One of the more exotic issues that seems to fade in and out of climate reporting is that of the missing “hot spot” in the tropics. A few less scrupulous onlookers have even used it as a prop in an effort to undermine the very foundation of human-caused global warming and the greenhouse effect right along with it [1
]. As we will see, even if the discrepancy is real (and unrelated to measurement uncertainty), it says next to nothing about greenhouse warming in general and the anthropogenic link in particular.
Before diving into the specifics of the hot spot, a brief word is in order about uncertainty. In climate science, uncertainty can be assigned either to the observational data or to the models (the latter of which are based on our best understanding of climate physics). Models aim to match observations, but you can have uncertainty on both ends. And the level of uncertainty depends on the specific variable under consideration.
For example, cloud effects and ocean variability represent key uncertainties in our climate models, while tropospheric temperature in the tropics is an area of uncertainty in our observations. When a mismatch arises, we should be cautious in assigning blame prematurely. The error may lie with the physics encoded in the models, but we also need to look to the observational side of the equation as a possible source of inaccuracy.
The Tropical Lapse Rate Discrepancy
The graphic below illustrates a well-known relationship between temperature and altitude. In the lower atmosphere—or troposphere—the air generally cools as you move away from the planet’s surface, until you get above the tropopause and into the stratosphere, at which point the relationship reverses. The rate of cooling in the lower atmosphere is known as the (positive) lapse rate. One atmospheric constituent that regulates the lapse rate is water vapor. Its role as a greenhouse gas allows the water molecules to absorb heat in the infrared released at the surface. It follows that regions with more moisture in the air should have a reduced lapse rate relative to regions with less moisture.
Tropospheric and stratospheric temperature are examples of positive and negative lapse rates, respectively.
This is where the infamous “hot spot”, otherwise called the tropical lapse rate discrepancy
, comes in. Fundamental physics tells us that a warming planet should not only bring higher surface and air temperatures, but a concomitant increase in water vapor (because warmer air holds more moisture
). The tropics have a unique physical profile compared with the polar regions and other extratropical latitudes in that you have more evaporation and more moisture. In NASA’s time series video
below, you can see the band of water vapor that concentrates tightly around the equator, with some seasonal variation.
We should thus expect a lower rate of cooling above the tropics: due to its surplus of moisture, the air there should show pronounced warming in response to a warming trend at the surface. This expectation of amplified warming in the tropics is what has come to be known as the “hot spot” in climate science. Gavin Schmidt of NASA puts it this way
“The increase in water vapour as surface air temperature rises causes a change in the moist-adiabatic lapse rate (the decrease of temperature with height) such that the surface to mid-tropospheric gradient decreases with increasing temperature (i.e. it warms faster aloft).”
Steve Sherwood of University of New South Wales echoes this physical relationship built into our climate models:
“The troposphere is expected to warm at roughly the same rate as the surface. In the tropics, simple thermodynamics (as covered in many undergraduate meteorology courses) dictates that it should actually warm faster, up to about 1.8 times faster by the time you get to 12 km or so; at higher latitudes this ratio is affected by other factors and decreases, but does not fall very far below 1. These theoretical expectations are echoed by all numerical climate models regardless of whether the surface temperature changes as part of a natural fluctuation, increased solar heating, or increased opacity of greenhouse gases.”
As we can see, the predicted phenomenon in the tropics is not linked exclusively to the release of fossil fuels. The high evaporative activity in the upper-air tropics serves as an accelerant for the generation of water vapor, amplifying the warming there (and adjusting the lapse rate downward) following any influx of heat at the surface. This additional heat can come from increased solar radiation, a reduction in aerosols, El Niño, etc. So a fall in the lapse rate, and thus the equatorial hot spot, was never considered a uniquely greenhouse signature to begin with. As Gavin Schmidt emphasizes: “It bears stating again that the expected amplification has nothing to do with the greenhouse effect – it is just a function of the surface warming.”
Resolving Observational Uncertainty
So the question now is whether this hot spot built into all of our climate models is reflected in the observational record. The answer, at least initially, was not exactly, the reason being that the observational record is not exactly perfect. Climate science veterans will be familiar with the many headaches involved with reconciling the atmospheric trend data with surface trend data.
Early on we had a more glaring discrepancy in that the tropospheric data were showing a cooling trend, not just at the tropics but at all latitudes. Thanks to greater collaboration among the scientists working with the data, the spurious cooling trends were chalked up to a combination of satellite drift, daytime heating and a failure to properly account for these factors when compiling data from the various sensor and hardware redesigns. Necessary adjustments were made to the processing, and the variance is now isolated to the rate of tropospheric warming and trends in the tropics.
