Climate Change & Tropospheric Temperature Trends

Current Revision Level

Rev. 1.3:    Jan. 21, 2009

Acknowledgements

I would like to thank the following for taking the time out of their already busy schedules to offer badly needed comments and suggestions regarding the content of this paper. Without their contributions, it would not have been possible. Thank you!

Dian Seidel    (NOAA Air Resources Laboratory, Silver Spring, MD)
Kevin Trenberth    (National Center for Atmospheric Research / Climate and Global Dynamics, Boulder, CO)
Jerry Mahlmann    (National Center for Atmospheric Research / Climate and Global Dynamics, Boulder, CO)
Rasmus Benestad    (Norwegian Meteorological Institute, Oslo, Norway; Contributing Editor for www.RealClimate.com)
Gavin Schmidt    (Goddard Institute for Space Studies, New York, NY; Contributing Editor for www.RealClimate.com)
William Connolley    (British Antarctic Survey, Cambridge, U.K.; Contributing Editor for www.RealClimate.com)
David Parker    (U.K. Met Office, Bracknell, Berkshire, U.K.)

Introduction

Shortly after the dawn of the nuclear age, Albert Einstein observed that for the first time in human history, the extinction of our own race and the destruction of the biosphere has been brought within the realm of technical possibilities. He was of course, referring to the advent nuclear weapons and their proliferation. But in the years since, his words have proven to be more prophetic than he imagined. The industrial revolution has given us the ability to globally alter the very processes that sustain all life on our home planet, including our own. With each passing day it seems, we are learning of new and unforeseen long-term changes that our activities are imposing on the biosphere, and increasingly, these are proving to be damaging. For the first time in our history, we are tinkering with the very umbilical cord that supports us, and doing so in ways that we do not fully understand and that many of us do not want to address. Nowhere is this more evident than in regard to global warming – or more appropriately, climate change.

Climate change is in many ways unlike any other dilemma the human race has ever faced. First, to a greater extent than any other environmental problem, it is truly global in nature. In this sense, global warming is not a misnomer. Though there is much debate about how severe its impacts will be and how they will play out in various global regions, few scientists today believe that the consequences will even be acceptable, much less positive – and they will be borne by all of us. Second, unlike other forms of pollution or environmental degradation, the response times associated with climate change are large compared with the human activities that are forcing them. It is often forgotten that climate change is not an atmospheric response to natural and/or anthropogenic forcing. The biosphere is a coupled system – atmosphere, ocean, and continental land – and it is as a system that it responds to being forced. The latent heat retention and transport capability of the world’s oceans is huge with respect to its atmosphere and continental land masses. As a result, when “kicked” it will respond much like a bowling ball being pulled by a slinky - it will be some time before we actually see motion, and once the ball is rolling it will take an equal or longer time to stop ¹. This presents an utterly unique and historically unprecedented moral dilemma. Significant sectors of the global economy have been built upon greenhouse gas emitting technologies and land use practices that are having global consequences. These however, are concentrated mainly in the developed world where decades, and even centuries of growth have inculcated a worldview that demand lifestyles dependent on them. At the time of this writing for instance, the United States comprises roughly 5 percent of the world’s population, yet it consumes over a third of its natural resources and generates some 20 percent of its pollution – including greenhouse gases. Few Americans do not have at least one automobile, and fewer still drive hybrids, or other alternatives that would mitigate some of the impact. The cost of shifting our economies away from these practices is will be significant – lifestyle changes, major economic adjustments, and though few want to admit it, a rethinking of the natural resource consumption levels of the world’s richest nations, their impact on its poor nations, and the roadblock this presents to allowing all nations and peoples to achieve a happy and productive existence. Thus, the costs of mitigating climate change impacts are likely to be enormous, and the burden of responsibility will fall mainly on those who are contributing most to the problem – the world’s developed nations, their industries that have created unprecedented wealth from greenhouse gas emitting technologies, and the First World consumer base that has grown dependent on them. Yet to avoid the most damaging impacts, these sacrifices will need to be made early rather than late in the process, and the payoff will not be obvious for decades to come. Delaying action until human cause global change is unavoidably obvious will almost certainly be too late.

