On March 15 I posted a note on the blog which covered the really bad state of the weather data. Towns 20 miles apart show vastly different temperatures--averaged over the entire year. As I emphasized, these are not daily temperatures but annual averages. One of the things that came up in the discussion, brought up by queen-of-fractal-beauty, was that we are dealing with microclimate. I don't care if it is due to microclimate, it is noise in the system that can't be fixed merely by belief that the editor knows what the true temperature is. But, queen-of-fractal-beauty did admit:
"Knowing how to throw out the noise so that you are left with meaningful information is a science unto itself. Raw meteorlogical data is nearly impossible to read. That's why all the charts and graphs put out by the community show corrected or smoothed data. They aren't "hiding" the truth. They're removing the noise so that we can see the truth."
Yes, we would agree that raw meteorological data is impossible to read because it is so incredibly noisy. And in this case, Queen-of-fractal-beauty thinks the noise is due to microclimate.
So, let's see how far the claim of 'microclimate' can be pushed. The Chinese data has mega-microclimate problems, up to 10 deg C, yet it too is used in global climate calculations and it too is 'fixed' by those who know the truth before they begin editing the data.
I lived in China for a year and a half, so I was interested if I could get the Chinese data. I could. I went to the Chinese governments climate center. here
There they give out monthly average temperature data from 1951 to 2008. One must down load and collate the individual months from individual stations. After doing all that work today, I then used the latitudes and longitudes to calculate distance. I used 70 miles per deg latitude (approx, it is 69 and some change) and I used the delta longitude x cos(latitude)*70 for the x distance. This should give mileage close enough for the purposes of this work.
The one unfortunate thing is that the names of the towns do not come across in ASCII in the readme.txt file. Thus, I don't know the names of the towns, only the station number, which is what I am using.
Then I searched through the matrix of distances for closely spaced station. It was easy finding them along the diagonal, but harder in the body of the matrix. I will assure the reader that in some of the really egregious temperature differences which I will show, I double checked their accuracy. I went all the way back to the downloaded months to check to see that the data in my table corresponds to the data I downloaded. It does.
Looking at the temperature charts, what can one say about the Chinese data other than oy vey? This stuff is utter crap. I have double checked the distance calculations as a gut check against the lat/lon, and if anyone wants to dispute that calculation I would be glad to hear of any correction. I have put the lat/lon on the chart for just that purpose. So, check away. I don't really care if I am off by a mile or two. I do care if I have screwed the distances up by a lot.
It is easy to see that neighboring towns in China have a worse correlation problem than anything I found in the US. One is tempted to think that the Chinese data is made up, because it is so bad.
Feel free to duplicate my work (it took all day today what with the double and triple checks because the data is so awful).
After seeing this data, one must ask how on earth can we use it to determine the climate? It is utterly worthless and the noisiness of it renders any claim that the earth has warmed by .8 C over the past century quite meaningless.
This one is on the Yellow River plain about 300 km south of Beijing.
This is a ridiculous amount of 'microclimate. Here is another
Likewise, this is a ridiculous amount of microclimate.
Now, I must emphasize that all these stations are in the eastern part of the country in order to avoid having temperatures taken at high altitudes. One can see that the closely spaced stations above don't even correlate in the direction they move.
Below are two nearby stations in SW China, Yunnan province, a place I have been to. The temperatures are about what I would expect for that area.
The problem with the data, as I see it is there is an offset in the time of the coldest temperature between these two nearby cities (about 80 miles apart). This is just what the data says. I can't help it, I think it is awful and unbelievable, but this is the data used for the land calculation of the global warming. Statistically, it is, as they say in the UK (another place I have lived), a dog's breakfast.
Here are two stations which look like they are on different planets. One warms and the other cools tremendously
Now, both of these are in the western part of China, the trend is terribly different and the difference eventually amounts to 22 deg C, or 40 deg F. Such a data set like this is what the climatologists are using to convince you that the world is warming. Do you believe the data can support such a bald assertion? I don't.
Below are also two stations in central China, south of Beijing, about half way along the Yellow River from Xi'an (where the terracotta soldiers are) to the mouth of the Yellow river in the Bohai Bay.
Subtracting these two stations we find
10 deg C as a difference is quite unbelievable for a yearly average, but here we find 13.5 deg C difference. That is a 14.46 deg C difference. In Fahrenheit that is a difference of 26 deg F FOR THE ANNUAL AVERAGE TEMPERATURE!!!. Anyone who believes this crap and claims that it supports global warming is simply either not skeptical, a hopeless believer or an idiologue. Scientists should go with the data. It is quite easy to check me out.
I must note something for those who don't know. I lived in China in 2005 and 2006. I speak Mandarin. The weather events depicted in the above picture (132-38) simply didn't happen because I was there, and personally know it didn't happen. It wasn't that cold for an entire year. These are the annual average temperatures, not daily temperature differences. The data is in deg C. The towns are 75 miles apart and that area of China is relatively flat. There is no way that station 38 was so cold as Beijing wasn't that cold for the entire year. The data is garbage.
For what it is worth, both of these stations show cooling since 1951, not the warming the hysteriacs claim.
One more example, from near Wuhan China. First the two temperatures plotted on the same chart. These towns are about 62 miles apart and look how different the temperatures are. This is not in an area of high elevation in China.
Now, lets look at the subtraction of these two stations.
Such differences, 6 deg C, means it is impossible to use this data to try to detect a .8 deg C change in temperature over the past century. Statistically it is absurd to say that the temperature has risen by .8 deg C plus or minus 6 deg C. Yet that is, effectively, what the climatologists are doing.
The average temperature of china from 1951 to 2008 is 12.6 deg C +/- 4.6 C. Since the earth has supposedly only warmed by .8 deg C, to see that kind of standard deviation in the data is bad. One can't really say, from the input of Chinese data, that the earth has warmed or cooled. the standard deviation in the measured data is far too great to detect such a tiny signal in the temperature stream. But, that does not seem to stop the global warming hysteriacs.
For those who might think changing to an anomaly representation would help, the SD doesn't change with the subtraction of a constant value from the data stream. So, while datumizing the data to the 1950-1970 average temperature will change the temperature into an anomaly, it won't change the deviations. Thus when one edits the data by 20 degrees, either in anomaly or temperature space, one still gets the same change of SD.
To me, the real question is why the climatologists, who see this kind of crap data daily don't feel ethically bound to say that the data won't support global warming. I would disagree with queen-of-fractal-beauty that reading raw meteorological data is impossible. It is very easy to read. The problem is that it doesn't support any basis upon which to claim that the earth is warming or cooling. The data is crap.