Airborne Digital-Image Data for Monitoring the Colorado River Corridor below Glen Canyon Dam, Arizona, 2009: Image-Mosaic Production and Comparison with 2020 and 2005 Image Mosaics.

摘要

Airborne digital-image data were collected for the Arizona part of the Colorado River ecosystem below Glen Canyon Dam in 2009. These four-band image data are similar in wavelength band (blue, green, red, and near infrared) and spatial resolution (20 centimeters) to image collections of the river corridor in 2002 and 2005. These periodic image collections are used by the Grand Canyon Monitoring and Research Center (GCMRC) of the U.S. Geological Survey to monitor the effects of Glen Canyon Dam operations on the downstream ecosystem. The 2009 collection used the latest model of the Leica ADS40 airborne digital sensor (the SH52), which uses a single optic for all four bands and collects and stores band radiance in 12-bits, unlike the image sensors that GCMRC used in 2002 and 2005. This study examined the performance of the SH52 sensor, on the basis of the collected image data, and determined that the SH52 sensor provided superior data relative to the previously employed sensors (that is, an early ADS40 model and Zeiss Imagings Digital Mapping Camera) in terms of band-image registration, dynamic range, saturation, linearity to ground reflectance, and noise level. The 2009 image data were provided as orthorectified segments of each flightline to constrain the size of the image files; each river segment was covered by 5 to 6 overlapping, linear flightlines. Most flightline images for each river segment had some surface-smear defects and some river segments had cloud shadows, but these two conditions did not generally coincide in the majority of the overlapping flightlines for a particular river segment. Therefore, the final image mosaic for the 450-kilometer (km)-long river corridor required careful selection and editing of numerous flightline segments (a total of 513 segments, each 3.2 km long) to minimize surface defects and cloud shadows. The final image mosaic has a total of only 3 km of surface defects. The final image mosaic for the western end of the corridor has areas of cloud shadow because of persistent inclement weather during data collection. This report presents visual comparisons of the 2002, 2005, and 2009 digital-image mosaics for various physical, biological, and cultural resources within the Colorado River ecosystem. All of the comparisons show the superior quality of the 2009 image data. In fact, the 2009 four-band image mosaic is perhaps the best image dataset that exists for the entire Arizona part of the Colorado River.