Climate Change & Tropospheric Temperature Trends

Part II: A Critical Examination of Skeptic Claims
“The computer models say that the troposphere should have warmed by +0.5 C in the last two decades. However, both NASA satellites and weather balloons show virtually no troposphere warming.

A new paper by Santer et al. attempts to debunk the satellite record. They claim that a satellite dataset produced by Remote Sensing Systems (RSS) in Santa Rosa, California, is more accurate than the dataset produced by climatologists Roy Spencer and John Christy at the University of Alabama in Huntsville (UAH). Why is the RSS dataset more accurate, according to Santer et al.? It conforms more closely to climate models. But data is supposed to confirm models, not the other way around. The UAH dataset agrees with a totally independent troposphere temperature record—weather balloon observations, which show about 0.08 degrees C of warming trend (see Figure 1) during the past two decades even when one includes the large warming contributed by the 1997-1998 El Nino event. The UAH results are plotted side by side with two independent determinations of the global temperature of lower troposphere in Figure 1. Note the near-perfect agreement (with correlation coefficients greater than 0.94 and 1 being perfect correlation) between the UAH satellite record and (a) balloon results from the U.K. Meteorological Office (marked HadRT) and (b) the assimilated global lower tropospheric temperature deduced by U.S. National Centers for Environmental Prediction (marked NCEP). The latest UAH effort in confirming the accuracy of the satellite temperature record and its error estimates are published in the May 2003 issue of the Journal of Atmospheric and Oceanic Technology (vol. 20, 613-629).”

(Ferguson and Lewis, 2003)

Once again, we see the same superficiality and cherry-picking that has become a trademark for organizations like these. First, radiosonde (“weather balloon”) records in general are not completely independent of the MSU record. Though the LKS record (Lanzante et al., 2003) cited here by GES is independent, most that are used in MSU intercomparison studies rely on MSU records to at least some extent to detect anomalous events. HadRT, which Ferguson and Lewis incorrectly report as independent, is one example (Parker et al., 1997; Free et al., 2002). While such corrections are generally minor for the products that use them, they cannot be ignored altogether.

Nor is it true that radiosondes are “a different measurement of the same quantity (the lower atmosphere)”. MSU and AMSU devices measure the bulk brightness temperature of broad atmospheric layers such as the middle troposphere and lower stratosphere. The narrowest measurement they can reasonably make is of the lower troposphere alone (MSUTLT), which is at least 7 km thick, and this observation can only be made with differencing methods that increase sampling noise (NRC, 2000). Radiosondes measure direct ambient air temperature using thermistors or bimetallic sensors of varying designs. These have a host of calibration and sensitivity issues of their own that are completely different from those of the MSU. Because they take snapshot readings of local temperature at discreet altitudes, they must be converted to equivalent MSU “brightness” temperatures by weighted summations over altitude and geographic region. There are many issues surrounding how this should be done as well. Many of the issues impacting the reliability of the radiosonde record are of precisely the type that will cause it to under-estimate actual air temperature (e.g. – the Phillips to Vaisala equipment shifts at many stations), so it is not surprising that these records are often lower than their MSU equivalents, as GES is emphatically asserts. None of this passes for measuring “the exact same things”.

There are also basic numerical misunderstandings in both the GES report and Ferguson and Lewis’ work. After yet another repeat of the bogus claim that Santer et al. (2003) attempted to “debunk” the MSU record, Ferguson and Lewis argue that there is “near-perfect agreement (with correlation coefficients greater than 0.94 and 1 being perfect correlation)” between UAH, and 2 other “independent determinations of the lower troposphere”. This is not only false, it reflects a serious misunderstanding of correlations. High correlations between MSU anomalies and those of other datasets imply only that they are not entirely independent. What is at issue are temperature trends, and correlation coefficients actually have very little to do with small trend differences. In this case rms errors and the confidence intervals derived from them are far more meaningful, and these typically show more a lot more variance between MSU and radiosonde products. The data that Ferguson and Lewis report as being a “near-perfect” fit to UAH Version 5.0 are the HadRT2.1 radiosonde analysis and the version of the National Center for Environmental Prediction (NCEP) reanalysis product 6 that was extant at the time. For UAH Version 5.0 comparisons to these products, the trend difference standard errors (which are based on annual anomaly difference rms errors) are +/- 0.0308 deg. K/decade for HadRT2.1 and +/- 0.0285 deg. K/decade for NCEP. The corresponding 95 percent trend difference confidence intervals are +/- 0.075 deg. K/decade for HadRT2.1, and +/- 0.067 deg. K/decade for NCEP (Christy et al., 2003 – see Figure 9). These values are large enough to include RSS Version 1.0. This is hardly “near perfect agreement”.