Unbeknownst to researchers in the mid-to-late 20th century, capturing temperature trends in the troposphere is a trickier undertaking compared with collecting in situ measurements at the surface. Atmospheric temperature data is derived from two different sources: radiosondes—measuring devices carried high into the air by weather balloons—and satellite-based instruments, with the former record beginning in 1959 and the latter in 1978.
Like any device used for climate study, both are susceptible to bias, and to a much greater extent than their land-based counterparts. Unlike thermometer readings taken on the ground, upper-air instruments do not measure temperature directly. Instead, temperature is backed into by sampling the radiance of Earth at infrared and microwave wavelengths, which is then run through an inversion algorithm. Along with the fact that algorithms vary, instruments aloft must contend with a slew of other uncontrolled effects that complicate accurate sampling. Some of these have already been mentioned, such as satellite drift from orbital decay and solar heating in daylight. Measurement uncertainty can also arise when ice from rain clouds gloms onto the temperature sensor, from poor spatial sampling of the radiosonde network in the tropics, and from cross-calibration among different equipment.
Consequently, even instruments of the same “breed” tend to differ in their readings. For example, the two leading datasets for upper-air temperatures—UAH and RSS—produce incongruous trend data, with neither dataset correlating completely with surface trends. This isn’t too surprising once we consider how much this equipment has changed over the years. The satellite-based instruments have gone through multiple iterations, each of which introduces new biases and peculiarities. (RSS alone has gone through 9 different iterations of their microwave sounding units (MSUs) and at least 8 versions of their AMSUs.) All of these data must ultimately be merged in order to derive trend data. Failing to correctly intercalibrate (or homogenize) the composite of outputs will result in a messy apples-to-oranges scenario and render the data unusable. As one can imagine, removing all of these systematic biases can be a logistical nightmare.
Such difficulties are nothing new. And considerable time and effort have been spent working out how to correct for them. While significant uncertainty still surrounds the satellite record, the fact that the two separate datasets are in disagreement with each other suggests instrument error is to blame and not faulty physics or surface measurements. A second indicator that the discrepancy is at least in part attributable to instrumental bias is that we see the same trend in both the troposphere and stratosphere, the opposite of what we should expect, as Gavin Schmidt notes here:
“The differences from the expected profile are all towards cooling – and this leads to even more cooling in the stratosphere than the models predict as well as cooling in the troposphere (the bias most often remarked upon). The fact that the bias is the same sign throughout the column (despite the very different physics in each region) is a clue that this is unlikely to be real.”
What about the hot spot? After all, even if the upper-air data are inconsistent with the surface data, we should still see evidence of a diminishing lapse rate in whatever data we have. Looking at seasonal and annual time scales, both radiosonde and satellite-based networks show the hot spot. Yet when we zoom out for the longer, multidecadal trends, we have seen some deviation from the models and physics-based expectations. Given the aforementioned heterogeneity of these networks since inception, this result is hardly enough to send climate scientists back to the drawing board. In fact, these less than ideal approximations help explain why theoretical expectations only diverge where the observational data is least reliable.
While we can’t go back in time and reconfigure all of the hardware for smoother comparison, science doesn’t stand still, and more recent reanalyses have brought the observational record into better alignment with global climate models. Notably, Sherwood et al’s May 2015 study has resolved the bias in the radiosonde trend data and confirmed the hot spot using a mix of Kriging and linear regression techniques. This study confirms that the troposphere is warming as expected and is in line with model predictions. For reference purposes, I will provide a list of the key studies in chronological order below:
—Contribution of stratospheric cooling to satellite-inferred tropospheric temperature trends (Fu, et al. 2004):
Here we show that trends in MSU channel 2 temperatures are weak because the instrument partly records stratospheric temperatures whose large cooling trend offsets the contributions of tropospheric warming…The resulting trend of reconstructed tropospheric temperatures from satellite data is physically consistent with the observed surface temperature trend. For the tropics, the tropospheric warming is ~1.6 times the surface warming, as expected for a moist adiabatic lapse rate.
—Satellite-derived vertical dependence of tropical tropospheric temperature trends (Fu, Johansen, 2005):
Our retrievals applied to satellite-observed MSU time series compiled by the RSS team for 1987–2003 demonstrate that the tropical troposphere is warming faster than the surface, and that tropical tropospheric temperature trends increase with height, which confirms the GCM predictions…We show that the T2LT trend bias can be largely attributed to the periods when the satellites had large local equator crossing time drifts, causing both large changes in the calibration target temperatures and large diurnal drifts.
—Radiosonde Daytime Biases and Late-20th Century Warming (Sherwood, et al. 2005):
The temperature difference between adjacent 0000 and 1200 UTC weather balloon (radiosonde) reports shows a pervasive tendency toward cooler daytime compared to nighttime observations since the 1970s, especially at tropical stations. Several characteristics of this trend indicate that it is an artifact of systematic reductions over time in the uncorrected error due to daytime solar heating of the instrument and should be absent from accurate climate records. Although other problems may exist, this effect alone is of sufficient magnitude to reconcile radiosonde tropospheric temperature trends and surface trends during the late 20th century.
