Global Warming Data

The simple physical theory in my previous post on global warming, coupled with measured data, is not quite sufficient to prove that global warming caused by man is not real. Nevertheless, it is highly suggestive that this is the case. We turn then to a purely empirical look at data, especially as used by the believers in Anthropogenic Global Warming (AGW).

First, we should look briefly at the historical record. This is a hazardous task to attempt since there is a lot of controversy on just what that record is. You can probably guess for yourself what some of the problems are. The value of a world-wide temperature average will depend a great deal on how you choose the places on Earth whose temperatures are averaged. If you choose samples in or close to cities, you will probably get temperatures greater than in rural areas (urban heat island effect). If one tries to form an average over the entire surface of the Earth, you will need to adopt some interpolation scheme to give temperatures at places where there are no sampling stations. The interpolated values will vary somewhat using different interpolation algorithms. Also, temperatures will vary with the time of day. Variations can even be caused by differences in the measuring equipment at different sampling stations. Calculating a world-wide average for temperature is not an easy problem.

The results become even more controversial when over the years some sampling stations are eliminated and replaced by stations at different locations. Using different ways of creating the average, one can get very different temperature anomalies (differences in temperatures with some standard temperature). Even the choice of the standard temperature from which the temperature anomaly is calculated can affect the result. In particular, the  global temperature datasets of NASA’s Goddard Institute for Space Studies (GISS) and of the National Climatic Data Center (NCDC) of the National Oceanic and Atmospheric Administration (NOAA)  have been criticized for inaccuracies (See here and here and here and here and here and here and here).

With all those warnings I will now show data from NCDC plotted from the year 1880 to 2014. I do this for two reasons: first, it shows qualitatively three historical features that are generally agreed upon; and second, it also shows some suspicious quantitative results. The data is plotted as temperature anomalies, but I do not know
NCDC Temp. Anomalies 1880-2014what the reference temperature is. Eyeballing the plot, the reference temperature is probably the average of all the temperatures in the data set.

It is generally accepted that the Little Ice Age began around the year 1300 and ended sometime between 1850 to 1870 (See here and here and here). Interestingly, the NCDC dates shows the world average temperature continuing to fall from 1880 until the year 1910. The average temperature then reversed and increased from 1910 to around 1940. From 1940 until 1975 there was global cooling. There was enough cooling of the planet, alarm was raised that another ice age was imminent. A famous 1974 issue of Time Magazine heralding the coming ice age is often cited as evidence of the national worry (see here and here). Yet the NCDC data shows only a slight cooling from 1940 until about 1950. Their record then shows the temperature anomaly staying close to zero until around 1975. Temperature then rises from 1975 to about 1999, after which temperature has stayed relatively constant in a widely recognized “pause” in global warming.

You may well recall claims by GISS and NCDC that most every year since 2000 has been the “hottest year” on record. Their own temperature record shows temperature fluctuating about a flat smoothed curve showing the trend. In answer to the claim that 2014 was the hottest year on record, Dr. Roy Spencer offered a devastating response. One often heard criticism of the government data sets is that the choice of temperature measuring stations together with how they interpolate these temperatures over the Earth’s surface does not provide an accurate average global temperature. Dr. Spencer heads a group at the University of Alabama, Huntsville (UAH) maintaining a satellite-measured dataset of world temperatures. Spencer points out that thermometers at dispersed ground stations can not accurately produce temperature averages; only satellites that can look at every point on the planet can directly measure averages of surface sections of the globe, which then can be averaged to form the global average. He then produced a plot of two satellite-produced data sets, one from his own UAH group and one from the Remote Sensing Systems (RSS) group. The plot, shown below, shows the global pause

UAH and RSS satellite temperatures between 1999 and 2014 with greater resolution. The satellite temperature measurements solve the twin problems of surface coverage and interpolation through their analogue measurements of lower atmosphere temperatures. These data sets should be regarded as intrinsically more accurate than the surface measured data sets of NCDC and GISS. Even if one considers the period of the global pause to contain the highest recorded temperature (a disputed position), the year 2014 is not that year.