But if Ferguson and Lewis misrepresented the lower troposphere record, GES positively botches it. They reported that Lanzante et al. (2003) found no trend difference between their radiosonde product (LKS) and UAH Version D, but failed to note that this was for Channel 2LT only, for adjusted and sign-averaged global data only, and only through 1997, neglecting nearly a fourth of the extant record at the time GES made these comments. The absolute value of the trend differences reported by Lanzante et al. was larger globally (0.011 deg. K/decade), and considerably larger for median values computed over individual station averages (0.246 deg. K/decade). GES then compares these values with RSS trends even though RSS has no 2LT product. The trend value of 0.16 deg. K/decade they report for RSS Version 1.0 is incorrect and appears to be the result of simple carelessness. From the looks of it, GES started with the Channel 2 difference between UAH Version 5.0 and RSS Version 1.0 (approximately 0.10 deg. K/decade - see the earlier quote from the same article), and then added this to the UAH 2LT trend rather than checking the value directly. The actual Channel 2 trend for RSS Version 1.0 is 0.097 deg. K/decade (Mears et al., 2003).

These arguments are typical of most skeptic forums. In recent years, most rely on the LKS radiosonde dataset (Lanzante et al., 2003) as vindication of UAH products (without mentioning that it only extends to 1997 and leaves out the significant warming of the latter 90’s and early 21st century). Some point to HadRT and Angell products, including older versions of both. But all use carefully selected portions of each record to “prove” that the UAH record is more reliable than its competitors, even to the point of comparing RSS middle troposphere products to lower troposphere UAH and radiosonde records for which they have no comparable record, as GES did in the above example. But in the summer of 2004, three of the foremost global warming skeptics who to date have concentrated mainly on anti-global warming publicity and industry funded lobbying, escalated the skeptic challenge to a new level when they published papers in peer-reviewed journals – two that challenges the upper air record, and one that challenges its relationship with the surface record. These papers take the radiosonde/MSU comparison and its relationship with AOGCM’s in skeptic forums to a new level that merits a discussion of its own.

Douglass, Singer & Michaels (2004)

The large majority of global warming skeptic publications are popular media pieces or summary papers for policy makers. One is very much like another, and those I have cited so far are typical examples. Skeptics seldom publish their work in peer-reviewed journals or contribute to the peer-review process. Yet there are a few notable exceptions. In July of 2004, David Douglass 7, S. Fred Singer 8, and Patrick Michaels 9 lead teams that published two papers in Geophysical Research Letter in which they claim to have demonstrated that there is a clear disparity between surface and lower troposphere temperature trends (Douglass et al., 2004), and that current state-of-the-art AOGCM’s cannot accommodate it (Douglass et al., 2004b). Their arguments differ from those considered so far in that they attempt to formally demonstrate both a disparity in the observational record and a model-observation as well. They were also significant in that both were peer-reviewed and published.

In the first of these papers Douglas et al. (hereafter, DEA) use MSU data, radiosonde data, and a reanalysis product applied to the period of 1979 to the present to argue that the disparity exists and that it cannot be accounted for by any known tropospheric dynamics. To do this, they start with global surface temperature data from Jones et al. (2001). These are monthly anomalies with respect to the 1961-1990 average of global surface air temperatures over land, and below-surface water temperatures for oceanic regions, as represented within a 5 deg. by 5 deg. grid cells. This record is then compared with lower troposphere trends taken from UAH Version D MSU2LT Data (Christy et al., 2000) and data from a new “2-meter” temperature product (R2-2m) derived from an updated version of the National Centers for Environmental Prediction - National Center for Atmospheric Research (NCEP/NCAR) Reanalysis 6 (Kanamitsu et al., 2002; Kalnay et al., 1996). The latter is selected for its consistency and completeness between the surface and 850 hPa layers, and because it is (they argue) a dataset that is independent of both the MSU record and the radiosonde products that have been used to date for tropospheric intercomparison studies (Christy et al., 2000; 2003; 2004; Seidel et al., 2003, 2004; Angell, 2003). In the second (2004b), they compare results from 3 AOGCM’s with surface temperature trends similar to those used in the first paper (but taken from Jones et al., 1999 rather than 2001), MSU2LT data from UAH Version D (Christy et al., 2000), radiosonde data from HadRT2.0 (Parker et al., 1997), and 50 year results from the NCEP/NCAR Reanalysis (Kisteler et al., 2001). From these datasets they argue that the models, which represent the current state of the art in AOGCM’s, cannot account for the observed troposphere and surface temperature trends. Predictions of global warming during the upcoming century are based on AOGCM predictions. Because these models typically show tropospheric warming that is equal to or greater than that of the surface, DEA claim that their papers prove that significant global warming is not happening now and will not happen any time soon.

