WSP00259 - Laserfiche WebLink

Home Browse Search

. . Attribute files were built using ERIS (Earth Resource Information System) (ESL, Inc., 1987a) for both the section and irrigation overlays. These files enabled every polygon to carry mUltiple attributes. The ERIS file for the section overlay contained ~ one variable de~cribing the section as having been defined by USGS quad, fence lines, , or visual estimation. Each field in the irrigation overlay carried eight attributes: ~ basin, quad, tqwnship, range, section, digitizer, photointerpreter, and irrigation '\ status. Digiti~er and photointerpreter attributes contained the last names of the ~ individuals who digitized and photointerpreted the data. The irrigation attribute ~ contained information on the irrigation intensity and status (high, low, or none) extracted from tpe 1987 satellite scenes. Attribute files are very useful for querying large databases because the user can quiCkly select information that is pertinent to his needs. For example, a user may want to tabulate acreages of all fields in basin 14020002 with high-intensity irrigation. A similar query might call for a plot of all fields in the Crawford quad with any irrigation, regardless of intensity. The database was designed to anticipate a wide variety qf queries and to permit updating in an efficient manner. Use of mUltiple-attribute files is not without its problems, though. Digitizers must perform intensive quality control and database checks. One measure we employed was to compile attribute files one quad at a time. Attributes were only incorporated in the total database after a 'quad had been completely entered and checked. This kept interactive procedures spee4y because file sizes were small. It also minimized risk to the attribute database as a whole through inadvertent purging or modifications of "good" records. Greenness Transformation Three SPOT and two Landsat HSS (multispectral scanner) satellite images were acquired for the:project over the Upper Gunnison Basin in June of 1987. Imagery acquisition was timed to coincide with the period of maximum crop cover prior to the first cutting of hay. Greenness images were prepared from the digitial MSS (Kauth, et. al.. 1979) and SPOT (Verdin, et. al., 1987) scenes. Greenness coefficients were adjusted for sun angle and applied: to each of the multiband images to create single band images with grey values indicating relative vegetative cover. The larger the grey value, the more robust and green the vegetative cover for a given pixel (ground-resolution unit). Ground control points located on the satellite images were used to transform the digitized field-boundary data to the satellite image pixel coordinate systems (20-meter cells for the SPOT images and 57-meter cells for the Landsat images). Rasterization converted the digitized vector data to raster, or cell, format equivalent with the satellite images. This was done so that the satellite greenness images could be used to determine the 1987 irrigation status of the fields interpreted from NHAP photography. Because of the lower resolution of the Landsat data, it was only used where SPOT coverage was lacking, which was approximately 10 percent of the total irrigated area. Graphics maskS of the digitized data were overlaid on the greenness images to aid in determining the' relation between greenness values and irrigation practice (Figure 2). The relative greenness values were then mapped to three values, or categories _ no irrigation, low intensity irrigation, high intensity irrigation - based on location with relation to delineated polygons, information gained during the summer field reconnaisance, and appearance on the NHAP photography.