It is difficult to imagine an environmental or moral dilemma tailored more perfectly to the most primal of human weaknesses – denial, rationalization, the passing of blame to others. So it should come as no surprise that the last 10 to 15 years have seen an unprecedented growth in challenges to the mainstream scientific consensus on global warming. Nor should it surprise us that almost without exception, these have come not from the mainstream scientific community, but from polluting and extraction industries and ultra-conservative special interests (who typically have worldviews rooted in free market values and the sanctity of business interests). Numerous front groups and think tanks funded by these interests have appeared in recent years seeking to disprove global warming, or at least to divert public policy away from mitigation efforts. Typically, these groups employ scientific consultants, all of which are drawn from the same pool of one to two dozen scientists who are well known for their contrarian views. None of them do original peer-reviewed research in climate science or any other field, and apart from one or two notable exceptions, little has been published by their consultants that seriously challenge the current consensus. Most of their efforts have been devoted to extensive public relations programs, ad campaigns, “educational” forums, and lobbying efforts at the state and federal level, where they have enjoyed wide support from the Bush administration. The lion’s share of their funding has come from industry coalitions (chiefly the fossil fuel, auto, mining, and coal fired power industries), and from ultra-conservative foundations and religious groups such as the John Mellon and Sarah Scaife Foundations, the Olin Foundation, the Coors empire, the Unification Church (the “Moonies”), and many others (Beder, 1998; 1999; Gelbspan, 1998; 2004).

So how is this relevant to troposphere temperatures? Given the maturity of surface station records of the last century, most global warming skeptics will now admit that the earth’s surface has warmed in recent years to at least some extent. Their main points of contention are a) that this warming has an anthropogenic component, and b) that the consequences of future greenhouse gas emissions are likely to be severe enough to justify mitigation efforts. If the observed warming is natural and not historically unprecedented, there is little we can do about it. If the likely impacts of warming over the next century are relatively mild (or even beneficial as some have argued), there is nothing to worry about. Either way, costly mitigation efforts and a technological shift away from fossil fuels will be unnecessary – a position that has obvious appeal to industry. Ultimately, this boils down to determining how much of the observed warming is due to anthropogenic greenhouse gases and land use, and what consequences can be expected from the status quo during the next century.

AOGCM’s have played a key role in the search for answers to these questions. The IPCC (2001) conducted a review of the best of these models, as evaluated by the Coupled Model Intercomparison Project, or CMIP (Meehl et al. 2000). On overview of how these models performed when compared against surface air temperature (Jones et al., 1999), precipitation (Xie and Arkin, 1996), and sea level pressure (ERA-15 reanalysis) is given collectively in Figures 15 and 16. All evaluated models were forced with a combination of natural and anthropogenic forcings, including greenhouse gas emissions. Volcanic and ENSO effects are included in most. Figure 41 is a Taylor diagram in which variance and correlation are plotted radially. The point marked “Observed” corresponds to observation. The further any given point is from the Observed point, the larger its overall RMS error in whatever variable is being measured. It can be seen that there have been issues with how well these models have replicated global precipitation and sea level pressure. But the best of them have done a good job of replicating the last century’s surface temperature evolution. When anthropogenic greenhouse gas emissions and land use activities are removed from these models, this agreement with observation is lost. This strongly supports both the reliability of these models in reproducing overall global surface temperature trends, and the reality of anthropogenic impacts on climate change. When these models are forced beyond the present and into the next century, all show significant warming that is appreciably curbed (but not removed) only by drastic greenhouse gas reductions.

Skeptic Arguments

Because of their ability to demonstrate anthropogenic impacts on global climate, and their predictions for the future, discrediting AOGCM’s has been a primary objective of global warming skeptics and their benefactors. Most of these models predict a strong surface-troposphere and show the latter warming at least as fast as the former. Their inability to comfortably reproduce the discrepancies that have been observed is an obvious weak point, and considerable effort has been expended by contrarians in attacking it. Selected portions of the MSU and radiosonde records have been the weapons of choice. The bulk of their literature has been published in popular books, press releases and editorials, and online. Most of it is little more than a superficial rehash of the same few claims. As there is a virtual avalanche of these publications today, and one is very much like another, I will not attempt to address all of them but will select a few representative examples that highlight the most common skeptic arguments.