Shortly after these papers were published (Aug. 12, 2004), Douglass, Singer, and Michaels jointly published an article online at Tech Central Station (Douglass et al., 2004c) in which they triumphantly announce that the science of global warming has been settled once and for all, and the climate change skeptics (themselves) have won. Challenging the existing scientific consensus, they ask us,

“How many times have we heard from Al Gore and assorted European politicians that ‘the science is settled’ on global warming? In other words, it's ‘time for action.’ Climate change is, as recently stated by Hans Blix, former U.N. Chief for weapons detection in Iraq, the most important issue of our time, far more dangerous than people flying fuel-laden aircraft into skyscrapers or threatening to detonate backpack nukes in Baltimore Harbor.

Well, the science may now be settled, but not in the way Gore and Blix would have us believe. Three bombshell papers have just hit the refereed literature that knock the stuffing out of Blix's position and that of the United Nations and its Intergovernmental Panel on Climate Change (IPCC).”

(Douglass et al., 2004c)

The first 2 “bombshell” papers they are referring to are the ones mentioned above (Douglass et al., 2004; 2004b). After this dramatic and inflammatory introduction, they go on to announce what they believe their papers have accomplished, and compare these results with the consensus view.

“The surface temperature record shows a warming rate of about 0.17˚C (0.31˚F) per decade since 1979. However, there are two other records, one from satellites, and one from weather balloons that tell a different story. Neither annual satellite nor balloon trends differ significantly from zero since the start of the satellite record in 1979. These records reflect temperatures in what is called the lower atmosphere, or the region between roughly 5,000 and 30,000 feet....

So, which record is right, the U.N. surface record showing the larger warming or the other two? There's another record, from seven feet above the ground, derived from balloon data that has recently been released by the National Oceanic and Atmospheric Administration. In two research papers in the July 9 issue of Geophysical Research Letters, two of us (Douglass and Singer) compared it for correspondence with the surface record and the lower atmosphere histories. The odd-record-out turns out to be the U.N.'s hot surface history. 

This is a double kill, both on the U.N.'s temperature records and its vaunted climate models. That's because the models generally predict an increased warming rate with height (outside of local polar regions). Neither the satellite nor the balloon records can find it. When this was noted in the first satellite paper published in 1990, some scientists objected that the record, which began in 1979, was too short. Now we have a quarter-century of concurrent balloon and satellite data, both screaming that the UN's climate models have failed, as well as indicating that its surface record is simply too hot.”

(Douglass et al., 2004c)

A closer examination of these “bombshell” papers reveals a very different story. In the introduction to the first, DEA tell us that,

“The globally averaged surface temperature (ST) trend over the last 25 years is 0.171 K/decade [Jones et al., 2001], while the trend in the lower troposphere from observations made by satellites and radiosondes is significantly less, with exact values depending on both the choice of dataset and analysis methodology [e.g., Christy et al., 2003, Lanzante et al., 2003]. This disparity was of sufficient concern for the National Research Council (NRC) to convene a panel of experts that studied the “[a]pparently conflicting surface and upper air temperature trends” and concluded, after considering various possible systematic errors, that “[a] substantial disparity remains”[National Research Council, 2000]. The implication of this conclusion is that the temperature of the surface and the temperature of the air above the surface are changing at different rates due to some unknown mechanism.

