||For the past two decades, monitoring and research teams in the physical, biological, and cultural resource programs within the Glen Canyon Environmental Studies (GCES), and now the Grand Canyon Monitoring and Research Center (GCMRC), have been monitoring and modeling the effects of the Glen Canyon dam flows on various ecological resources within the Colorado River ecosystem (CRE). The overall objective of these programs is to determine flow regimes that maintain the resources at recent levels, and possibly restore the resources to pre-dam conditions. The research and monitoring has been performed mostly by in situ measurements, supplemented by annual airborne image data provided by the information technology (IT) program. The image data that were acquired generally consisted of analog, stereo black-and-white photography (color-infrared or CIR photography in particular locations) at 11-cm spatial resolution. These data were point-perspective (unrectified) without pointing or camera information necessary to rectify (georeference) the data to make accurate maps or to perform photogrammetry to derive accurate topography. Correct use of these image data by scientists required a complex process to transform the distorted, point-perspective analog data into an undistorted (rectified), map-projected digital form so that accurate information could be obtained for any particular area. The complexity of the process did not encourage many scientists to use the image data to its fullest potential or accuracy. Therefore, the approaches that were used by GCMRC cooperators before the year 2000 were similar to approaches used by image scientists in the early 1970's. The GCMRC monitoring and research programs (i.e., physical, biological, and cultural) were reviewed by external protocol evaluation panels (PEP) within the past few years (Wohl et al., 1999; Doelle et al., 2000; Urquhart et al., 2000; Anders et al., 2001; Jones et al., 2001). In general, these panels recommended that these programs conduct more integrated, corridor-wide monitoring in order to more accurately determine the effects of dam flow on ecosystem resources. In addition, the Remote Sensing PEP for the IT Program (Berlin et al., 1998) recommended that more modern, advanced remote-sensing technologies be examined to provide better data to the research programs. These two factors prompted GCMRC to establish a remote-sensing initiative whose purpose was to determine the most appropriate remote sensing technologies and approaches that could increase the capabilities and efficiencies of the research scientists in order to help them perform more integrated, less-invasive, corridor-wide studies. The first step in that initiative, which started in the fall of 2000, was a review of the types of ecological parameters being monitored, the collection methods being used, the precision required for each parameter, and alternative remote-sensing and GIS approaches for such monitoring. The latter aspect involved a review of published literature to determine technologies and approaches that produced useful results for problems analogous to those faced by GCMRC. The useful approaches were reviewed in Davis (2002a), which also includes a table of over 100 operational airborne and spaceborne sensor systems that lists relevant characteristics of the sensors. The sensor table was used to select appropriate sensors for consideration and possible evaluation, based on their capabilities for meeting the requirements for a particular program parameter. Although the review found that many resource parameters that are currently monitored could not be adequately approached using airborne remote-sensing technology, these being mostly chemical characteristics of water, the review also found an equal number of parameters that might be approached, if data of the correct type and format were collected and provided to the scientists. During this initial fact-finding process, we found that the level of detail recorded by previous airborne data collections was not being used during scientific analysis, despite initial claims by scientists that they needed the high resolution provided by historical data. This initial review also produced a table of CRE resource categories whose monitoring might be enhanced by remote sensing, along with the types of remote-sensing data that might satisfy required measurement accuracies; the types of data subject to investigation are listed in Table 1 in order of increasing complexity and generally increasing cost. Personnel involved in the remote-sensing initiative collected and analyzed remote-sensing data listed in the table, starting with the least complex and proceeding to the most complex data, until a particular data set was found to provide acceptable accuracies for a particular resource parameter. This approach was followed because cost of data is an issue. The remote-sensing initiative was supervised by Mike Liszewski (IT program manager) and coordinated by Philip Davis (research scientist at the U.S. Geological Survey). The initiative involved all of the IT personnel and many scientists from different disciplines, whose expertise was required for evaluation of specific remote-sensing technologies, data-provider performance, data-analysis methods, and resulting accuracies. This process is now near completion. This report reviews the GCMRC program objectives and measured parameters and the results from our remote-sensing investigations on those parameters that might benefit from improved data acquisition and/or analysis. The resources are discussed in order of the increasing capabilities found by remote-sensing approaches. Thus, the order of our discussions proceeds from the cultural resource program to the biologic resource program and then the physical resource program. Remote sensing of radiation on Earth is limited to the wavelength region from the visible to microwave energies and it is this broad energy region that we have investigated for monitoring applications within the CRE. The different types of remote-sensing data sets that were collected and investigated for the different environmental parameters are listed in Table 2. Radar data were not included in this evaluation because the aircraft used for radar data collections are large, impossible to maneuver in the canyon, and provide too low spatial resolutions (about 5 m) when flown above the canyon rim. In addition, the walls of the canyon can produce secondary radar reflections that interfere with the primary reflections and make the image data unintelligible. The collection of numerous wavelength bands by multi- and hyperspectral sensors limits the spatial resolution that can be achieved by such sensors because of data-rate limitations of current storage devices. Spectral resolution refers to the wavelength band width for a particular image, whereas spatial resolution refers to the surface dimension of a single picture element within that image. For example, multispectral sensors that record 12 wavelength bands with 50 nm spectral resolution can obtain image data at a 1-m ground resolution, while hyperspectral sensors that record up to 220 wavelength bands with 10 nm spectral resolution can only obtain data at a 2-4 m ground resolution. The cost for image data increases with the number of wavelength bands collected, which effects the benefit/cost ratio and makes use of more sophisticated data difficult to justify, unless these data provide information that cannot be obtained by more simple, less expensive data. Therefore, our evaluations proceeded from the simpler to the more complex data sets, until a viable data set was found for a particular resource parameter. The final section presents the team's recommendations for future remote-sensing monitoring activities based on all of our investigations.