I’ve started putting up the source code and more detailed analysis in a few links. I’m just going to list them here and give them a “one line” description. The original posting continues just a bit below.
When the GHCN data set is reduced to the 3000 thermometers with the longest records (cut off at about 64 years worth of data for the station), the “global warming” signal is not present.
When we look at the data on a seasonal basis, we find that the “warming signal” is present in the winter months, but not in August. This can not be a function of decade long solar changes nor of long term accumulation of CO2. Changes in the sun may be responsible for other things, and CO2 may well cause some other effects (like improved plant growth) but the global warming signal is too seasonal to be either of them:
UPDATE: Yes, an “update to the update” ;-) I’ve found the cause for the thermometer deletions. There was a push from another UN committee to make a network of thermometers focused on Climate Research. (CRN, RCS various names in different countries) that as near as I can tell has lead to the thermometer deletions. NASA / NOAA deleted the thermometers from the GHCN. I look at the global pattern of deletions by continent and by major country here:
A recent posting on The March of the Thermometers that looks at GIStemp boad zones vs a better set of zones, and what happens to thermometers over time and space is here:
That CO2 has the ability to only work in winter is, er, odd? Or maybe it isn’t CO2 that’s the issue…
What about those short lived stations? When we select the 10,000 stations (representing less than 1/2 the data) with the shortest lives, we get a very strong warming signal in the data. For a future bit of work, I’ll be looking in more detail at exactly which stations contribute the most, and when. But it is pretty clear that the warming signal comes with the addition of thermometers… The spacial, in addition to the temporal, distribution of “warming” do not allow for the cause to be a diffuse broad acting agent like CO2.
A finer grained look, by quartiles of age, with a “bonus look” at the 10% best stations is here:
And a look into how much of this signal gets through the “temperature” steps of GIStemp (everything up to zones) is here:
The crib note is that a lot makes it to the zonal stage and with a general overall warming of the data set by about 1/2 C but the “tilt” seems to come from being a poor filter for the data profile, not from a coding failure. Realize that this “warming of the data” is in addition to the warming signal in the “raw” data. GIStemp is acting as an amplifier up to this point, not as a filter.
Want to “try this at home”? Here’s a listing of the code I wrote to do some of this:
This is a copy of a letter that I posted on WUWT under the “tips” tab.
If there is no followup, I’ll start posting the details of the code, methods, and conclusions here.
I have been “characterizing” the GIStemp process and how the data are transformed as they go through the process. Along the way, I have discovered a couple of Very Interesting Things. I would like someone to verify these findings (since they may be suitable for publication). A couple are fairly simple (but have significant implications). One is, IMHO, a bit of a “doozy”…
I used the “raw GHCN” data from STEP0 of GIStemp as the seed for my “benchmark”, then put together a couple of FORTRAN “hand tools” to see what happened to the trends. What I found was:
1) There is a pronounced seasonal variation in the “Global” average temperature. This means that the GAT is decidedly biased to the Northern Hemisphere. Hemispherical changes can easily bias the “GAT”. (i.e. effects of change of axial tilt, of precession (which pole is close to sun at perigee) etc.) For example:
GAT year: 1900: (by month, starting in January)
0.6 0.4 4.6 10.8 15.8 19.7 21.5 21.7 17.8 13.6 6.2 2.4
2) In calculating these “Global Average Temperatures” I found that the exact method of calculation strongly changes the result. Do you average all the separate valid records and then divide by the count of valid records? Or do you calculate a yearly GAT, then average those to get a decade or total data series GAT? This implies that the number of thermometers active in any particular period of time has a strong impact on the GAT in that time. For example:
0.2 1.6 4.7 9.5 14.0 17.7 19.6 18.9 15.6 10.7 5.5 1.8 10.0 2.6 3.9 7.3 11.8 15.8 18.9 20.7 20.3 17.4 13.1 7.9 3.9 11.97
These two series are the average of all station records, by month, with an average of all temperatures in the 13th field. The difference is that the first record is adding the individual temperatures, where the second series is from averaging the individual years first, then averaging those averages (much as GIStemp does). It’s clear that a degree C (or 2!) in the GAT is an artifact of the number of stations used in any given average and the order of averaging.
3) Finally, this lead me to the idea of selecting only those stations with a long history (I now have a FORTRAN program that takes the GHCN format files, counts the records for each station ID, then sorts them in rank order by total years of data and lets you select a “cutoff” value. I chose to use a 3000 station cut off (that gives about 64 years for the “short lifetime” stations) but similar results happen with 1k, 2k, and even 4k stations. The result is that almost all of the AGW “signal” goes away. The conclusion is that the AGW “signal” is an artifact of the arrival and departure of thermometers from the scribal record. The addition of more thermometers in the Southern Hemisphere followed by the loss of Siberian thermometers with the collapse of the Soviet Union. The thermometer count rises from 1 in 1701 to over 9000, then drops back to under 3k today. That has an impact… I calculated “decade Global Average Temperatures” from a data set reduced to the 3000 longest lived thermometers. A sample of the records are below. The next to last field is an average of all data for the decade, while the last field is the number of thermometers (station IDs) active in this group from that 3000 in the data set. You can see that the GATs don’t change much from decade to decade anymore.
DecadeAV: 1890 0.6 2.2 5.8 11.9 17.0 20.9 23.2 22.4 19.0 13.2 7.2 2.9 12.2 1174 DecadeAV: 1940 0.3 1.3 5.4 10.4 15.3 19.1 21.6 20.9 17.5 12.3 6.2 2.0 11.0 2851 DecadeAV: 1990 -0.1 1.4 5.7 10.8 15.4 19.3 21.6 20.9 17.2 11.9 6.1 1.0 10.9 2186 DecadeAV: 2009 1.7 2.2 6.6 11.7 15.9 19.9 22.3 21.7 18.0 12.3 6.9 2.1 11.8 209
I have no explanation for why the long lived thermometers drops to 209 in the last data.
To say that I think this is of some importance is a bit of an understatement.
So my question to you is simple: Would you or someone you work with like to duplicate these results and potentially collaborate in producing a paper suitable for publication? I don’t have the PhD / credentials to publish, and would benefit from someone with skills at making tables into graphics anyway ;-)
I will be putting a copy of this letter up on my blog, and I will be putting up the code and a bit more detailed set of data / results over time (unless someone wants to “vet” the code and results first for publication). Please let me know what, if anything, you would like to do.