—The Effect of Diurnal Correction on Satellite-Derived Lower Tropospheric Temperature (Mears, Wentz, 2005):
Satellite-based measurements of decadal-scale temperature change in the lower troposphere have indicated cooling relative to Earth’s surface in the tropics. Such measurements need a diurnal correction to prevent drifts in the satellites’ measurement time from causing spurious trends. We have derived a diurnal correction that, in the tropics, is of the opposite sign from that previously applied. When we use this correction in the calculation of lower tropospheric temperature from satellite microwave measurements, we find tropical warming consistent with that found at the surface and in our satellite-derived version of middle/upper tropospheric temperature.
—Amplification of Surface Temperature Trends and Variability in the Tropical Atmosphere (Santer, et al. 2005):
The month-to-month variability of tropical temperatures is larger in the troposphere than at Earth’s surface. This amplification behavior is similar in a range of observations and climate model simulations and is consistent with basic theory. On multidecadal time scales, tropospheric amplification of surface warming is a robust feature of model simulations, but it occurs in only one observational data set. Other observations show weak, or even negative, amplification. These results suggest either that different physical mechanisms control amplification processes on monthly and decadal time scales, and models fail to capture such behavior; or (more plausibly) that residual errors in several observational data sets used here affect their representation of long-term trends.
—Biases in Stratospheric and Tropospheric Temperature Trends Derived from Historical Radiosonde Data (Randel, Wu, 2006):
Detailed comparisons of one radiosonde dataset with collocated satellite measurements from the Microwave Sounding Unit reveal time series differences that occur as step functions or jumps at many stations…The fact that the jumps occur at different times for different stations suggests that the problems originate primarily with the radiosondes rather than the satellite data…The net effect of the jumps is a systematic tendency for spurious cooling in the radiosonde data at each of the identified stations….As a result of these jumps, the radiosondes exhibit systematic cooling biases relative to the satellites. A large number of the radiosonde stations in the Tropics are influenced by these biases…Comparison of trends from stations with larger and smaller biases suggests the cooling bias extends into the tropical upper troposphere. Significant biases are observed in both daytime and nighttime radiosonde measurements.
—The Answer is Blowing in the Wind (Thorne, 2008); paper: Warming maximum in the tropical upper troposphere deduced from thermal winds (Allen, Sherwood, 2008):
The uncertainty with respect to upper air temperature estimates in the tropics is so substantial that we can draw no meaningful conclusions as to whether or not there is a discrepancy between long-term trends in the real world and our expectations from climate models…In order to gauge upper air temperature change in the tropics in a fundamentally different way, Allen and Sherwood exploit the thermal wind relationship, in which vertical gradients in wind are linked to horizontal gradients in temperature…Allen and Sherwood use radiosonde-derived wind data to reveal that temperatures in the tropical upper troposphere are very likely to be increasing as global surface temperatures rise…The new analysis adds to the growing body of evidence suggesting that these discrepancies are most likely the result of inaccuracies in the observed temperature record rather than fundamental model errors.
—Toward Elimination of the Warm Bias in Historic Radiosonde Temperature Records—Some New Results from a Comprehensive Intercomparison of Upper-Air Data (Haimberger, et al. 2008):
Both of the new adjusted radiosonde time series are in better agreement with satellite data than comparable published radiosonde datasets…A robust warming maximum of 0.2–0.3K (10 yr)−1 for the 1979–2006 period in the tropical upper troposphere could be found in both homogenized radiosonde datasets…Both RAOBCORE, version 1.4, and RICH data show a robust upper-tropospheric warming maximum in the tropics. Therefore, both datasets support the arguments of Santer et al. (2005) and Thorne et al. (2007) that the apparent inconsistency in the vertical profile of tropical temperature trends between earlier homogenized radiosonde datasets (HadAT2; RATPAC) and satellite temperature products is to a large fraction caused by residual biases in these radiosonde observation time series…In the tropical upper troposphere, where the predicted amplification of surface trends is largest, there is no significant discrepancy between trends from RICH–RAOBCORE version 1.4 and the range of temperature trends from climate models.
—Robust Tropospheric Warming Revealed by Iteratively Homogenized Radiosonde Data (Sherwood, et al. 2008):
While this effort appears not to have detected all artifacts, trends appear to be systematically improved. Stronger warming is shown in the NH where sampling is best. These results support the hypothesis that trends in wind data are relatively uncorrupted by artifacts compared to temperature, and should be exploited in future homogenization efforts… Despite this, the adjusted tropospheric temperature trends agree roughly with physical expectations.
