This Layout is showcasing a dasymetric model of 1959 cropland percentages in south-central Ohio counties, using four variables.

In Ohio in 1959, the pattern of cropland was largely in the Central and Northwest areas. The Southeast had much smaller percentages of cropland in each county. This was based on a number of variables including terrain and economy, predominantly. The Northwest area consisted of mostly level terrain with mixed cash crops as the main economic type. The Southwest corner had more hilly terrain and was used more for livestock ranching. These two variables can tell us where cropland is more likely to be.

This pattern of cropland is very closely associated with both the limiting and related variables. The two variables I talked about before, terrain and economy, are the related variables. In our cropland case, the related variables are combined to limit the amount of cropland. The terrain variable limits what kind of geography that crops can actually be grown on. There will be little to no areas in the “steep/broken terrain” class that have cropland. The economy variable doesn’t limit as strictly as the terrain variable, but still can show us where cropland is more likely to be. We should expect a large amount of cropland in a livestock and dairy dominated area. The limiting variables are our woodland and urban variables. These can also provide information about cropland density. Our urban variable is extremely telling because where there are large urban areas there will be no cropland. The woodland variable is similar in that where there are dense woodlands there will not be cropland. Our limiting variables dominate the pattern wherever there is an urban area. This is one of those variables that is binary so there is an urban area or there isn’t. This is a powerful variable because it fully blocks out sections of the map. I think the related variables dominated more in the area that showed more cropland. The two variables added up to show that that area would be the most likely for cropland which ended up lining up with our choro_crop. This solidified that the cropland would be in the northwest corner.

The Choro_crop layer shows the percentage of total area classes as cropland for each county. It is derived from the 1959 Census of Agriculture. This layer is a simpler technique that is best for comparing relative magnitudes across the surface of our map. The dasy_predict layer is much more complicated. It is built off of the choro_crop layer and includes limiting and related variables to make a dasymetric model. By building up this model, you get a much more specific and interesting model. By adding the limiting variables, we can see where there would be absolutely no cropland and by adding in our related variables we can see where cropland is most likely to be. This gives us a much more fully thought-out model instead of a single value for a whole county.