A number of studies have suggested explanations for the disparity. Lindzen and Giannitsis [2002] have ascribed the disparity to a time delay in the warming of the oceans following the rapid temperature increase in the late 1970s. Hegerl and Wallace [2002] have concluded that the disparity is not due to El Nino or cold-ocean-warm-land effects. Other authors [Santer et al., 2000] have suggested that the disparity is not real but due to the disturbing effects of El Niño and volcanic eruptions, a conclusion that has been critiqued by Michaels and Knappenberger [2000]. Still others argue that the disparity results from the methodology used to prepare the satellite data [Fu et al., 2004, Vinnikov and Grody, 2003]; however, only the results from Christy et al. [2000] have been independently confirmed by weather- balloon data [Christy et al., 2000, Christy et al., 2003, Lanzante et al., 2003, Christy and Norris, 2004].”

(Douglass et al., 2004)

So the trend from satellite and radiosonde products is “significantly less, with exact values depending on both the choice of dataset and analysis methodology”. In fact, only UAH Products are significantly less, and while it is true that the exact values do depend on dataset choice, DEA’s wording makes it sound as if the differences are little more than minor adjustments. In fact, the variations are considerable. Notice also that after commenting on the potential impact of methodology on the MSU record, they carefully avoid any mention of RSS Version 1.0 (Mears et al., 2003; 2003b; 2003c) which is not only one of the best characterized MSU products in existence, but one that agrees well with state-of-the-art AOGCM predictions. Instead, they cite Vinnikov and Grody (2003) whose analysis is quite unique, has many open questions yet to be resolved, and also, coincidentally, has the largest trend predictions of any MSU product by a factor of at least 2. They also cite Fu et al. (2004) regarding satellite dataset analysis method differences when in fact, Fu’s team made no statements about this. What they did was demonstrated that the MSU Channel 2 data were contaminated by stratospheric emissions and quantified the degree of this effect.

Then, after this rather deft diversion, the 2 sources they do cite for the MSU and radiosonde records just happen to be the ones that are closest in agreement for the period 1979 to 1997 and low in trend – UAH Version 5.0 truncated to 1997 (Christy et al., 2003) and the LKS radiosonde product (Lanzante et al., 2003). Figure 11 shows tropospheric temperature trends for UAH, RSS, LKS and HadRT2.1 for 3 layers and by global region (Seidel et al., 2004). It can be seen that there is very good agreement between LKS and both UAH products for MSU2, though regionally the confidence intervals are large enough to accommodate RSS outside of the southern hemisphere for which the UAH-RSS discrepancy is largest. So not surprisingly, the southern hemisphere contributes most to the discrepancy. HadRT2.1 shows significant disagreement with both. For MSU2LT however, the LKS dataset shows noticeable discrepancies with UAH products, but agreement with HadRT2.1 is improved. In this case the largest regional discrepancy is again with the southern hemisphere, where now LKS shows more warming. This is particularly significant, as it is in this region that we expect the 2LT product to be most impacted by Antarctic sea-ice and summer melt pools (Swanson, 2003). Thus, even though UAH median trend estimates tend to be closer to comparably adjusted radiosonde products, agreement varies significantly by layer and region, and confidence intervals tend to be large.

When we extend the record another 4 years the picture changes yet again. Figure 12 shows the same troposphere temperature trends by layer and region as Figure 11, but for 1979-2001. In this case, UAH and RSS products are compared with HadRT2.1 (the LKS record ends in 1997). Now we see that both UAH and RSS products are in relatively good agreement with each other, and both disagree with HadRT2.1 globally and in all regions except the southern hemisphere, where UAH products are closer the HadRT2.1 than RSS. For the MSU2LT layer, both UAH products and HadRT2.1 agree well, but the confidence intervals for each are as large as the trends being measured (Seidel et al., 2004). It is worth noting that until 1997, LKS trends in all regions and globally were consistently warmer than their HadRT2.1 counterparts. Given the 1997-98 El Nino and its impact on all trends, it would have been surprising if this had not continued if LKS had been extended to 2001.