—Consistency of modelled and observed temperature trends in the tropical troposphere (Santer, et al. 2008):
We find that there is no longer a serious discrepancy between modelled and observed trends in tropical lapse rates. This emerging reconciliation of models and observations has two primary explanations. First, because of changes in the treatment of buoy and satellite information, new surface temperature datasets yield slightly reduced tropical warming relative to earlier versions. Second, recently developed satellite and radiosonde datasets show larger warming of the tropical lower troposphere.
—Critically Reassessing Tropospheric Temperature Trends from Radiosondes Using Realistic Validation Experiments (Titchner, et al. 2009):
The homogenization system consistently reduces the bias in the daytime tropical, global, and Northern Hemisphere (NH) extratropical trends but underestimates the full magnitude of the bias. Southern Hemisphere (SH) extratropical and all nighttime trends were less well adjusted owing to the sparsity of stations…The implications are that tropical tropospheric trends in the unadjusted daytime radiosonde observations, and in many current upper-air datasets, are biased cold, but the degree of this bias cannot be robustly quantified. Therefore, remaining biases in the radiosonde temperature record may account for the apparent tropical lapse rate discrepancy between radiosonde data and climate models.
—Atmospheric temperature change detection with GPS radio occultation 1995 to 2008 (Steiner, et al. 2009):
Existing upper air records of radiosonde and operational satellite data recently showed a reconciliation of temperature trends but structural uncertainties remain…The observed trends and warming/cooling contrast across the tropopause agree well with radiosonde data and basically with climate model simulations, the latter tentatively showing less contrast.
—Changes in the sea surface temperature threshold for tropical convection (Johnson, Xie, 2010):
We conclude that, in contrast with some observational indications, the tropical troposphere has warmed in a way that is consistent with moist-adiabatic adjustment, in agreement with global climate model simulations.
—Tropospheric Temperature Trends: History of an Ongoing Controversy (Thorne, et al. 2010):
Particular focus is given to the difficulty of producing homogenized datasets, with which to derive trends, from both radiosonde and satellite observing systems, because of the many systematic changes over time…It is concluded that there is no reasonable evidence of a fundamental disagreement between tropospheric temperature trends from models and observations when uncertainties in both are treated comprehensively.
—Revisiting the controversial issue of tropical tropospheric temperature trends (Mitchell, et al. 2013):
Using these approaches, it is shown that within observational uncertainty, the 5–95 percentile range of temperature trends from both coupled-ocean and atmosphere-only models are consistent with the analyzed observations at all but the upper most tropospheric level (150 hPa), and models with ultra-high horizontal resolution (≤ 0.5° × 0.5°) perform particularly well. Other than model resolution, it is hypothesized that this remaining discrepancy could be due to a poor representation of stratospheric ozone or remaining observational uncertainty.
—Atmospheric changes through 2012 as shown by iteratively homogenized radiosonde temperature and wind data (Sherwood, Nishant, 2015):
We present an updated version of the radiosonde dataset homogenized by Iterative Universal Kriging (IUKv2)…This method, in effect, performs a multiple linear regression of the data onto a structural model that includes both natural variability, trends, and time-changing instrument biases, thereby avoiding estimation biases inherent in traditional homogenization methods…Temperature trends in the updated data show three noteworthy features. First, tropical warming is equally strong over both the 1959–2012 and 1979–2012 periods, increasing smoothly and almost moist-adiabatically from the surface (where it is roughly 0.14 K/decade) to 300 hPa (where it is about 0.25 K/decade over both periods), a pattern very close to that in climate model predictions. This contradicts suggestions that atmospheric warming has slowed in recent decades or that it has not kept up with that at the surface. Second, as shown in previous studies, tropospheric warming does not reach quite as high in the tropics and subtropics as predicted in typical models.
From the above we can see that the hot spot imbroglio is yet another example of excitable contrarians missing the forest for the trees. Since the models are physics-based (and very well-established physics at that), it is overwhelmingly more likely that the physics is right and the measurements are not. Efforts to properly aggregate the discontinuous data sets have affirmed this basic picture. And contrary to claims by partisan media, a missing hot spot would not seal the fate on anthropogenic warming.
In terms of the human fingerprint, there are several discernible signals, including the parallel effect of tropospheric warming and stratospheric cooling, the escalating C-12 ratio in the atmosphere, a rising tropopause, reduced atmospheric oxygen levels, ocean acidification, and the to-date release of fossil carbon, all of which powerfully confirm that we are moving massive amounts of carbon from the “slow” domain into the “fast” domain, kicking the planet’s natural systems into overdrive.
—Tropical tropospheric temperature, instrumental and proxy trends via The Way Things Break
—My lay-friendly primer on the science of global warming: A Climate of Change, also available for download at my research page, in addition to the current post.)