
As of October 2005, the USDA Natural Resources Conservation Service (NRCS) has entered into a five year Memorandum of Understanding with the Geographic Information Science & Applied Geography (GISAG) Research Center of the Department of Geography and Planning at the University of Toledo, Ohio.
Crop
Type Identification by Use of Remote Sensing
Dr. Kevin Czajkowski, Dr. Patrick Lawrence, Jim Coss, Phil Haney, Katie Swartz, and Rumiko Hayase
Work performed will assist NRCS in implementing the Maumee
Watershed project, including sub watershed rapid resource assessments,
watershed and area planning, on farm conservation planning and delivery of
conservation technical assistance and conservation cost-share programs
authorized by the 2002 Farm Bill that are of mutual interest to University of
Toledo and NRCS.
The tasks will generally consist of:
1) Annually determining land cover and crop rotations via remote sensing techniques.
2) Combining Ohio, Indiana, and Michigan data layers to establish Maumee Watershed Project Area GIS data layers for the project.
3) Establish and maintain a Maumee Watershed Project GIS Website to provide educational and informational outreach to share the data and information with other project partners, resource managers, and the general public.
GIS Data layers to include (but not limited to):
DEM (Digital Elevation Model)
SSURGO Soils
Stream Network
Land Use Cover
Watersheds
Quaternary and Bedrock Geology
Wetlands
Source Water Protection Areas
Groundwater data
FEMA Floodzone data and 100 Year floodplains
Climate zones
Soil drainage
Roads and Transportation
Pipelines and Transmission Lines
Recreational Areas and Parks
Landfill and Dumps
Various boundaries (e.g. states, cities and
counties)
Weather and Climate Stations
Stream Hydrologic Stations
Census Information
Historical Sites and Cultural Resources
Crop type classification for the Maumee River project is being carried out using multitemporal Landsat 5 satellite imagery. Images from several times during the growing season will allow us to differentiate between the different crop types, in particular corn, soybeans, wheat and pasture.

Once collected, the images undergo cloud screening and then are stacked in Erdas IMAGINE remote sensing software package. Training sets of crop type have been collected using a driving survey of the watershed and located with Global Positioning System (GPS) readings. These training sets are used to create spectral signatures in Erdas and then a supervised classification is performed using the Maximum Likelihood classifier.

Validation of the classification will be performed using separate points collected during the driving survey.
|
|
|
|
|
Last revised: May 3, 2006