Once again, we see that agreement depends on layer and region, and confidence intervals tend to be large in comparison to the trends being measured. This is particularly true of the 2LT layer that is of most interest to DEA. Furthermore, which layer is in agreement and to what degree appears to be strongly driven by the length of record being examined. Note also that DEA carefully avoid any discussion of the issue of limited radiosonde coverage, particularly in regions such as the southern oceans that have the most impact on differences between UAH and RSS trends. They also avoid any discussion of Antarctic sea-ice and melt pool impacts which will be of particular importance for the lower troposphere 2LT trends that they are most concerned with. It is difficult to see how any of this adds up to a “confirmation” of UAH MSU products at the expense of its competitors! All this is neatly obscured by DEA’s subtle language, limited citations, and the ensuing, rather pensive discussion of a few exotic theories that might explain the ”discrepancy”.

But the selectivity doesn’t end there. To demonstrate their irreconcilable “disparity” DEA chose a common time frame over which to compare the three datasets. Their choice was based on the datasets they had selected and the point they were trying to prove. Introducing their methods, they state that,

“Since we wish to examine the disparity in the temperature trends among these three datasets, we limit our analysis to a common observational time series. The starting point in our analysis will be 1979, which is the beginning year in both the R2-2m and MSU data. We truncate the analysis at December 1996 which avoids the snow cover issue in R2-2m. This also avoids the anomalously large 1997 El Nino event in the tropical Pacific which Douglass and Clader [2002] showed can severely affect the trend-line. We will show later in this paper that it is likely that our conclusions would change little had we been able to use data though 2003.”

(Douglass et al., 2004)

In other words, even though the extant MSU records from both UAH and RSS extend to the present, DEA purposely choose to truncate their analysis to omit nearly a third of it! The stated reasons for doing this are to exclude a known issue with snow cover contamination in R2-2m and the “anomalously large” ENSO event of 1997, but these arguments are unconvincing. There were at least 4 other ENSO events during the satellite era (1982-83, 1986-87, 1991-92 and 1994-95). The 1982-83 event was one of the largest of the 20th century and occurred during the tropospheric/stratospheric impact of the El Chicon eruption (see Figures 6-8). These were not omitted even though the 1982-83 event was almost as large as the 1997 event. Furthermore, there is at least some evidence that a relationship may exist between global warming and ENSO events, particularly their frequency (Meehl and Washington, 1996; Knutson et al., 1997; Timmermann et al., 1999; Collins, 2000). Though the jury is still out on this (Zhang et al., 1997; Knutson et al., 1997; Boer et al., 2000), there is enough evidence of a possible relationship between the two that we certainly cannot avoid them prima facie in upper-air climate change studies! Likewise, avoiding the snow cover issue is also unconvincing as the MSU2LT record is impacted by this as well, particularly in those regions where UAH and RSS products differ significantly (Swanson, 2003). Even if neither of these things were an issue, we are still left with an analysis of only 2/3 of the relevant upper-air record being used to evaluate products that cover the entire period.

The truncation of DEA’s analysis period raises another point. DEA specifically compare lower troposphere trends as determined by UAH MSU2LT products with surface and upper-air trends from other records. The online community encyclopedia Wikipedia ( has a section that discusses the MSU record that includes a table that shows these trends as a function of the record ending year from 1992 through 2003 (Wikipedia, 2004). A check of this table reveals that the year DEA decided to truncate their analysis, 1996, just happens to be the last year for which UAH 2LT products show a negative lower troposphere temperature trend. How convenient! Had they used data up to the present they would have observed a warming trend that in the last several years has been moving more or less steadily in the direction of restoring long-term agreement with the surface record. By truncating their analysis to 1996 they have,

  • Omitted a full third of the MSU record and including only that portion of it for which a negative lower troposphere temperature trend can be derived. Longer 2LT records show warming trends that are moving in the direction of restoring long-term agreement with the surface record.
  • Allowed themselves to directly compare the UAH MSU2 record with the one record which is truly independent of MSU products and shows the best agreement with UAH for that period, LKS (Lanzante et al., 2003). The LKS record does not extend beyond 1997.

  • Top

    Page:   << Previous    1    2    3    4    5    6    7    8    9    10    11    12    13    14    15    16    17    18    19    20    21    22       Next >>
Global Warming Skeptics
Climate Astroturfing
OISM Petition Project
Leipzig Declarations
Climate Denial 101
Christianity & the Environment
Climate Change
The Web of Life
Managing Our Impact
Caring for our Communities
The Far-Right
Ted Williams Archive