Christopher E. Soulard 20080910 remote-sensing image USGS Data Series 376 Menlo Park, California U.S. Geological Survey http://pubs.usgs.gov/ds/376/ Soulard, C.E., and Raumann, C.G., 2008, Historical orthoimagery of the Lake Tahoe basin: U.S. Geological Survey Data Series 376. The U.S. Geological Survey (USGS) Western Geographic Science Center has developed a series of historical digital orthoimagery (HDO) datasets covering part or all of the Lake Tahoe Basin. Three datasets are available: (A) 1940 HDOs for the southern Lake Tahoe Basin (27,937-ha area), (B) 1969 HDOs for the entire Lake Tahoe Basin (roughly 83,500-ha area), and (C) 1987 HDOs for the southern Lake Tahoe Basin. The HDOs (for 1940, 1969, and 1987) were compiled photogrammically from aerial photography with varying scales, camera characteristics, image quality, and capture dates. The aerial photographs used to produce the HDOs are as follows: 1940 U.S. Forest Service (USFS) black-and-white photographs (CNL, July 1940), 1969 USGS black-and-white photographs (VCHH, August 1969), and 1987 USFS natural-color photographs (LTBMU FS 615170, July 1987). Precision-corrected Ikonos multispectral satellite imagery (1-m pixel size, acquired July 2002) was used as a substitute for HDO/DOQ for the latest imagery date, but these data are not available for download in this series due to restrictions in the licensing agreement. The resulting datasets have a 1-meter horizontal resolution. The projection of these data is set to UTM Zone 10, NAD 1983. The data for each of the three available dates are clipped into files that spatially approximate the 3.75-minute USGS quarter quadrangles (roughly 3,000 to 4,000 hectares), and have roughly 100 pixels (or 100 meters) of overlap to facilitate combining the files into larger regions without data gaps. The files are named after 3.75-minute USGS quarter quadrangles that cover the same general spatial extent. These files are available in the ERDAS Imagine (.img) format. Since 2001, researchers with the U.S. Geological Survey (USGS) Western Geographic Science Center, the USGS Water Resources Discipline, and the Desert Research Institute have recently completed an interdisciplinary research project in the Lake Tahoe Basin in California and Nevada designed to (1) map the current and historical state of the land surface, (2) analyze patterns, rates, and trends in urbanization and land-use change using GIS and other tools, and (3) assess the causes of land-use change and the possible implications to water quality and ecosystem health. This research focused on the southern Lake Tahoe Basin that includes the Upper Truckee River, Trout Creek, and Bijou Creek watersheds and three intervening areas that directly contribute to lake inflows (27,937-ha area). The USGS is currently in the process of expanding this research to the entire Lake Tahoe Basin (roughly 83,500-ha area). To complete this goal, historical digital orthophotos (HDOs) were produced by using multiple sources of archived aerial photographs (for the years 1940, 1969, and 1987). Precision-corrected Ikonos multispectral satellite imagery (1-m pixel size, acquired July 19, 2002) was used as a substitute for HDO/DOQ for the latest imagery date, but these data are not available for download in this series due to restrictions in the licensing agreement. The U.S. Geological Survey has also produced a map illustrating the land-use/land-cover change (manually interpeted from the HDO data) in the southern Lake Tahoe Basin for 1940, 1969, 1987, and 2002. This map is available for download at http://pubs.usgs.gov/sim/2007/2962/. 071940 081969 071987 ground condition Complete None planned ISO 19115 Topic Category ImageryBaseMapsEarthCover None historical aerial photography orthophotography None Lake Tahoe Basin California CA Nevada NV Hydrologic Unit Code HUC 16050101 None 071940 081969 071987 None Acknowledgement of the U.S. Geological Survey would be appreciated in products derived from these data. Christopher E. Soulard U.S. Geological Survey Physical Scientist mailing and physical address

