WSP00259 - Laserfiche WebLink

Home Browse Search

. . Flood-irrigation practices in the area covered by one of the SPOT images interfered with ttie ability of the greenness transformation to fully delineate green ~ vegetation. Consequently, a brightness transformation was applied to the SPOT image. " The greenness and;brightness images for this area were then combined using a decision ~ rule based on the ,relations between brightness, greenness, healthy vegetation, and wet (':1 soils. Areas with high-intensity greenness values and high-brightness values as well (T.) as areas with low~intensity greenness values and low-brightness values were classified CP as high-intensity irrigation. This accounted for wet conditions in some high-intensity irrigated fields. Areas with low-intensity greenness values and high brightness values were classified as low-intensity irrigation. After the image greenness values were mapped to denote intensity of irrigation, the images were cqmbined with rasterized field-boundary files to produce tabular ERIS files. A digital overlay procedure was used to accomplish this. The files were organized so that :the unique number assigned to each polygon by GES could be used as a cross reference w~th the ERIS attribute files generated from the greenness images. A majority rqle criterion was applied to the greenness tabular files to establish irrigation intens~ty and status for all field polygons. By looking at whether the majority of the ]greenness cells in a given polygon had high- or low-intensity irrigation values, it was possible to make separate files for the two irrigation categories. Polygons with equal numbers of high- and low-intensity cells were placed into the high-intensity-irrigation category. Polygons with zero greenness values were discarded. The high- and low-intensity-irrigation greenness files were then merged with the attribute files, 'using the unique polygon numbers as the matching variable. This permitted preparation of acreage summaries using the areas stored automatically by GES with each field ,polygon, rather than by counting satellite image pixels. Such acreages are of 1a higher quality because field boundaries are more accurately portrayed in a veqtor format than in a raster format. Using the attributes for each polygon, it was then possible to create acreage summaries based on basin number, county, quadrangle', township/range, and section. Tables 1-3 show the types of acreage summaries that were calculated for the irrigated lands in the Upper Gunnison basin. RESULTS Acreages were calculated for all irrigated fields based on 1987 satellite data. Small areas of the, project area in Hinsdale, Montrose, and Gunnison counties were not / covered by the satellite imagery. In these areas (1599.5 acres) irrigated acreages were reported as p~tentiallY irrigated based on the photointerpreted data. After looking at the satellite imagery with the digitized field boundaries overlain, it was ; determined that several fields were being counted as completely irrigated when only parts of the fields were actually irrigated. Many of the fields were flood irrigated and run-off water appeared to cross into other fields. Complete irrigation coverage was also not achieved in some of the flood irrigated fields. To account for these hon-irrigated acres, a correction factor was applied to the data. A percent irrigated 'value was calculated by dividing total irrigated pixels by total pixels for each field. The percent irrigated value was then mUltiplied times the vector acreage for all fields with less than 90% irrigation to get a better accounting of actual irrigated acreage. Field boundaries were determined by location of irrigation ditches, fence lines, roads, and other ~an-made and natural features. This resulted in a large number of field polygons. ,Field complexity could have been significantly reduced by not ,