Laserfiche WebLink
<br />00849 <br /> <br />we have found that image resolution for the physical resource program was more a function of <br />required vertical accuracy than resolving powcr for geomorphic mapping. <br /> <br />Before the remote-sensing initiative. all topographic data were acquired by field survey <br />measurcments, which is not only expensive but also time consuming. which limited the area that <br />could be monitoled within any given year. Since 2000, various alternative airborne approaches <br />have been tested. the results of which are discussed below. As additional study sites were added. <br />or when more historical data were requircd to extend the period of observation. photogrammetry <br />and geomorphic mapping were applied to historical photographic data. The historic <br />photogrammetric analyses partly overlap timc periods of land-based topographic surveys to <br />provide ground truth data for photogrammetry. Before 2000, almost every annual image <br />acquisition consisted of analog film and prints, which degrade over time, were lost or misplaced, <br />and were extremely difficult to use by non-resident scientists. Thus, preservation of and access to <br />the historical image archive were two critical factors considered within the remote-sensing <br />initiative. An operational airborne approach for topography must satisfy the most stringent <br />vertical accuracy requirements within the GCMRC programs. This requirement resides within <br />the physical resource program and is 25 cm, which is based on the minimum amount of change in <br />sand bar height that is deemed "significant" and needs to be recorded (Schmidt et aI., \999). This <br />elevation accuracy value was our targeted objective with all the various topographic remote- <br />sensing evaluations that we performed during the past two years (Table 2). <br /> <br />Surficial geology of terraces and debris tlows are mapped into units based on elevation <br />(terrace level), hill slope, grain size, relative age, and composition. Historically, this infonnation <br />was extracted from (poorly rectified) aerial photographs using photointerpletation and field <br />investigalions. In general, dcbris flows ale monitored and mapped on an annual. system-wide <br />basis. whercas terrace deposits allong-tenn monitoring sites for fine-graincd sediment are <br />generally mapped once and monitored annually. Water resources personnel who monitor CRE <br />debris tlows have examined hyperspectral data to determine ifmass movement could be predicted <br />from the surficial geologic compositions provided by hyperspectral data. . A Ithough this is an <br />intriguing problem, the remote-sensing initiative focused on a more fundamental issue to <br />determine Ihe type of data that is most appropriate for detecting and mapping debris tlows. <br /> <br />One of the primary objectives of the remote-sensing initiative was to determine if there <br />were beller cost-effective, data-colleclion approaches for all the various parameters that are <br />monitored by the physical resource program so that scientists could be more productive and their <br />data would be at least as accurate as that obtained by past approaches. The most appropriate data <br />would provide the highest mapping capability. Overall, Ihere are lour remote-sensing issues that <br />needed to be resolved for this program and they are addressed in the following section. <br /> <br />.-- <br /> <br />4.2.1. Preservation of GCMRC Image Archive <br /> <br />An initial step in the remote-sensing initiative was an inventory of all image data that had <br />been collected by GCMRC. During this process, we discovered Ihat the film for one annual <br />image collection had been lost and could not bc localed. We also observed that the photographic <br />print collection was deteriorating due to age (discoloration) and use (markings, tears. wrinkles). <br />In addition. access to the archive required a physical presence into order to view, select and <br />duplicate necessary data. The initiative therefore strongly recommended that all future image <br />collections either be obtained with digital sensors or be immediatcly scanned to digital format. <br />The initiative also strongly recommended that the original photographic film be converted to <br />digital imagery and. in the case of lost film. that the prints be digitally scanned. Within the <br />initiative. Davis et al. (2002b) determined by a scries of scan tests that a scan resolution of 15 <br /> <br />23 <br />