Details:

Journal Title:
Remote Sensing of Environment
Personal Author:
Enwright, N.M. ; Cheney, W.C. ; Evans, K.O. ; Thurman, H.R. ; Woodrey, M.S. ; ... More +
Enwright, N.M. ; Cheney, W.C. ; Evans, K.O. ; Thurman, H.R. ; Woodrey, M.S. ; Fournier, A.M.V. ; Gesch, D.B. ; Pitchford, J.L. ; Stoker, J.M. ; Medeiros, S.C. Less -
NOAA Program & Office:
NOS (National Ocean Service) ; NCCOS (National Centers for Coastal Ocean Science)
NOS (National Ocean Service) ; NCCOS (National Centers for Coastal Ocean Science) Less -
Description:
Irregularly flooded wetlands are found above the mean high water tidal datum and are exposed to tides and saltwater less frequently than daily. These wetlands provide important ecosystem services, such as providing habitat for fish and wildlife, enhancing water quality, ameliorating flooding impacts, supporting coastal food webs, and protecting upslope areas from erosion. Mapping irregularly flooded wetlands is challenging given their expansive coverage and dynamic nature. Furthermore, coastal wetlands are expected to change over the coming century due to sea-level rise and changes in the frequency and intensity of extreme storms. Consequently, coastal managers need baseline information on the spatial distribution of wetlands along with efficient and repeatable methods for observing changes. In this study, we used coastal wetlands from existing land use land cover data, best available lidar-derived digital elevation models, and Monte Carlo simulations to incorporate elevation uncertainty to create a probabilistic map of irregularly flooded wetlands along the northern Gulf of Mexico coast (USA). Our approach integrated findings from a review of coastal wetland elevation error in lidar datasets and an analysis of spatial autocorrelations of wetland elevation. We found a positive correlation (r = 0.563, p < 0.0001) when comparing the probability estimated from a digital elevation model and in situ elevation observations. The differences in probability had a mean bias error of −0.04 (i.e., digital elevation model-based probability tends to be slightly lower), a mean absolute error of 0.20, and a root mean square error of 0.26. Beyond this overall validation, we explored error metrics for land cover classes and lidar collection details. To quantify areal coverage of the probabilistic output, we classified the probability values into equal bins using an interval of 0.33. The areal coverage of the lowest probability bin (“unlikely”; probability ≤0.33) was separated into the upper and lower portions of the irregularly flooded wetland zone. Of the coastal wetlands along the northern Gulf of Mexico coast about 38% were classified as unlikely and low with the greatest coverage in south Louisiana and the Everglades and around 33% were classified as unlikely and high with the greatest coverage in the Everglades and Texas. The relative coverage within the highest probability bin (“likely”; probability >0.66) covered around 13%, with the greatest coverage in south Florida, south Louisiana, and Texas. The framework developed in this study can be transferred to other coastal wetland areas and updated to observe changes with sea-level rise.
Keywords:
[+]
Coasts Monte Carlo Method Wetlands Coastal Wetlands Elevation Uncertainty Lidar Monte Carlo Simulations Spatial Autocorrelation
Source:
Remote Sensing of Environment 287 (2023) 113451
DOI:
https://doi.org/10.1016/j.rse.2023.113451
Format:
PDF
Document Type:
Journal Article
Funding:
Grant no. NA19NOS4510195
Place as Subject:
Mexico, Gulf of
Rights Information:
Accepted Manuscript
Compliance:
Submitted
Main Document Checksum:
[+]
urn:sha256:d9baeae66a1a23a679d75509e8e2b3a3224f1524a34bc814e5ec00980cce3baa
Download URL:
https://repository.library.noaa.gov/view/noaa/56092/noaa_56092_DS1.pdf
File Type:

Supporting Files

• No Additional Files

More +