An analysis of farm ground in the Imperial Valley lost to 1, 2, and 3 mile buffers from feed lots. Made with QGIS overlap analysis and MS Excel.

Acres of Farm Ground: 453,489
Acres Lost to 1-mile rule: 72,784 (16%)
Acres Lost to 2-mile rule: 181,255 (40%)
Acres Lost to 3-mile rule: 271,450 (60%)

Currently, the LGMA rule is that leafy green vegetables can only be grown on fields that are more than 1 mile from a feed lot due to the risk of E. Coli. I wanted to know how much of the valley’s produce ground would be unable to grow leafy greens if the rule is increased to 2 or 3 miles.

I located the IID’s REST server online and pulled the data into QGIS. Their REST server has a feature layer that has a polygon for every field in the Imperial Valley which include the canal & gate name, acres, etc. for each polygon.

I had an old KML file that contained the locations of every feed lot in the county, along with a 1, 2, and 3 mile radius around each one, but it measured the radius from the center, which isn’t useful since food safety departments will measure from the closest edge. I imported the KML file into QGIS and drew a polygon around each feed lot.

Then, using QGIS’ buffer tool I was able to generate the buffer zones around each feed lot using the actual edges of the polygons I drew, rather than from a center point in the feed lot.

With all those layers in place, I was able to use the QGIS overlap analysis tool to analyze the overlap between the field boundary polygons and the feed lot buffers. This generated a CSV file which gave me a percentage value for the overlap of each individual polygon.

In Excel, I opened the CSV and used some simple formulas to apply the percentage values to the acreage of each polygon and determine how many acres of each field are within 1, 2, or 3 miles of a feed lot and how many acres of each field were beyond (or safe from) each buffer zone.

This is only a rough estimate and the reality is slightly worse because some shippers won’t allow you to split a field if part of it is within the buffer zone. And the split wouldn’t be exactly along the buffer, generally it is a line drawn along the mid point or other point in the field, but this gives a general idea of the effect each rule would have. I should change the formula to assume that if 50% or more of a field’s area is within a buffer then all of it is not safe.

Here is a table of each field and it’s values. Unfortunately the REST server returned IID’s unique IDs and not the more common full name. For instance, “DAH_66B” rather than “Dahlia 66B”. I think later, I will work on a lookup table that fixes this so I can use it on future projects pulled from IID’s data.