Virginia Carter Nancy B. Rybicki Justin T. Reel Henry A. Ruhl David W. Stewart
In addition, total biomass, total biomass minus periphyton biomass, and total LAI were calculated for each individual quadrat. The quadrats were grouped according to species composition and subsequently into density classes based on total biomass minus periphyton: sparse = 0-500 gdw/m 2, medium = 500-1000 gdw/m 2, dense = 1000-2000 gdw/m 2, and very dense = >2000 gdw/m 2.
A variety of remotely sensed products were available for developing the vegetation cover maps, including 1:12,000-scale color digital orthophoto quadrangles and the color infrared (IR) aerial photographs from which these were made, a 68-class 1993-94 Landsat vegetation cover classification map of southern Florida developed by the former National Biological Service and the University of Florida, and several vegetation maps of parts of the Taylor Slough model area (Rintz and Loope, 1978; Olmsted et al., 1980; Olmsted et al., 1981). In addition, we acquired a set of 1997 Landsat Thematic Mapper (TM) images that covered the model area.
A geographic information system graphical user interface (DBView), which was developed specifically to assimilate and interpret spatial data (Stewart,1997), was used to manipulate and recombine the 68 classes in the south Florida Landsat map into six vegetation cover classes plus water using the color IR photographs, digital orthophoto quadrangles, and vegetation maps for guidance.
Following a detailed examination of this vegetation map and correlation of the map with ground-truth information, a second vegetation cover map was developed using a January 1997 TM image. The Landsat TM instrument records both reflected (six bands) and emitted (thermal band) energy, respectively, for each ground area sampled. The ground spacing of reflected light measurements is nominally 30 m,while each thermal measurement represents an area 120 m on a side. Typically, the reflected and thermal data are processed separately. However, for this effort, the thermal data were oversampled to the 30-m resolution of the reflected bands. All data points within the resulting 7-band image were statistically grouped into 20 land-cover classes. The result of this process was then geometrically rectified to match the coordinate system used for all other field and remote sensing data collection. Using DBView, vegetation data collected in the field, and field observations, the 20 land-cover classes were further grouped into the seven vegetation classes and one water class.
Evaluation of these two vegetation cover maps required field trips to many sites within or on the periphery of Taylor Slough, including the area to the east that includes the C-111 canal area criss-crossed by drainage canals and the area to the west along the main park road to Flamingo and the Old Ingraham Highway. A special field reconnaissance was made to northern Florida Bay to check the map classes along the Buttonwood Embankment (Craighead, 1969) and the edges of the tidal embayments where Taylor Slough flows into Florida Bay. In addition, the GPS locations of the samples were plotted directly on the vegetation cover maps using DBView to identify the vegetation class from which each sample came.
U.S. Department of the Interior, U.S. Geological Survey, Center for
Coastal Geology
Comments and suggestions? Contact: Heather
Henkel - Webmaster
Generated by mp version 2.7.19 on Mon May 12 14:23:32 2003