Here’s something interesting: I’ve collated the GISS raw(dset=1) and GISS adjusted (dset=2) versions and then calculated the range of adjustments. The largest positive adjustment was over 8 deg C and the largest negative adjustment is greater than -6 deg C. I separated out the stations that had no adjustments (max adjustments under 0.01 deg C either way) and plotted their locations in the first figure. I then plotted figures showing stations with adjustments. (Nearly 40% of the 5990 stations with adjustments had zero adjustment.)
Here’s a plot of the stations with zero GISS adjustments. Notice the difference in pattern in North America between the stations with zero NASA adjustment and the stations with non-zero NASA adjustment (for now, I’ve not examined whether the difference between only positive, only negative and two-sided adjustments is relevant.) There’s a sharp distinction at the Canadian-US border with Canadian data being almost entirely NASA unadjusted, while US data is all adjusted to some degree. This has nothing specifically to do with USHCN adjustments as these have all been done upstream of the NASA adjustments.
Hansen et al 1999 describes their adjustment as follows:
An adjusted urban record is defined only if there are at least three rural neighbors for at least two thirds of the period being adjusted. All rural stations within 1000 km are used to calculate the adjustment, with a weight that decreases linearly to zero at distance 1000 km.
I must say that I find it hard to picture an implementation of this adjustment that yields the different patterns in the figures below.
Maybe Hansen’s adjustments have something to do with border security and you have to clear customs to get adjusted by NASA. Who knows.
It’s not just the Great White North. Look at the predominance of unadjusted data in Australia and China – I guess Hansen didn’t get the memo that there was urbanization going on in China. Maybe someone will send him a picture of Shanghai then and now.
Here’s the same graphic for stations which had adjustments.
I’ve added a comparison showing the difference between the last 10 years and the 1930s – I need to figure out a way of showing the color coding, but it’s the usual color coding – red hot and blue cold. As you can see, the pattern in the U.S. is different.
Here’s Chris’ version: