Laserfiche WebLink
<br />one taken near the upper end of the class range would be small enough to <br />be acceptable to the user. The problem is thus to select class ranges for <br />all terrain factor variables in such a way as to minimize the effect of <br />stratification. This ideal must, of course, be tempered by practicality; <br />class intervals too small to be identified by any practical interpretation <br />or mapping process must be avoided. In this case, because the mathematical <br />models relating the gap crossing system to the environment had not been <br />formulated, the basic or primary data were recorded in actual values; that <br />is, measurements of factors were not in classes but rather in absolute <br />values. The overriding reason is that data collected and stored in this way <br />can be classified in a number of different ways and are thus useful for a <br />wide variety of models or other purposes, whereas data collected in terms <br />of classes can be used only for the one purpose for which the classes were <br />designed. Because the mathematical models of the various systems were not <br />available, considerable judgment had to be used in the selection of the <br />class intervals for the various factors when they were actually mapped. <br />The class ranges for the significant factors are identified in Table 2. <br /> <br />Acquisition of basic data <br /> <br />10. The numerical values for the factors listed in Table 1 were <br />obtained by photogrammetry, photo interpretation, literature search, and <br />by field measurements in the study area (fig. 1). Photogrammetry was used <br />where possible to obtain measurements of water width, bank height, bank <br />angle, gap widt~, number of trees per unit area, and vegetation band width. <br />Often, however, the geometric properties of the gaps could not be measured <br />because of obscuration by vegetation and photographic scale limitations. <br />In such instances, the descriptive data were estimated using photo inter- <br />pretation techniques. The reliability of photo interpretation is directly <br />proportional to the quantity and excellence of "ground truth" data, which <br />are the links between ground conditions and the photo image. Accordingly, <br />every effort was made to obtain abundant ground truth data from both liter- <br />ature sources and field measurements. <br /> <br />11. Published literature on the required factor data was found to be <br />almost nonexistent, so arrangements were made for a German-speaking WES <br />employee to visit the various German agencies dealing with hydrographic <br />studies to obtain existing long-term hydrographic and engineer records. <br />The records were very valuable to the study, but they had significant gaps <br />and constraints. For example1 most sites were descriptions of locations <br />chosen specifically for gaging stations, and thus tended to be in straight <br />reaches where the channel was artificially constrained and geometrically <br />regular, or at bridges or other works which produce anomalous but regular <br />channel configurations. Thus, the data on water level fluctuations at <br />gaging stations could not, for example, be directly applied to the natural <br />and irregular reaches of the channel. The records also lacked soil strength <br />and vegetation data. <br /> <br />16 <br />