Laserfiche WebLink
<br />00844 <br /> <br />environments. such as return-current channels, shoreline embayments, and Ihe mouths of <br />tributaries. The presence and number of backwaters within or near return-current channels. <br />shoreline embayments, and tributary moulhs are monitored during quarterly, system-wide <br />vegetation surveys. which map the presence of dry and wet marshes that contain different wetland <br />vegelation. The temperatures within the backwater areas are monitored periodically by <br />lhemlistor-string lield surveys al selected backwater areas and by water gages at several sites <br />within Ihe CRE. By their nature, the surveys and gages sample a small portion of the CRE. <br />Detection and mapping oflhese warm backwaters should be more easily accomplished using <br />appropriate remote-sensing data. We investigated this possible remole,sensing application by <br />collecling a multispectral (12-band) data set containing two TIR bands for a 44,mile segment of <br />the CRE between river miles 30 and 74 (coincident with a ground survey in the area). This 44- <br />mile airborne survey lOok 20 minutes to complete using a helieopter flying at 365 m AGL (above <br />ground level). Even at Ihis low AGL, the multispectral sensor could provide only I-m spalial <br />resolution. TIR image data record surface radiant temperature across the entire channel, which <br />can augment site-specilic thermistor surveys. Our TIR data were collected in early July 2000 <br />when the ambient air temperature was so high (38l<!:) that the TIR cooling system could not reach <br />the required absolute zero degrees (for a 0.1 degree sensitivity), which resulted in a 0.3 degree <br />sensitivity in the TI R image data, but even this sensitivity proved more than adequate to detect <br />and map the warm backwater areas (Davis, 2002b). Derivation of water radianl temperature <br />required calibration of the airborne sensor signals to water temperatures, which was accomplished <br />using coincident water-gage temperature data. The airborne TIR data showed a linear relation <br />with main-stem water temperature (Figures 7 and 8) and easily mapped all warm backwater areas <br />within the 44-mile river segment, which included eddies with rather steep thermal gradients, <br />tributary mouths with gradual and abrupt thermal gradients. eddies formed by reattachment bars, <br />return-current channels, and isolaled backwaters that are diflicult to discern in visible-wavelength <br />imagery (Figures 9 and 10). Although the I-m spatial resolution (and possibly the 0.3 degree <br />sensitivity) of the TIR imagery was insufficient for the detection of archaeological structures. il <br />was found to be totally adequate for detection and mapping the warm-water areas within the 44, <br />mile study area. Unfortunately, the 44,mile study area covered by the multispectral data only , <br />included wet marches. Therefore, we could not test the capability of these data for discriminating <br />and mapping dry and wet marsh areas based solely on the I-m reflectance band data. <br /> <br />At the present time, the limiling factor for multispectral sensors is their spatial resolution, <br />which is not bener than 50 cm. Additional protocols of the physical resource program, and <br />possibly of the biologic resource program, require a higher spatial resolution (S 20 cm). These <br />additional requirements are discussed in following sections. Withoul any additional applications <br />of multispectral data at resolutions near 50-100 cm, the cost ($530 per river km) for such data for <br />just mapping warm backwaters would be difficult to justifY, However, there may be additional <br />needs for this lower resolution, multispectral image data in order to more accurately and <br />efliciently inventory Ihe terrestrial vegetation reSources within the eRE. This issue is addressed <br />in the following section, <br /> <br />3.2 Terrestrial Envirollmell/ <br /> <br />The biological resource program periodically monitors the composition, area, and volume <br />of vegetation habitats within the CRE each year. Composition is monitored 10 determine changes <br />in plant populations due to the invasion of exotic (non,native) species and 10 dam flow operations <br />during the year. Area and volume are primary indicalors of the suitabilily of vegetal ion slands as <br />faunal habitats. Monitoring of these characteristics is performed mostly by ground surveys. <br /> <br />13 <br />