Spencer: Spurious warming demonstrated in CRU surface data ( raw data vs adjusted data)

Spurious Warming in the Jones U.S. Temperatures Since 1973INTRODUCTION
As I discussed in my last post, I’m exploring the International Surface Hourly (ISH) weather data archived by NOAA to see how a simple reanalysis of original weather station temperature data compares to the Jones CRUTem3 land-based temperature dataset.

While the Jones temperature analysis relies upon the GHCN network of ‘climate-approved’ stations whose number has been rapidly dwindling in recent years, I’m using original data from stations whose number has been actually growing over time. I use only stations operating over the entire period of record so there are no spurious temperature trends caused by stations coming and going over time. Also, while the Jones dataset is based upon daily maximum and minimum temperatures, I am computing an average of the 4 temperature measurements at the standard synoptic reporting times of 06, 12, 18, and 00 UTC.

U.S. TEMPERATURE TRENDS, 1973-2009
I compute average monthly temperatures in 5 deg. lat/lon grid squares, as Jones does, and then compare the two different versions over a selected geographic area. Here I will show results for the 5 deg. grids covering the United States for the period 1973 through 2009.

The following plot shows that the monthly U.S. temperature anomalies from the two datasets are very similar (anomalies in both datasets are relative to the 30-year base period from 1973 through 2002). But while the monthly variations are very similar, the warming trend in the Jones dataset is about 20% greater than the warming trend in my ISH data analysis.

This is a little curious since I have made no adjustments for increasing urban heat island (UHI) effects over time,.....

From the article linked at post #7:

********************************EXCERPT**********************************

Jan 20 2010

Published by AJStrata at 10:46 pm under All General Discussions, CRU Climategate, Global Warming

As I have learned more and more about the ’science’ and ‘math’ that is the foundation of the man-made global warming (AGW) theory, the more I realize how amateurish that science and math really is. For those of us who deal day in and day out in complex physical systems, driven by multiple natural forces that are working on scales that few humans can get their heads around (e.g., the exploration of space), the case for global warming is incredibly flawed.

What I will do in this long and technical leaning post is identofy where the global human induced warming(AGW) math completely falls apart, bringing down the entire house of cards that is the AGW theory. This is different from what other skeptics are doing by trying to reproduce the flawed process of creating a global index for today and then running back to 1880. In my view, the approach used to create a global index from land based sensors is fatally flawed, and we can prove it beyond a shadow of any doubt.

My basic premise is the very large and complex set of natural systems driving Earth’s climate cannot be measured sparsely and then declared to be completely understood on a global scale. The core of the problem with the current theories is in the calculation of uncertainty or error – which directly reflects on the confidence of the results. The process of globalizing these sparse measurements causes this error to rapidly explode when extrapolated to small regions, let alone global scales.

The basic question is whether the noise in the raw data is so high that it is impossible to detect a global rise of less than 1° C over a century (the current claim of the AGW crowd). I will show that even today’s computations of a global temp index are incapable of an accuracy to 1°C, let alone to that same confidence back to 1880.

Where the AGW ’scientists’ make their fatal mistake is not understanding how temperature accuracy decays over distance. Once you measure this decay rate (which is easy to do) it becomes clear we cannot develop a global temperature index of even modest precision from land based sensors

Here is the first of two diagrams that lit the light bulb for me. it is the global temperature map using methodologies employed by the IPCC and AGW theorists for years and is based on land based measurements extrapolated up to global scales.

The thing you have to understand about this map is it is 99+% guesstimates! Only a tiny fraction of the map is actually measurement. And those guesstimates are very imprecise.

The CRU, GISS and NCDC all derive global profiles by making some incredible and unproven assumptions about how a temperature measurement can be extrapolated to represent a city, a region, a country, a hemisphere and the globe. For example, in the CRU data they assume you can create a 500 km ‘grid’ from one or a small number of temperature stations. Even worse, they assume you can create a neighboring 500 km grid that has no stations from nearby grids (which are poor guesstimates in the first place, adding insult to injury).

It is pure folly, as a simple analysis can show on any given day of the year.

From the comments at website article linked at #7....a new website to me:

Monday, January 18, 2010 Mapping global warming

*****************************EXCERPT*******************************

[Important Update: After coming across this thread entitled 'Comparing Surface Temperature Records' on the TreesfortheForest blog and therefore being reminded of the warming/cooling trends in the MGST anomaly chart, I have decide dto re-analysis teh trends and maps for the time period from 1880 to 1939. It is evident from MGST anomaly chart that 18880 was a cooling period followe dby a warming period from 1910 to 1939. Consequently I've now split the reviou sanalsys for 1880 to 1939 into two disticnt time periods i.e from 1880 to 1909 and 1910 to 1939. The results are very interesting so please re-read the whole of this thread]

Recently I've been working in collaboration with Verity (vjones) on 'mapping' possible global warming/cooling in both the NOAA and GISS GCHN raw and adjusted datasets. This work has resulted from the recent work we've been doing on developing a database and user interface to the data (TEKTemp) for storing all the NOAA and GISS raw/adjusted data. TEKTemp is described in more detail in another thread here.

