{ "culture": "en-US", "name": "", "guid": "", "catalogPath": "", "snippet": "This dataset was created as part of a land-cover mapping project for Vermont using object-based image analyis techniques in conjunction with high-resolution remote-sensing data, including LiDAR and multispectral imagery.", "description": "

This dataset depicts wetlands in Vermont. It was created with an automated feature-extraction process that relied on high-resolution LiDAR and multispectral imagery. In particular, a Compound Topographic Index (CTI) layer derived from LiDAR was used to identify landscape features that have suitable topography and flow potential for wetlands. Moderate-scale (10 m) statistical models developed by Patrick Raney of Ducks Unlimited were also used during classification in a data-fusion approach that maximized the value of individual inputs. After initial identification, mapped features were assigned to one of three primary wetlands classes: Emergent, Scrub\\Shrub, and Forested. The class assignments were based primarily on vegetation height (as estimated from LiDAR) and spectral characteristics (e.g., features with short vegetation that appeared very bright in leaf-off multispectral imagery were assigned to the Emergent category). The initial map was then generalized to eliminate unnecessary detail using a minimum mapping unit of 0.1 acres. In a final step, the automated output was manually reviewed against multispectral imagery and obvious errors of commission and omission were corrected. More than 57,000 manual corrections were incorporated into the final layer. Overall, the combination of automated feature extraction and manual corrections was biased toward over-prediction, focusing on capture of borderline features whose functional status cannot be definitvely established with remote-sensing data alone (i.e., it is generally easier to discount false wetland features than it is to locate omitted ones). Known areas of overestimation include managed forestlands with extensively-modified drainage patterns and wide river and stream channels. This map is considered current as of 2016, the year of the most-recent multispectral imagery used during manual review.<\/SPAN><\/P><\/DIV><\/DIV><\/DIV>", "summary": "This dataset was created as part of a land-cover mapping project for Vermont using object-based image analyis techniques in conjunction with high-resolution remote-sensing data, including LiDAR and multispectral imagery.", "title": "Wetland_subshed_Uni_Diss", "tags": [ "airborne LiDAR", "environment", "eCognition", "CTI", "GIS", "Compound Topographic Index", "Vermont", "automated feature extraction", "wetlands" ], "type": "", "typeKeywords": [], "thumbnail": "", "url": "", "minScale": 150000000, "maxScale": 5000, "spatialReference": "", "accessInformation": "Developed by the University of Vermont Spatial Analysis Laboratory with funding from: State of Vermont Clean Water Fund, Vermont Agency of Natural Resources, Vermont AGency of Transportation, Lake Champlain Basin Program, and the Vermont Center for Geographic Information.", "licenseInfo": "

Spatial Analysis Lab, Vermont Center for Geographic Information, and the State of VT make no representations of any kind, including but not limited to the warranties of merchantability or fitness for a particular use, nor are any such warranties to be implied with respect to the data. There are no access constraints<\/SPAN><\/P><\/DIV><\/DIV><\/DIV>" }