GENERATION OF COASTAL MARSH TOPOGRAPHY WITH RADAR AND GROUND-BASED MEASUREMENTS (REPRINTED FROM THE JOURNAL-OF-COASTAL-RESEARCH, VOL 13, PG1335-1341, 1996)

Citation
Ew. Ramsey et al., GENERATION OF COASTAL MARSH TOPOGRAPHY WITH RADAR AND GROUND-BASED MEASUREMENTS (REPRINTED FROM THE JOURNAL-OF-COASTAL-RESEARCH, VOL 13, PG1335-1341, 1996), Journal of coastal research, 14(3), 1998, pp. 1158-1164
Citations number
16
Categorie Soggetti
Environmental Sciences","Geosciences, Interdisciplinary","Marine & Freshwater Biology
Journal title
ISSN journal
07490208
Volume
14
Issue
3
Year of publication
1998
Pages
1158 - 1164
Database
ISI
SICI code
0749-0208(1998)14:3<1158:GOCMTW>2.0.ZU;2-P
Abstract
A topographic surface of a low lying coastal marsh was created by usin g three flood extent vectors digitized from ERS-1 SAR images and two e levation contours from U.S. Geological Survey topographic quadrangles. Point measurement of water depth at the times of the SAR collections allowed conversion of the radar measured flood extent vectors to topog raphic contours. Generation of the topographic surface was accomplishe d with a surface gridding algorithm, SAR and on-site measures. Errors in the generated topography were mainly associated with the lack of in put contours covering narrow to broad plateaus and topographic highs a nd lows. The misplacement of SAR derived flood extent vectors also cau sed errors in sparsely vegetated high marsh, at convoluted marsh-fores t boundaries, and at topographic depressions. Overall, the standard de viation of differences between measured and predicted elevations at 74 7 points was 19 cm. Excluding the above mentioned abrupt boundaries an d topographic highs and lows outside the range of available contours, standard deviation differences averaged about 14 cm, but most often av eraged about 8 cm. This suggested a 5 to 9 factor improvement over the 150 cm topographic resolution currently available for this area.