In my last post (part 1), I performed the tests using the 1970-2000 population data from NHGIS and comparing the temperature trend and trend of log population of nearby stations in the USHCNv2 data set.
One thing that has been nagging me is that while we might expect a population increase at a station to create a warming bias, I don’t expect a cooling bias of the same magnitude when we see a population decrease. This is because we’re really attemping to use population in this case as a proxy for development using “urban” materials, and when a population decreases we don’t expect to necessarily lose any of these urban surfaces. From the stations I used, about 35% of them actually had a population decrease over the period.
Thus, my first test was to replace the dP value of any station < 0 with 0. You can get the data (which is the same as last post) and the updated R file from here. For you R programmers out there, I’d love to know how to do this more efficiently than using my for loop.
One reason to be careful here is that this modification can greatly affect what we perceive to be the magnitude of the bias/effect. After all, by replacing negative dPs with 0 we are decreasing our differences in the nearby station comparison, and thus increasing our dT/dP slope. Furthermore, when it comes time to compute the population-adjusted temperature at each station to get our new average, we’ll be ignoring the cooling bias (if it exists) that might offset the warming bias.
Also, since some of the UHI can be attributed to waste heat, we might expect at least that portion of it to decrease with a decrease in population. After playing around with a couple different numbers, I settled on replacing dP with 1/4 * dP when dP is negative for a station. Here are the tabled results between unadjusted, replacing negative dP with 0, and replacing negative dP with dP/4:
As you can see, in the tests above the best correlation seems to consistently appear when we use the 1/4 replacement. However, it is somewhat troubling that the magnitude can be affected so much, with only slight differences in correlation. Anyhow, here are some graphs with the 1/4 replacement.