345 Middlefield Road, MS-531

Menlo Park California 94025 United States (650)329-4317 csoulard@usgs.gov Christopher E. Soulard, Tom Coons, Robert Vitales, Rachel Sleeter, U.S. Geological Survey Zeiss ZI, BAE Systems SocetSet 5.2, Global Mapper 7, and ESRI ArcCatalog 9.2 Digital orthophoto quadrangles and quarter-quadrangles meet National Map Accuracy Standards. The accuracy of an HDO dataset is, at best, as accurate as that of the input data which the ortho control was derived. Additional inaccuracy can occur in the ortho-control-collection process because even slight measurement inaccuracies of the ground features selected for ortho control can affect the final accuracy. To fully assess the accuracy of the HDO, the accuracies of the DEM, DOQ, and/or map from which the HDO was rectified need to be considered. Final quality-assurance/quality-control measures consist of reviewing the DOPS program execution and accuracy reports, HDO/HDOQ header verification, on-screen visual inspection of each individual HDO/HDOQ, and a comprehensive check of all HDO/HDOQs within the project, including the examination of image-adjacency geometry and general image radiometry. Data are limited to areas included in the original aerial photography that, following orthorectification, met National Map Accuracy Standards. Digital orthophoto quadrangles and quarter-quadrangles meet horizontal National Map Accuracy Standards. N/A National Map Accuracy Standards requires that 90 percent of the well-defined points tested must fall within 40 feet (1/50 inch) at 1:24,000 scale and 33.3 feet (1/30 inch) at 1:12,000 scale. The DOQs and maps used to establish control meet National Map Accuracy Standards, and the resulting HDOs closely resemble the accuracy of these control data. Digital orthophoto quadrangles and quarter-quadrangles meet vertical National Map Accuracy Standards. N/A National Map Accuracy Standards requires that the vertical accuracy of the source DEM is equivalent or better than a level 1 DEM, with a root-mean-square error (RMSE) of no greater than 7.0 meters. The DEMs and maps used to establish vertical control meet National Map Accuracy Standards, and the resulting HDOs closely resemble the accuracy of these control data. U.S. Forest Service 071940 remote-sensing image CNL 20000 remote sensing image 071940 ground condition 1940 aerial photography U.S. Forest Service, Project CNL U.S. Geological Survey 081969 remote-sensing image VCHH 30000 remote sensing image 081969 ground condition 1969 aerial photography U.S. Geological Survey, Project VCHH U.S. Forest Service 071987 remote-sensing image LTBMU FS 615170 24000 remote sensing image 071987 ground condition 1987 aerial photography U.S. Forest Service, Project LTBMU FS 615170 Before initiating HDO production, issues revolving around the nature of the photography needed to be resolved to determine the most appropriate historical photography dates to use. These issues included: (1) the quality of the film diapositive or photographic print to be scanned; (2) the existence of any camera data, including information regarding the calibrated focal length, flight height, flight direction, scale, and calibrated fiducial measurements; and (3) the ability to acquire control data (for example, DOQs, USGS DEMs, and USGS 7.5-minute topographic-quadrangle maps) that coincided with each of the historical photograph dates. The most important consideration that can affect HDO production is the presence and distribution of well-defined photo-identifiable feature points shared by both the historical photographs and the data used for control (primarily DOQs). Changes in the landscape throughout time can affect the production team's ability to identify well-defined ground control points shared by both the historical photographs and the control data, and the lack of common points can make orthorectification largely inaccurate. Although more historical photography exists for the Lake Tahoe Basin beyond the 1940, 1969, and 1987 dates used in this study, these dates were determined to be the most appropriate based on the above criteria. Initial steps of HDO production included project planning, source-data research, and data acquisition. Data research and acquisition steps included obtaining existing standard USGS first-generation DOQs, USGS DEMs, published USGS 7.5-minute topographic-quadrangle maps, flightline diagrams, historical aerial photographs, and camera-calibration reports for the historical aerial photographs. The photograph centers and coverage extents were plotted on 7.5-minute topographic quadrangle maps for general reference and orientation to the data used for control. This photograph layout also helped determine the optimum image coverage for the 3.75-minute image mosaic. The number of single orthorectified images, or "chips," needed to make a HDOQ depends on the photographic scale and the location of photograph exposure. This varied considerably between each dataset, since the photographic scale (from 1:20,000 to 1:30,000) and flightlines varied between the three historical photography projects. After determining which historical photographs required rectification, the photographic paper prints or diapositives were digitally captured by using a high-precision image scanner with a scanning aperture of 28 microns, unless otherwise noted from the contributing agencies (for example, the California Tahoe Conservancy provided scanned imagery). Since camera data for historical photographs were generally incomplete or nonexistent, scanned images with clear fiducial marks were precisely measured for generating a camera calibration file for each date of imagery. The transformation software used to generate the camera-calibration file also calculated an approximate photograph scale. After transformation, all image coordinates had their origin at the image center. Existing standard DOQs were primarily (although not exclusively) used as control for the scanned historical images as a substitute for aerotriangulation, since historical flight control records were either insufficient or no longer available. The selected control points had to be visible on both the existing DOQ and the historical image. By displaying the two images on screen, ideally nine or more points of common, well-defined features in both images were selected and measured to provide ground control for each historical image. At least two points were selected on an image edge to ensure commonality with two or more adjoining overlap and side images. It is important to note that although a historical image can be controlled with a minimum of four points, the selection of more points ensures that the horizontal accuracy of the final HDO is more consistent with the control data. Occasionally, the visual and geographic differences between the historical image and the control DOQ were so great that correlation of common features for control points was difficult, if not impossible. This problem required using a control source closer to the date of the source historical photograph. Alternative control sources included historical and current topographic maps (hard-copy or digital) that were compiled using aerial photographs that are temporally closer to the date of the historical photographs being used to make the HDO. Camera-calibration files, control files, a DEM (which is generally the same elevation source as that used to produce the existing standard DOQ selected for control), and a scanned historical image are the inputs for the USGS-developed Digital Orthophoto Processing System (DOPS), which was used to perform orthorectification. A parameter (.par) file containing relevant project information was generated by the DOPS. Essential data describing each individual chip to orthorectify was recorded in the parameter file to ensure the proper interaction between inputs and for later program execution. Once rectified, the individual image chips required mosaicking to construct the 3.75-minute HDOQ tiles. Image chips were commonly used repeatedly where chip coverage extended into multiple 3.75-minute quadrangles. When data gaps were present in the mosaicked HDOQ as a result of underestimating the image-chip coverage, the original chip-domain size was adjusted to extend the image area and was again rectified and mosaicked. The mosaic software application locally adjusted the radiometric values of the images at join lines (appearing as seam lines in the software interface) to minimize tonal variations between adjacent images. After the image chips were mosaicked, the HDOQ overlap was calculated and trimmed, and the HDOQ corner crosses were embedded in the mosaicked image. Roughly 100 pixels (or 100 meters) of overlap were added to each 3.75-minute tile to facilitate combining the files into larger regions without data gaps (despite the appearance of black areas in many of the tiles, these areas will not prevent users from mosaicing the tiles into larger regions). Finally, a standard USGS keyword header was created for the HDOQ to include information such as the aerial-photo-source provider and acquisition date. Final quality-assurance/quality-control measures consisted of reviewing the DOPS program execution and accuracy reports, HDO/HDOQ header verification, on-screen visual inspection of each individual HDO/HDOQ, and a comprehensive check of all HDO/HDOQs within the project, including the examination of image-adjacency geometry and general image radiometry. 2002-2007 Christopher E. Soulard U.S. Geological Survey Physical Scientist mailing and physical address

345 Middlefield Road, MS-531

Menlo Park California 94025 United States (650)329-4317 csoulard@usgs.gov N/A meters North American Datum of 1983 Christopher E. Soulard U.S. Geological Survey Physical Scientist mailing address

345 Middlefield Road, MS-531

Menlo Park California 94025 United States (650)329-4317 csoulard@usgs.gov IMG WinZip http://pubs.usgs.gov/ds/376/ None Data are available online at no charge via Internet download. Acknowledgement of the U.S. Geological Survey would be appreciated in products derived from these data. None 20080910 Christopher E. Soulard U.S. Geological Survey Physical Scientist mailing and physical address

345 Middlefield Road, MS-531

Menlo Park California 94025 United States (650)329-4317 csoulard@usgs.gov FGDC Content Standards for Digital Geospatial Metadata None None http://pubs.usgs.gov/ds/376/ 2008092314152300{7716D523-D980-49B0-A981-0B723E6FBD4D}2008092313522200TRUE20080923file://\\IGSWMCWNWS01138\C\Documents and Settings\csoulard.WRUSGS\Desktop\tahoe_metadata_templateLocal Area Network