Science Service System
You are here : Home : Proposals_Summary

Summary of Proposal HYD0427

TitleWinter habitat for salmon, mapping ice cover and river channels from SAR data
Investigator Guritz, Rick - Geophysical Institute, University of Alaska Fairbanks, Alaska Satellite Facility
Team Member
Dr. Burnett, Kelly - U.S. Forest Service, Pacific Northwest Research Station
Mr. Vance-Borland, Ken - Oregon, Department
Dr. Miller, Daniel - Earth Systems Institute, Non-Profit
Dr. Zimmerman, Chris - United States Geological Survey (USGS), Alaska Science Center
Dr. Bernier, Monique - University Du Quebec, INRS - Centre Eau Terre Environnement
SummaryHabitat quantity and quality are key drivers of salmon abundance and distribution in freshwater. Advancing limited knowledge about how these drivers affect salmon in the Arctic-Yukon-Kuskokwim (AYK) system is the overarching goal of this project. However, current geospatial data on rivers are not sufficiently accurate and lack attributes for detailed mapping and modeling of salmon habitats over the vast AYK region. Chief among these attributes is the extent to which rivers freeze during winter, potentially limiting space for egg incubation and juvenile rearing. This project intends to examine the feasibility of using SAR data to map river networks and to detect areas of ice and water below ice in rivers. Specific objectives for the scientific use of the requested TerraSAR-X data are to: 1) identify ice classes (open water, ice over water, and frozen to the bottom) for the purpose of characterizing potential over-wintering salmon habitat in rivers; and 2) to delineate and compare high-resolution digital river networks obtained from TerraSAR-X and ALOS PALSAR fine-beam SAR data, by interpreting ALOS PRISM optical data, and by inferring flow paths, along with hydrogeomorphic characteristics (e.g., gradient, drainage area, and valley width), from digital elevation models (DEM) (Clarke et al. 2008) developed from satellite data. We will produce a monthly time series of SAR imagery for the Nome River from December through April using TerraSAR-X Strip mode multi-pole SAR (HH-VV at 3-m resolution). An evaluation of different products/modes will identify the best data source to meet the stated objectives. Variations of backscatter will be evaluated for ice classification. Field crews will validate the ice classification in a representative set of river reaches and collect control points with survey grade GPS. We will identify river and stream channels using threshold segmentation from coherence imagery. Topographic features of the area are subtle; therefore, we think that TerraSAR-X Strip mode SAR data at 3-m resolution is needed to construct a detailed and accurate river network. Hydrogeomorphic characteristics (e.g., channel gradient, valley width) will be modeled when river networks are derived from DEMs. This project will produce new, higher-resolution digital maps of the Nome River network and document the value of TerraSAR-X data for mapping of ice classes. We will deliver at least two peer-reviewed publications describing: 1) the feasibility of mapping ice classes from TerraSAR-X and PALSAR backscatter analysis; and 2) tradeoffs in accuracy and cost when using available river networks, river networks delineated directly from remote sensing data, and rivers derived from DEMs. Funding ($473,073) was secured from the Arctic-Yukon-Kuskokwim Sustainable Salmon Initiative (AYK-SSI) for a project entitled “Landscape predictors of coho salmon.” We originally intended to use RADARSAT-1 data and seek additional funds for TerraSAR-X data. Subsequent to the award of the AYK-SSI proposal, RADARSAT-1 data become unavailable and we submitted a proposal to the Pacific Coastal Salmon Recovery Fund to purchase TerraSAR-X data. Funding has been tentatively approved and is contingent on acceptance of the current proposal for scientific use of the TerraSAR-X data.

Back to list of proposals

© DLR 2004-2016