Some empirical evidence that the 1200km GISS data gets rounded up to enhance the appearance of warming. Below are identical graphs of July, 2007 anomalies. First pair is the raw data at 250km rounding, and the second pair is the GISS 1200km rounded data. Three problems which exaggerate warming become obvious.
- The rounded maps interpolate red through vast regions where there is no data - like most of Canada and the arctic. There is not a single data point shown in the arctic regions of Canada, yet the rounded map paints a broad brush of red across those regions.
- The rounding algorithms push all of the temperature regions upwards. Note the graph of the raw (250km) data shows average temperatures below normal for most of the southern hemisphere
- The map projections greatly exaggerate the size of polar regions and give the appearance of large areas of warmth, which are in fact much smaller than they appear.
Figure 1. Raw data rounded to 250km
Figure 2. Most of the area 0-60 south latitude shows below normal temperatures.
Figure 3. GISS rounded data to 1200km. There are vast areas of red which have no data points to back them up. The impression from this map is that July, 2007 was very hot, yet as you can see from the raw data above in figure 1, that was not the case in most places. Most of Antarctica appears bright red, yet this is based on a single location at the south pole where the temperature averaged colder than 65 below zero Farenheit. The Antarctic penninsula was about eight degrees below normal, but that is missing from this map. See figure 1.
Figure 4 Rounded data shows most of the southern hemisphere from 0-60 S latitude as above normal.
Figure 5 shows the rounding errors. Blue regions are latitudes where temperatures were rounded up, and brown regions are rounded down. There is clearly a much greater region of rounded up temperatures than rounded down. Maybe by a factor of 10x.
