Sign up to receive email updates, action alerts & health tips from EWG. [Privacy]

Stream Banks

Failure to maintain buffer zones worsens farm pollution

Read more below

Appendix: Methodology

Broken Stream Banks: Appendix: Methodology

April 25, 2014

The methods used to execute this assessment had three components:

  • Creating an accurate representation of waterways adjacent to agricultural land.
  • Establishing which waterways were identified as perennial in the Minnesota Public Waters Inventory GIS layer.27
  • Quantifying the extent to which waterways adjacent to agricultural land maintained the required 50-foot vegetative buffer.

Detecting Agricultural Shorelines

The infrared and red bands were manipulated to highlight hydrographic features in the high-resolution photography obtained from MDNR. This enabled the construction of an accurate picture of standing shorelines in southern Minnesota. The shorelines along surface water were used to identify those stream segments that were within 200 feet of land tracts classified as agricultural in the Farm Service Agency’s Common Land Unit shape file.28

Classifying Public Water

Once a segment of a river or stream intersecting agricultural land was identified, it was classified as “public” and “perennial” based on the Minnesota Public Waters Inventory GIS layer for all overlapping areas. Waterways that were not classified as public and perennial were ignored. The remaining rivers and streams were then named using the Minnesota Public Waters Inventory and the USGS’ National Hydrography Dataset. If no name existed it was logged as unnamed and given a unique ID number. 29

Assessing Agricultural Buffers

Buffers of 25, 50 and 75 feet were mapped along all the perennial waterways. Only buffers associated with agricultural land, based on the agricultural common land units, were used to establish the universe of required buffers. All buffer areas were then analyzed individually for the presence of vegetation using the Normalized Difference Vegetation Index, a product developed from the aerial photography.

The acreage within 50 feet of the waterway was then assessed to determine how much of the required buffer acres were actually maintained with the required permanent vegetation. The stream segment, entire waterway and watersheds were given a grade based on the percent of required permanent vegetation present. The acreage summary statistics represent the area within the 50-foot required buffer zone that lacks permanent vegetation.

The length of stream segments that were missing part or all of the required buffer acres was also determined in order to assess the amount of unbuffered shoreline (For example, one mile of stream could potentially have two miles of unbuffered shoreline along both banks.)

The areas missing part or all of the required buffer acres were intersected with trout streams, highly erodible land tracts, impaired waterways and watersheds in order to analyze the data from numerous angles.

We manually checked every mile of perennial rivers and streams for errors in our analysis by comparing the 2011 remotely sensed GIS data outputs with 2012 or 2013 imagery. If we found that buffers that were listed as missing in 2011 were present in 2012 or 2013, we lowered the estimate of missing buffer acreage accordingly. If we found, however, that buffers present in 2011 were missing in 2012 or 2013, we did not increase our estimates of missing buffer acres. This means our results likely underestimate the extent to which cropland is currently encroaching into the required buffers.