1)   The MSU record is the only reliable satellite analysis.

Skeptics argue that the satellite record is the only one accurate enough to determine global temperature trends. The surface record, they say, is plagued by urban heat island effects, poor data quality, and a host of other issues that are avoided in MSU products. Robinson et al. (1998) ² are fairly typical of the skeptic literature on this point. They argue that,

“Since 1979, lower-tropospheric temperature measurements have also been made by means of microwave sounding units (MSU’s) on orbiting satellites. Figure 6 shows the average global tropospheric satellite measurements – the most reliable measurements, and the most relevant to the question of climate change.”

(Robinson et al., 1998)

The Figure 6 they refer to presents a time series of monthly global troposphere temperature anomalies based on UAH Version C (Christy et al., 1998), though their citation is to a letter in Nature that briefly discusses the accuracy of MSU detectors (Christy and Braswell, 1997). They do not directly cite UAH Version C – which included a discussion of the recently discovered spurious warming introduced by POES orbital decay (Wentz and Schabel, 1998) and specifically notes that a correction for it was not included in that dataset. They go on to say,

“Disregarding uncertainties in surface measurements and giving equal weight to reported atmospheric and surface data and to 10 and 19 year averages, the mean global trend is minus 0.07 ºC per decade. In North America, the atmospheric and surface records partly agree. Even there, however, the atmospheric trend is minus 0.01 per decade, while the surface trend is plus 0.07 ºC per decade. The satellite record, with uniform and better sampling, is much more reliable.”

(Robinson et al., 1998)

We’re told that the MSU record has “uniform and better sampling, [and] is much more reliable”, yet no details are given. We aren’t shown any specifics regarding the problems claimed for either record or any comparisons that might allow us to judge the relative uncertainties in each. Skeptic claims about flaws in the surface record are beyond the scope of this paper and have been dealt with in detail elsewhere (NRC, 2000; IPCC, 2001). As for the MSU record, it is not at all clear that its uncertainties are any less problematic. The one clear advantage of the MSU Record is that it is truly global whereas the surface and radiosonde records are not. Beyond that its uncertainties are legion - impacts of sampling error on the evaluation of diurnal drift (UAH Record) and synthetic Channel signals like 2LT and TLT, imperfect characterization of various instrument non-linearities and calibration issues, the shortened service life of some NOAA POES spacecraft (most notably, NOAA-09), uncertainties in the characterization of Instrument Body Effect, potential signal contamination from surface and stratosphere emissions, and various complications surrounding merge calculation methodologies. These were covered in detail in Part I of this paper. Some of these errors are less severe than others, and efforts to correct for them have resulted in genuinely low confidence intervals compared to the surface and radiosonde records. Others however, are more significant. The differing merge methodologies of RSS and UAH products alone accounts for at least 65 percent of the difference between their trends. When smoothing philosophies are included as well, the difference is larger still. The anomalously large values derived by the UAH team for the NOAA-09 target factor are a particular point of concern. Yet the RSS analysis can comfortable account for the best characterized AOGCM results which shows that it is even possible that differing data reduction philosophies may be able to explain most of the discrepancy between the records (Santer et al., 2003). Then there is the fact that MSU products measure bulk layer temperatures rather than altitude specific ones, and the layers measured imperfectly represent the lower and middle troposphere temperatures that are needed for AOGCM comparisons. In particular, stratospheric noise on Channel 2 accounts for as much as 20 percent of its signal. MSUTLT avoids much of this pollution, but the differencing method it is based on more than doubles its sampling noise compared to Channel 2 (NRC, 2000; Mears et al., 2003c). Furthermore, because the stratosphere is cooling faster than the troposphere is warming, there will be an even larger impact on the trends themselves. It now appears that stratospheric contamination may contribute up to 0.08 deg. K/decade to the Channel 2 trend (Fu et al., 2004). This is nearly half of the expected trend if the surface and upper atmosphere are strongly coupled, and more than half of the observed RSS MSU Channel 2 trend. These results can hardly be considered as less problematic than those of other products.

