A GIS Approach to Buffer Strip Analysis

A GIS Approach to Buffer Strip Analysis

Abstract

Issues related to the use of chemicals in the farming business are of growing concern. One key area of focus deals with the potential for applied chemicals to drift from the intended use area to other areas where their presence may be undesirable and possibly hazardous. Numerous interest groups field concerns related to the various aspects of this issue. The US Environmental Protection Agency (EPA) may have the last word concerning spray drift. The idea of no spray zones is being considered in high-risk areas. Included in these high-risk areas are aquatic environments. These areas may be sources of drinking water and sensitive ecosystems. An analysis of theoretical buffer strips was examined for an area in eastern Kansas related to this issue. Findings show that policy imposed by the EPA may have significant impact to farming practices and economic viability of growers in this area. Findings also show that some of this policy may be mitigated by the presence of forested areas which could serve as a natural barrier to spray drift. For more detailed findings concerning buffer composition, additional analysis will need to be conducted.

Contents

Introduction

The image used in this project was extracted from a Landsat 5 Thematic Mapper (TM) scene taken over northeast Kansas on May 23, 1994. Ground resolution for this image is 30 meters. The following image shows a false color composite of the study area utilizing TM bands 3, 4, and 5 seen in BGR respectively. This area covers Clinton Reservoir along with numerous other small lakes and ponds. In this composition vegetation appears in bright green while some of the agricultural fields are displayed in a pink color. There is also a considerable amount of grassland in this scene displayed with a combination of the green and pink colors.



Landsat 5 TM Band Descriptions
(Jensen 2000)
BandWavelength (µm)Spectral Region
10.45 - 0.52Visible Blue
20.52 - 0.60Visible Green
30.63 - 0.69Visible Red
40.76 - 0.90Reflective Infrared
51.55 - 1.75Mid-Infrared
610.40 - 12.50Thermal Infrared
72.08 - 2.35Mid-Infrared


This specific study scene was selected for several reasons. First, it contains a relatively large amount of water and related shoreline. There are several small community and state lakes in the image as well as many small farm ponds. Of greater significance is the presence of Clinton Reservoir. This area also contains a good deal of tilled farm ground as well as a considerable amount of forested areas. This scene represents nicely what type of land cover can be found in eastern Kansas. Farming is one of the biggest industries in this area with growers producing a variety of cool and warm season crops. To more efficiently grow these crops, pesticides, herbicides, and fungicides are often applied to the cropped areas. The direct result of these chemical applications is higher crop yields. Chemical labels provide the user with specific instructions and policy associated with when and how they can be applied.

Numerous reports have been filed linking pesticide spray drift to hazardous effects such as health concerns, fish kills, and non-target plant damage. Due to these concerns, the EPA Office of Pesticide Programs is considering imposing policy that would require chemical applicators discontinue applications in certain circumstances in an effort to minimize the potential harmful effects of spray drift. By examining theoretical buffer zone requirements using a GIS approach, visual and quantifiable results can be analyzed. This project serves to provide practical data to growers with theoretical conditions that may be imposed by the EPA. These conditions include buffer strips (no spray zones) around all water bodies, flowing or standing. Theoretical buffer strips of 30, 100, 250, and 500 meters will be analyzed. The following list includes questions that will be addressed in this report.

  1. What percentage of the area is composed of the above mentioned land cover classes?
  2. What percentage of crop ground falls within the proposed buffer distances to aquatic environments?
  3. What percentage of buffer strips is composed of forested areas?

Image Processing

Top


Idrisi image processing and GIS software was used to classify and analyze the data. Various techniques were used to classify this image into the land use classes of interest. These classes are: Water, Row Crop, and Forest. From this image it can be seen that there is a considerable amount of agricultural ground in close proximity to the various water bodies. This feature is of particular importance and will be discussed later.




Boolean image of water in the study area. 1150 individual water bodies were identified in this image.








Boolean image of row crop in the study area.








Boolean image of forest in the study area.







Image Composition of Relevant Land Classes
Land Class Area (ha)Percent of Total Scene
Water39974.1
Row Crop2314324.0
Forest1644717.0
Other5298754.9
Total96574-

Creating the Buffer Area

Top

Boolean images of the various buffer strips (30, 100, 250, and 500 meters) were also created. An example of the transition is shown below using the buffer strip of 250 meters as an example.




The image on the right above shows potential no spray zones that coincide with ground currently used for row crop production. This overlap is of particular interest since it is these areas that would be most significantly impacted if a buffer strip requirement were to come into effect. Policy may dictate that these areas no longer would be allowed to receive treatment of chemicals. The effect that this would have can only be negative for crop yield. This may significantly effect crop yeild and farming practices.



This image shows all aspects of the overlay process. The areas of row crop that overlap with the buffer areas are seen in purple. This image is a close up of the inlet to Clinton Reservoir.







Image Analysis

Top

From the processed imagery, details associated with buffer strips can be quantified. As mentioned earlier, several questions were to be addressed. The following table outlines what was discovered concerning questions 2 and 3 with respect of composition and impact of buffer strips.

Land Cover Statistics
Buffer Width (m)Total Buffer Area (ha)Crop-Buffer Overlap (ha)Impacted Cropped Area
(Percent of Total)
Area of Buffer Forested (ha)Percent of Buffer Forested
30107160.03353.3
10054253291.495617.6
25023752330914.3442518.6
500598321269454.91054517.6



Summary

Top

Farming is indeed a large industry in eastern Kansas. In this analysis alone, row crop represents nearly a quarter of the land use. In the current day of low crop prices and farm consolidation, farming efficiency is of extreme importance. Potential disruption in current farming practices may send even more farmers out of business.

From this data it can be seen that a significant portion of row crop areas will be impacted with no spray zones of 250 and 500 meters from aquatic environments. It is also of interest the relatively high percentage of forested areas falling within the buffer zones. This statistic represents only the fact that the no spray zone, whether they are overlapping crop ground or not, do contain significant portions of forest. This characteristic may be analyzed further to determine if the forested areas may act as natural barriers of potential spray drift to aquatic environments. Additional questions that may be addressed concenring this issue follow.

  1. What is the composition (roads, vegetation type, etc...) of the buffer zones?
  2. What is the width of the naturally occurring vegetation and other features occupying the buffer zones?
  3. What is the exposure potential for each individual direction of spray drift?


References