Attached is the July of 2019 global temperature anomaly from the CFSv2 per WxBell. It has the global anomaly at only +0.305 C vs 1981-2010. What struck me as surprising in the map is how much of the CONUS is pretty solidly cooler than normal. So, I decided to investigate.
I looked at numerous US cities’ July, 2019, mean anomalies starting with some in SD and NE. Then when they almost all came in warmer than that CFSv2 showed, especially in the eastern halves of those states, I decided to expand my analysis to cities in states such as MN, IA, IL, AZ, CO, LA, AR, TN, MS, AL, GA, FL, and the Pacific coast states. I made sure I was comparing apples to apples such as making sure the normals were based on 1981-2010 and I converted from F to C. I found only a very small % that were close, mainly in AR (where the coolness verified pretty well) and FL (the
@pcbjr (and for a short time my own) home of Gainesville did actually come in near -0.5 F). But the vast majority came in 1-2 F or even warmer than the CFSv2 map showed!
After seeing this, I found out from Radiant/Maxar that Moscow came in 3.6 cooler than its 1981-2010 normal. If you look closely at the map, Moscow is on the edge of the green, some of the coldest in the northern Hem. The edge of the green is -3 C or -5.4 F. Comparing the -5.4 F suggested by the map to the actual of -3.6 F tells me that even at Moscow it is close to 1.8 F too cool!
When you folks get a chance, check for yourself rather than take my word for it. Why is this so important? Because WxBell is essentially claiming based on this CFSv2 map that the globe was only 0.305 C warmer than the 1981-2010 globe average. But if one were to adjust for the average error in the too cool direction, the actual could very easily be more like 1 F warmer than +0.305 C. If so, that would bring the actual anomaly closer to +1 C vs 1981-2010 normals! Even if the average error in the map were only 0.5 F too cool, that would still be enough to warm the anomaly from +0.305 C to near +0.55 to +0.60 C. This is a big deal to me and should be investigated. I do not trust Weather Bell at all. I also do wonder about the methodologies of the CFSv2 to determine this map.
Any other opinions?
View attachment 21431