Further evidence is provided by taking each of the major data sets of measured average world temperatures and finding their trends by doing a least-squares fit of each of the data sets to low order polynomials. This task was performed for us by a gentleman named Chip Knappenberger to whom we all owe our thanks. The data sets themselves are fairly noisy, as seen below. The orangish-yellow curve is the NCDC data set, the dark maroon belongs to the GISS data set, while the light purple curve is of the CRU data set from the Climatic Research Unit of the University of East Anglia. The CRU global temperature data set is ground-measured.  The UAH data has a light cyan color while the RSS data is colored dark blue. When the least squares fit is performed to extract the trends from the noisy curves, we get the smooth curves below

Comparisons of Major Temperature Data Setsthe raw data of the global temperature averages.  If you look at the units of the trends(deg C per year), it is clear what Knappenberger means by a trend are the slopes of the fitted temperature averages, i.e. their time rate of change. That means whenever a fitted trend curve falls below zero, the temperature in that series is falling. For readers who know a little bit about calculus, he took the first derivative of the fitted function and plotted that. The top, almost horizontal red line is the prediction from the computer climate models that is sloping upwards very slightly. The zero trend line that the trends would follow if the temperatures were constant at the reference temperature is shown as a thin grey line. As you can see, every one of the data sets, including the ones from ground measurements shows decreasing average planetary temperatures with time.

Trend lines of major temperature data sets All major measured global average temperature data sets show global cooling, not global warming! This is despite the fact that some of the ground measured data sets are suspected of being fudged! If they have been fudged, reality has not given them sufficient flexibility to achieve what they want. Note the more accurate, satellite-measured data of UAH and RSS give the lowest temperatures.

It may seem very ungenerous to suspect the GISS, NCDC, and CRU data sets of being put together with intellectual deceit, but there are some reasons for believing that. A very useful recitation of this growing suspicion by a number of individuals has been put together by Francis Menton, who owns the manhattancontrarian.com website. In short, it is believed by some that algorithmic adjustments were found to systematically adjust older temperatures down with the temperatures of the last two or three decades adjusted upwards to emphasize global warming with fictitious warming trends. It is possible this merely showed a mistaken, intellectual enthusiasm for AGW rather than intellectual fraud, but the uncovered evidence is rather convincing. I invite you to read Menton’s six posts on the subject and make up your own mind. In chronological order, they can be found here and here and here and here and here and here.

CO2 in the atmosphereThe measured evidence, together with the theoretical estimates for the time a typical infrared photon stays within the atmosphere on the order of 100 μsec, is a convincing case. Infrared radiation is not trapped in the atmosphere, but although greenhouse gases (mostly water vapor) slow it down, the radiation still escapes fairly quickly. Yet throughout the last several decades carbon dioxide has steadily increased as shown below. We can only conclude that global temperatures are not as sensitive to CO2 concentrations as the global warming alarmists would have us believe. But then what could have caused the periodic intervals of global warming, such as between 1975 and 1999, that we have seen? We will look at that question in the next post in this series.

Views: 7,326

GO TO HOME

0 0 votes
Article Rating

Sharing is caring!

Subscribe
Notify of
guest

This site uses Akismet to reduce spam. Learn how your comment data is processed.

2 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Cheese

I’d be interested in your take on the following blog post:

https://tamino.wordpress.com/2010/12/16/comparing-temperature-data-sets/

which shows that the differences in the sources mostly disappear if you simply use the same time period as the reference point for the global average temperature. The data also suggests that the satellites data’s relatively greater drop after El Nino was because the troposphere temperature is more sensitive to el Nino warming and la Nina and volcanic cooling activity.

blank

I have to admit I have been worried about whether the data sets had a common zero, and it is gratifying to see their similarities once a common zero is picked and a filter for the noise applied. It would have been better had the author applied a cubic spline or piecewise quadratic fit, rather than a 12-month moving average, as the moving average weights past data too much. However, significant differences between the satellite data sets (UAH and RSS) and the ground station data sets (GISS, NCDC, and HadCRUT3v) remain. Just “eye-balling” the data suggests to me the satellite… Read more »

2
0
Would love your thoughts, please comment.x
()
x