Laserfiche WebLink
Three adjacent Landsat TMS images were processed to reflectance using header data <br />packaged with each satellite scene and equations provided in Irish (1999). Reflectance is <br />the ratio of the reflected light measured at the satellite divided by incoming solar <br />radiation. Reflectance is an important remote sensing measure because it incorporates a <br />series of calculations that correct for sensor differences and for the strength of solar <br />radiation that is highly influenced by time of year. <br />Following calculation of reflectance, the TM images were geocorrected. Three images <br />were received in NLAPS format from RTi that, together, provided complete cloud-free <br />coverage for the phreatophyte study area. Landsat images are spatially referenced by the <br />Worldwide Referencing System (WRS), with apath/row naming convention that enables <br />a user to easily obtain the correct satellite image for a particular area. The three satellite <br />images that provided coverage for the area of interest are WRS Path/Row 32/32, 33/32, <br />and 33/33 (Figure 3). These images were chosen because they captured peak growing <br />season expression of phreatophyte vegetation and were cloud free. <br />i~ <br />1 <br />.j <br />~~ "~ <br />' ' ~~~ j` ~ 32132 <br />~ 33132 <br />~ 33133 <br />~ Study Area <br />N <br />~ so ~©0 2ao <br />Kilometers <br />Figure 3. <br />Landsat TM scenes that <br />cover the South Platte <br />study area (blue) <br />defined by the mapped <br />boundary of the South <br />Platte alluvium. The <br />alluvium boundary <br />defined the outside <br />edge of the potential <br />area for evaluation of <br />riparian ETg. <br />Imagery dates are <br />32/32 -July 16, 2001 <br />33/32 -July 7, 2001 <br />33/33 -July 7, 2001 <br />Each image was registered to the DOQQs using ground control points from known, <br />permanent locations, such as bridges and major intersections visible on both DOQQs, and <br />TM images. A measure of the accuracy for geocorrection is the root mean square error <br />(RMS). RMS is the approximate scale relative to a pixel; therefore, the average <br />geocorrection error was nominally on the order of about 0.20 of a TM pixel for the entire <br />