I don't like seeing raw (or for that matter adjusted data) presented in the form of 'anomalies' from a so called 'reference period' and so have chosen thus far instead to break with convention and do all my trend analysis of the NOAA/GISS data using actual raw/adjusted temperatures. One of the great advantages of computer programming is that it enables you to repeat repetitive tasks very easily (a DO .... LOOP) and so it's been possible to create charts that show the raw and adjusted data (and trend lines for each) for the many thousands of stations in the NOAA/GISS datasets.

After performing the trend analysis (linear regression) for each station I then needed a convenient user-friendly way in which to present these seevral thousand charts to anyone who is interested in looking at indiviudal stations and so I came up with the idea of using an interactive map. After a bit of Googling I managed to identify a nice free-to-use Flash map application called 'DIY Map' which as you'll see does 'just the job' I need it to do.

The first thing to note is that I've analysed the NOAA GHCN and GISS GHCN datasets separately and so have produced completely separate 'interactive maps' for each dataset. The NOAA maps are here

NOAA raw/adjusted data interactive maps

and the GISS map are here

GISS raw/adjusted data interactive maps

The map uses an XML file which is loaded into the DIY Map application. The application is written in Flash (Actionscript) and so it can take a while for the map to load all the XML data and display it. As the data is loaded you will most likely see a prompt which says that it is taking sometime for Abobe Flash 10 Player to load the application data. If so then when prompted please click 'No' on the prompt that appears and eventually the full colour map be displayed.

The interactive maps (accessible using the links above) allow you to zoom in to a given country just by clicking on a part of that country and you can just as easily zoom out again. Note that there is a colour coded 'dot' on the map for each station in the NOAA/GISS dataset. When you hover over a dot it pops up a little 'summary info' window which tells you the name of the station, its WMO station code and raw/adjusted trends. If you then click on the dot a new browser window will open showing a chart of the raw/adjusted data for that station. As I get more time I'll add further info to this station details page.

Each map has a legend which explains the colour coding for the station dots. Most of the maps show colour coded raw/adjusted slopes for the linear trends in the stations raw/adjusted temperature data. Note the units for these slopes are 'deg.C/century'.

To save some space on this thread and since the legend is the same for all the maps that show raw/adjusted temperature trends here is an image of the slope legend (Figure 1). Note the colours and the corresponding slope ranges when you view the maps that follow on later in this thread.

Figure 1 - Colour coded slope ranges - Map legend

Now let's make a start by looking at the raw NOAA data as shown in the following JPEG captured image of the NOAA 1880 to 2010 raw data map (Figure 2).

This comment is from the Digging in the Clay website...but referencing the AJStrata comment on his website...see link at #7....

******************************EXCERPT*****************************

Oefinell (02:45:04) :

ajstrata (19:47)
Interesting post but misguided. Your blog correctly states that it is not possible to get an accurate global surface temperature from current data. In fact that is well understood by Climate Scientists. That is why they do not measure the absolute temperature at all, they measure the temperature change or variance.

Provided the methodology is consistent from month to month then you can accurately see how temperatures have changed over time. You do not need to calculate an absolute temperature at all.

There are lots of further questions over how their estimates of variance are calculated (data adjustments etc) but they are not addressed in your blog post.

Your final idea is a good one though – comparing surface temperature measurements with the satellite data. There are lots of papers on this subject if you care to Google a bit. In fact all the data is publicly available and it is very simple to calculate a variance from the satellite data and compare it with a variance from the GISS data – since 1979 at least.

They actually correlate pretty well, although the satellite trend suggests it is warming a touch slower than the surface stations but not so much that it invalidates the GISS methodology.

Graphs and links to data here:
http://www.climate4you.com/GlobalTemperatures.htm#Comparing%20surface%20and%20sattellite%20temperature%20estimates

I think the sceptic community needs to get over all this controversy with the surface station data. There are other much bigger and more important questions – in particular whether the climate sensitivites included in the IPCC models are correct. These are largely calculated from paleoclimate studies so the data from surface stations is irrelevent.

A recent study by some eminent climate scientists led by Steven E. Schwartz, including one from the NASA team, highlight the problems titled:
“Why Hasn’t Earth Warmed as Much as Expected?”
(good question!)
http://ams.allenpress.com/perlserv/?request=get-abstract&doi=10.1175%2F2009JCLI3461.1

If you dig around you can find links to the full report – one in a comment here if I recall correctly.

This paper is remarkably clear of AGW spin and points out the urgency of finding more accurate estimates for the effect of Aerosols (air pollution) which are usually blamed by the modellers for the global cooling that happened 1940 – 1970. If the effect of aerosols turns out to be too insignificant to account for this period of cooling then the paper states that the climate sensitivity in the models must be too high.

Of course this is only one among a ton of papers on the subject, however it is refreshing that some climate scientists are prepared to stick their necks out and ask the question the rest of us have been asking ourselves for some time!

Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.