How Much of the Area is "Close to the Edge?Well, that rather depends upon what you call "close". I suppose it is "close" whenever you want to use a half-plot near the border of the stand. The best way to sort this out is to ask the GIS people to put a boundary of width "W" around the inside of each of the forest polygons, and determine what that area is. That border also goes along "typed out" areas like roads or small water bodies in the polygon as well. When you compare that to the total area for the polygons you will know what percent of the forested land base is "close" to a stand edge. You might be surprised at how large that percent is. Make a guess before going further in this article or finding out for yourself, using whatever distance to the edge you consider "close". If the GIS people are too busy, there is a simpler request you could make, or perhaps you can get it yourself from the database. Ask them for the area of the forest polygons, and their perimeter. Now you have to calculate the area of the strip along the inside of the polygon. That is almost the length of the perimeter times the strip width "W" (make "W" a variable in your spreadsheet so you can test several widths). As the strip bends around, it "squeezes out" little triangles. By the time you get all the way around, these small triangles make a circle with radius "W". The final equation for the area of the border strip is therefore
This is slightly complicated by the fact that a polygon might have some narrow spots or be quite small, so make sure that the border area does not exceed the total area of small polygons. The GIS could do this calculation exactly, but this rough method will probably get you close enough to put the problem into perspective. Add up all the border areas, compare this to the total polygon areas, and you have what you need. We did this recently for a client, and found that almost 30% of the forest area was within 10 meters (32 feet) of a stand edge. Roughly, it was 1% of the area for every 1 foot of distance to the edge, although this dropped off starting at about 50%. It might be interesting to check for yourself. The percentage of area cannot be ignored, so that means you should be using a proper procedure to correct for "edge effect" - or at least have some proof that an incorrect procedure is not going off track. This simple calculation would be of interest to any inventory department. Doing it with several strip widths would be even more enlightening. If the GIS department cannot do it for you, do it yourself. A graph of your rough results might get them interested in providing a more precise answer. (Note : let us know if you do it both ways - we would be curious to find out the result for some additional databases - and let us know if it is OK to pass on your results in a future edition of the newsletter). |
Originally published October 2003