Monitoring Winter Cover Crops with Satellite Imagery

Project: Promoting NASA Satellite Data Products for the Reduction of Agricultural Nutrient Loading in the Chesapeake Bay via Winter Cover Crop Prioritization Mapping

Software: ENVI

For this project, I worked in a team with 4 other interns and 3 NASA science advisors to study winter cover crops in the Chesapeake Bay using satellite imagery as a GIS intern with the DEVELOP program at NASA Goddard during 2009-2010.

One cause of the decline Chesapeake Bay water quality is nutrient loading, excess nutrients that cause eutrophication from low dissolved oxygen. At the time of the project, the Chesapeake Bay Watershed modeling system included hydrologic datasets to study run off in the bay, but lacked land cover datasets that could monitor nutrient run off from farming.

Cover crops have been considered a best management practice for farmers to assist with reducing nutrient runoff. However, there was a need for remotely monitoring temporal and spatial change in the planting of crops over time across the whole Chesapeake Bay Watershed.

Leaf Off Landsat Imagery Mosaic of Chesapeake Bay Watershed

To address this issue, I assisted my team with creating leaf-on and leaf-off mosaics of the Chesapeake Bay using Landsat 4-5 and Landsat 7 imagery to track winter cover crops.

First, I downloaded Landsat 4-5 TM and Landsat 7 SLC-off images from the online USGS Visualization application (GloVis) that covered the Chesapeake Bay Watershed for November 2008-April 2009 (leaf-off) and May 2009-September 2009 (leaf-on). After putting the imagery through the Large Area Scene Selection Interface Tool (LASSI) to optimize scene selection and LEDAPS for atmospheric correction, I created a mosaic of the Chesapeake Bay Watershed in ENVI. Next, I found the NDVI of the scenes to find the green areas with healthy vegetation. Lastly, the National Land Cover Dataset 2006 (NLCD) was clipped to the pasture/hay and cultivated crops land cover types and then overlaid on top of the NDVI mosaics to identify agricultural areas for crop classification during the leaf off season.

Leaf On Landsat Imagery Mosaic of Chesapeake Bay Watershed

There was a visual noticeable difference between the Landsat leaf off and leaf on mosaics due to the vegetation seasonality. Though the imagery was processed by LASSI, the quality of some of the images was limited by cloud and snow cover especially during the leaf off period.



NDVI overlaid by National Cropland Data Set in Eastern Shore, MD

However, the leaf off Chesapeake Bay Watershed mosaic offered improved detection of winter crop plantings on farms that were likely to have high nutrient levels following fall harvest. Green coloration on the mosaic and NDVI values greater than 0.35 indicated the presence of healthy vegetation during the leaf off season. By combining the NLCD dataset with the leaf off mosaic, a land cover map was created to show the possible locations of winter cover crops.

Cropland vector data intersection with NDVI > 0.35