|
snippet:
|
Lakebed maps were created offshore the Bayfield Peninsula, Wisconsin, Lake Superior to characterize habitats, substrates, and geoforms. These maps were created to support management decisions intended to address multiple littoral issues including increases in coastal erosion and sedimentation, degradation of native fish habitat, coastal development, water quality changes, harmful algal blooms, and species invasions. The intended users of these maps are lake managers and scientists working on restoring the Great Lakes. |
|
summary:
|
Lakebed maps were created offshore the Bayfield Peninsula, Wisconsin, Lake Superior to characterize habitats, substrates, and geoforms. These maps were created to support management decisions intended to address multiple littoral issues including increases in coastal erosion and sedimentation, degradation of native fish habitat, coastal development, water quality changes, harmful algal blooms, and species invasions. The intended users of these maps are lake managers and scientists working on restoring the Great Lakes. |
|
accessInformation:
|
US DOC; NOAA; NOS; National Centers for Coastal Ocean Science (NCCOS) |
|
thumbnail:
|
|
|
maxScale:
|
5000 |
|
typeKeywords:
|
[] |
|
description:
|
<DIV STYLE="text-align:Left;"><DIV><DIV><P><SPAN>Geomorphology, substrate and habitats were classified on the lake bottom using unique patterns observed in bathymetry and backscatter collected by a Teledyne-Reson 7125 multibeam echosounder mounted to the Research Vessel (R/V) Echo. The Coastal and Marine Ecological Classification Standard (CMECS; FGDC 2012) was used to organize and interpret lakebed bottom types. </SPAN></P><P><SPAN>Within the CMECS framework, the lakebed is classified into geoform levels 1 and 2, substrate units for origin, class and subclass and biological units for setting, class, subclass, group and community. Classification was performed on lakebed feature-based segments derived from automated partitioning of bathymetry, backscatter, and depth derivatives (e.g., slope, rugosity, and curvature) in the ENVI (v.4.7) image processing analysis software. The smallest segment area or minimum mapping unit was 100 square meters, defined by the scale of discernible surface patterns in remotely sensed data.</SPAN></P><P><SPAN>Substrate and biological units were classified according to predictions derived from random forest models trained on 127 annotated underwater lakebed videos distributed across the study area. A second independent set of annotated underwater videos were used to assess the accuracy of model predictions for substrate subclass and biotic community. Overall class accuracies for substrate subsclass and biotic community are 80% and 74%, respectively. When accuracy is calculated proportional to class area, the accuracies are 83% and 80%, respectively. </SPAN></P><P><SPAN>Geoforms were classified by interpretation of substrate prediction and remote sensing data by a mapping expert. No accuracy was quantified for geoforms since their scale is larger than the footprint of available in situ video annotations.</SPAN></P></DIV></DIV></DIV> |
|
licenseInfo:
|
|
|
catalogPath:
|
|
|
title:
|
NCCOS_BayfieldAPIS_LakebedMaps |
|
type:
|
|
|
url:
|
|
|
tags:
|
["National Centers for Coastal Ocean Science","NCCOS","Northwestern Michigan College","Office for Coastal Management","Great Lakes Restoration Initiative","GLRI","Ground Truthing","Accuracy Assessment","In Situ Data","Underwater Video","Benthic Habitat Mapping","Lake Superior","Apostle Islands National Lakeshore","Bayfield Peninsula","Wisconsin","Lakebed","Sediments","Substrate","Biotic","Geoform","CMECS","Random Forest","Predictions"] |
|
culture:
|
en-US |
|
name:
|
|
|
guid:
|
|
|
minScale:
|
150000000 |
|
spatialReference:
|
|