Thursday, January 13, 2011


One of the variables that has a considerable impact on yields, both for arable and livestock, is the type of soil. It is clear why for crops such as wheat, but not so obvious for livestock. Don't they just eat grass? Well, yes, but grass won't grow everywhere, or perhaps not as luxuriantly as the livestock would prefer. Malcolm has identified some soil data, and on the map below I've layered the soil data with the 648 parishes. The next step is to add more soil layers (water, slope etc) and then extract the readings for each parish. Then build a model of yields using regression. We'll also add metereological data, and Mi is helping me with that.

Just from a glance at the map you can see that most of our parishes lie on fairly sandy soil (2) while those to the north-east lie on more clayey soil. Usually clayey soil is better for arable. It will be interesting to test this.

You'll see that there is a gap in the extreme south-west. That is the county of Cornwall, which I wasn't originally going to include. But it looks as though the thesis of 'market integration' is going to be something we'll run with, so Malcolm is working on that data now. That will bring our total number of observations to nearly eight hundred.

No comments: