Tony Heller's Dishonest Graphs

In the past I've shared how Tony Heller lies about bias correction procedures applied to CONUS temperatures and as well as by using simple averages of station data with no gridding, meaning that he fails to accounting for station moves and closures. I've also been shown these graphs from him below (located on his blog here) that he uses to support both of these lies. This graph is intended to show that scientists are conspiring to fabricate global warming through bias correction. In his words, "the data is being tampered with precisely to match the increase in atmospheric CO2." To make his case, Heller shows difference of USHCN final - USHCN raw "temperatures" (it's actually not temperature; see below) plotted with CO2 concentrations. Since the plot shows a roughly linear increase with CO2 concentration, Heller suggests this is evidence that scientists are just engaging in criminal fraud by manipulating USHCN final temperatures to fabricate a warming trend consistent with CO2 concentrations.

Here's another graph of Heller doing the same kind of thing but with the year on the x-axis instead of CO2 concentration. Notice the acceleration apparent in the adjustments when plotted as Heller has plotted them.

These graphs are wrong for lots of reasons, and it's probably good to summarize them:

1. Heller uses a simple average of station data. He takes the mean of all Tavg final station values and the mean of all Tavg raw station values and subtracts them. Any station moves, adds, or closures add bias to the data. About 30% of the ~1200 USCHN stations have closed over the last 30 years or so. The mean station altitude and latitude has substantially increased with these closures, which again adds bias to Heller's data. If Heller properly calculated his averages, this wouldn't be a problem for him, but since he insists on his dumb averaging, this adds a significant bias to his calculations, meaning that his Final - Raw values are not meaningful.
   

   

   Average Latitude and Altitude have Increased
   from "Dale Gribble" on X

   
   
2. USHCN was deprecated in 2014 in favor of nClimDiv (with ~10,000 stations), and the old network is no longer maintained, though data tables on NOAA's ftp site still exist and station logs are still uploaded there. Heller should switch to nClimDiv but he refuses to do so because that would be more difficult for him to lie.
3. Heller has no units for the y-axis on either of his above graphs. Is he presenting his Final - Raw data as °C or °F? He doesn't say, but I'm now pretty certain it's neither. He appears to have normalized the differences to values between -1 and and +1; he seems to have set the smallest Final - Raw value to (296, -1) and the largest Final - Raw data to (412,+1) and then scaled the differences between those two values. The only problem with this is that the y-axis is wrong, since the data point he labels (206,-1) is plotted below -1 on the y-axis. And the same mistake occurs on the other graph. His (1902, -1) data point is plotted below -1 on the y-axis. By not indicating his methodology on the graph, Heller lets the reader assume these are units of temperature (either °C or °F), but this is clearly not the case.

So I decided to plot correct versions of each graph. The graph below shows the (now deprecated) USHCN Final - UHSCN Raw in actual units of °C. I maintained Heller's scale, though with proper units of -1°C to +1°C, so that you can see how the correct numbers don't show the slope of Heller's graph.

Likewise, the graph below shows USHCN raw and final and the difference of Final - Raw in units of °C. Again, accelerating curve in these differences disappears when they are properly plotted.

Quite clearly, Heller is using these tricks to lie about CONUS temperatures. He's fabricating "evidence" for his baseless conspiracy theory by deliberately manipulating CONUS temperature data through 1) using a deprecated dataset, 2) his dumb averaging methodology, and 3) removing units from his graphs and "normalizing" the differences to values between -1 and +1. Heller appears to be hoping that his followers are gullible enough to believe him with no skepticism or critical thinking.