to Update Existing Land Use Databases.

684-1 Classification of Orthophotos to Identify Vegetative Buffer Strips to Update Existing Land Use Databases.



Tuesday, 7 October 2008
George R. Brown Convention Center, Exhibit Hall E
Myriam Larose, Agronomy, Purdue University, 275 South Russell Street, West Lafayette, IN 47906, Gary Heathman, 275 S. Russell St., USDA-ARS, USDA-ARS-MWA Purdue University, National Soil Erosion Research Laboratory, West Lafayette, IN 47907-2077 and L. Darrell Norton, 275 S. Russell Street, USDA-ARS, Purdue University-USDA-ARS-MWA, National Soil Erosion Research Lab, West Lafayette, IN 47907
Vegetative buffer strips (VBS) including grass buffer, forested buffer (FBS), among others, are important best management practices implemented in agricultural areas to control the movement of potential agricultural contaminants into water bodies The importance of VBS has led the Natural Resource Conservation Service and the United States Department of Agriculture’s Farm Service Agency to administer programs that fund the implementation of VBS by farmers. One such program is the Conservation Reserve Program that primarily funds the use of VBS in USA. Although many farmers have implemented VBS, there are limited science-based resources available to quantify the actual area of VBS implemented and their environmental benefits at the watershed level. The environmental benefits of VBS at the watershed level are difficult to quantify without detailed and accurate inventory of existing VBS. The lack of large scale studies in this area of research is mainly due to no optimal methods be available for determining the exact location and width of existing VBS. The spatial resolution of most available remote sensing imagery, such as the Landsat (30 m), is not of sufficient detail to identify VBS location and characteristics. Therefore, the goal of this study was to use multi-year high resolution orthophotos to identify VBS using segmentation based classifier from the eCognition Definiens software in northern Indiana, USA. The results show that more than 75% of the watershed had either VBS or FBS between the adjacent field and the drainage ditch.  Grass buffers ranged from 5 to 26 m wide.  Forested buffer strips were typically > 30 m width; however, a few were found to only be as wide as 6 m. This study is beneficial to measure the extent of VBS implementation, as well as, to incorporate VBS into existing land use databases for hydrologic and water quality modeling.