Though Robinson and his co-authors avoided any specifics regarding issues with MSU datasets, others have attempted to dispel concerns about them. For instance, in 1998 shortly after Wentz and Schabel published their paper on spurious cooling due to POES orbital decay (1998), it came under immediate attack from numerous industry front groups. Typical of the bullets fired was an August 1998 press release from the California based National Center for Public Policy Research in which it was claimed that,

“A study released yesterday suggesting that satellite data showing a drop in the earth's temperature over the past 18 years is wrong is fatally flawed. The study thus has no impact on the ongoing global warming debate.

The study, written by Frank J. Wentz and Matthias Schabel, claims that because NASA's orbiting satellites can lose altitude as they circle the globe, temperature data collected by these satellites has been inaccurate. Wentz and Schabel further suggest that, with these altitude drops factored in, the temperature of the planet has warmed 0.13 degrees per decade rather than declined by 0.09 degrees per decade.”

(NCPPR, Aug. 13, 1998)

NCPPR goes on to argue that Wentz and Schabel ignored “false warming caused by other factors” and concluded that,

"The Wentz/Schabel study is fatally flawed and is thus of little use in the current global warming debate," said David Ridenour, Vice President of The National Center for Public Policy Research. "The bottom line is that satellite data -- which has consistently shown no warming trend -- remains the most reliable means of measuring the earth's temperature. Satellites cover 99% of the surface of the planet. By contrast, reliable ground temperature data over the past 100 years covers just 18.4% of the planet."

(NCPPR, Aug. 13, 1998)

Not surprisingly, the argument has been carefully cherry-picked. The “false warming” factors they are referring to include updated corrections for diurnal drift and IBE. These two effects collectively add up to 0.07 deg. K/decade of spurious cooling in MSU2R (the lower troposphere synthetic channel used in Versions B and C that is analogous to 2LT used in later Versions). The corresponding spurious cooling of MSU2R from POES orbital decay described by Wentz and Schabel was about -0.10 deg. K/decade. The net effect of the two is an increase in observed MSU2R warming from 0.03 deg. K/decade to 0.06 deg. K/decade (Christy et al., 2000). A corresponding loss of warming was observed on MSU2 (from 0.08 deg. K/decade to 0.04 deg. K/decade) but the end result was still a warming troposphere. The confidence interval on these values includes zero, but the most likely result is warming. The statement about the Wentz/Schabel study being “fatally flawed” is flat out incorrect. Both corrections are needed in MSU products. When the NCPPR published this press release UAH Version C was the extant UAH MSU analysis product. Version C notes the orbital decay issue, but did not include it, as the discovery was made after the final pre-publication paper had gone to galley print. Because the corrections for diurnal drift and IBE (i.e. “false warming”) that were included are actually smaller than the orbital decay correction, Wentz and Schabel were actually closer to the truth than UAH Version C. Later versions of UAH and RSS products corrected for both errors, and as we have seen, the evolution in MSU observed trends has been upward ever since. Though UAH Version D had not been published at the time of this press release, all the information necessary for this comparison had been (Wentz and Schabel, 1998; Christy et al., 1998). Ridenour and the NCPPR simply did not do their homework.

Note that the previous statements all implicitly assume that the surface and tropospheric records are measuring the same things. They are not. We have already seen in Part I that there is likely to be at least some decoupling of the two in the tropics and extra-tropics (Trenberth and Stepaniak, 2003). The differing impacts on each from volcanic eruptions and ENSO’s were well known at the time Robinson et. al. and the NCPPR published their statements. Furthermore, though AOGCM’s have assumed a strong degree of surface-troposphere coupling over the long term, the short-term forcings and evolution of surface and tropospheric temperatures are vastly different and may show considerable variation both regionally and temporally. So in regard to evaluations of global warming, there is no justification to giving the two “equal weight” on decadal time scales even in principle. The concluding statement contrasting the global coverage of the MSU record with “reliable ground temperature data” from the last century, which we are told covers “just 18.4% of the planet.” But only a few paragraphs before they stated that,

“The satellite measurements showing no global warming have been corroborated by weather balloon measurements. If this satellite data were significantly off, satellite and weather balloon measurements should have diverged…”

(NCPPR, Aug. 13, 1998)