
“Fearless Felix” Baumgartner ascended to the stratosphere and stepped into the void from 24.2 miles above the Earth. His speed during the fall reached Mach 1.24, and the Austrian adventurer nailed the landing. October 14, 2012 Wired
Introduction
Murry Salby is also totally committed to the atmosphere. He is a scientist with such deep and broad knowledge of atmospheric physics that he has written multiple textbooks on the subject. And yet he is not fearful for the future of our climate system, in contrast to many of his colleagues. By stepping away from “consensus” climate alarms, he has shown unusual courage by speaking plainly about the atmosphere and climate, despite attempts to silence him.
Dr. Salby’s latest textbook is entitled Physics of the Atmosphere and Climate (here). I got a copy and have been reading in it to understand where he comes down on various issues related to climate change. In particular I wanted to know what explains his divergence from IPCC climate scientists.
H/T to Kenneth Richard and No Tricks Zone
Synopsis
In reading the textbook, I find two main reasons why Salby is skeptical of AGW (anthropogenic global warming) alarm. This knowledgeable book is an antidote to myopic and lop-sided understandings of our climate system.
- CO2 Alarm is Myopic: Claiming CO2 causes dangerous global warming is too simplistic. CO2 is but one factor among many other forces and processes interacting to make weather and climate.

Myopia is a failure of perception by focusing on one near thing to the exclusion of the other realities present, thus missing the big picture. For example: “Not seeing the forest for the trees.” AKA “tunnel vision.”
2. CO2 Alarm is Lopsided: CO2 forcing is too small to have the overblown effect claimed for it. Other factors are orders of magnitude larger than the potential of CO2 to influence the climate system.


Lop-sided refers to a failure in judging values, whereby someone lacking in sense of proportion, places great weight on a factor which actually has a minor influence compared to other forces. For example: “Making a mountain out of a mole hill.”
Overview
Salby’s textbook presents all of the physical complexity of the climate system in contrast to simplistic global warming theory. And he provides his sense of the Scales of the various processes, balancing any lopsided overemphasis on CO2 effects.
From the Preface:
Despite technological advances in observing the Earth-atmosphere system and in computing power, strides in predicting its evolution reliably – on climatic time scales and with regional detail – have been limited. The pace of progress reflects the interdisciplinary demands of the subject. Reliable simulation, adequate to reproduce the observed record of climate variation, requires a grasp of mechanisms from different disciplines and of how those mechanisms are interwoven in the Earth-atmosphere system.
What is today labeled climate science includes everything from archeology of the Earth to superficial statistics and a spate of social issues. Yet, many who embrace the label have little more than a veneer of insight into the physical processes that actually control the Earth-atmosphere system, let alone what is necessary to simulate its evolution reliably. Without such insight and its application to resolve major uncertainties, genuine progress is unlikely.
The atmosphere is the heart of the climate system, driven through interaction with the sun, continents, and ocean. It is the one component that is comprehensively observed. For this reason, the atmosphere is the central feature against which climate simulations must ultimately be validated.
The treatment focuses upon physical concepts, which are developed from first principles. It integrates five major themes:
1. Atmospheric Thermodynamics;
2. Hydrostatic Equilibrium and Stability;
3. Radiation, Cloud, and Aerosol;
4. Atmospheric Dynamics and the General Circulation;
5. Interaction with the Ocean and Stratosphere.
Lessons from Dr. Salby: Essential Elements of a Balanced Climate Understanding
Below I show some of the written statements from the textbook to illustrate how his knowledge counteracts myopic and lopsided thinking.
Focusing on CO2 from burning fossil fuels is myopic: A multitude of natural sources drive atmospheric concentrations.

Plate 24 Estimated global carbon cycle, illustrating stores of carbon,
in GtC, and transfers in GtC/yr, where 1 GtC=109 tons of carbon.
Source: Design by Philippe Rekacewicz, UNEP/GRID-Arendal (11.07.10).
Pg.545-6
The storage of CO2 is illustrated in Fig. 17.11. Except for deep sedimentary rock, which is sequestered, most of the carbon is stored in the ocean. It accounts for some 40,000 gigatons (1012 kg) of carbon (GtC), in the form of dissolved CO2 and organic matter. Most resides in the deep ocean, where cold water supports the greatest observed concentrations. There, dissolved CO2 is controlled by the thermohaline circulation. Land and the adjoining biosphere account for only about 2000 GtC. The atmosphere contains less than 1000 GtC, concentrated in CO2. Hence, the store of carbon in the ocean is two orders of magnitude greater than the store in the atmosphere.
Equally significant are transfers of carbon into and out of the ocean. Of order 100 GtC/yr, they exceed those into and out of land. Together, emission from ocean and land sources (∼150 GtC/yr) is two orders of magnitude greater than CO2 emission from combustion of fossil fuel. These natural sources are offset by natural sinks, of comparable strength. However, because they are so much stronger, even a minor imbalance between natural sources and sinks can overshadow the anthropogenic component of CO2 emission (cf Secs 1.6.2, 8.7.1).
The values in Fig. 17.11 can be used to estimate the effective turnover time of atmospheric CO2. At an absorption rate of 100 GtC/yr, the ocean will absorb the atmospheric store of CO2 of 1000 GtC in about a decade. That absorption of CO2, which is concentrated in cold SST at polar latitudes, is nearly offset by emission of CO2 from warm SST at tropical latitudes. Warming of SST (by any mechanism) will increase the outgassing of CO2 while reducing its absorption. Owing to the magnitude of transfers with the ocean, even a minor increase of SST can lead to increased emission of CO2 that rivals other sources (Sec. 8.7.1). Further, if the increase of SST involves heat transfer with the deep ocean, the time for equilibrium to be reestablished would be centuries (Sec. 17.1.2).
Attributing rising temperatures to fossil fuels is lop-sided: Natural sinks respond to warming by releasing CO2 in far greater quantities.

Net emission rate of CO2, r˙ CO2 = d dt rCO2 (ppmv/yr), derived from the Mauna Loa record (Fig. 1.15), lowpass filtered to changes that occur on time scales longer than 2 years (solid). Superimposed is the satellite record of anomalous Global Mean Temperature (Fig. 1.39), lowpass filtered likewise and scaled by 0.225 (dashed). Trend in GMT over 1979–2009 (not included) is ∼0.125 K/decade.
Pg.65ff
Net emission of CO2 closely tracks the evolution of GMT. Achieving a correlation of 0.80, the variation of GMT accounts for most of the variance in CO2 emission.

Plotted in Fig. 1.43b is the rate of change in isotopic composition, d dt δ13C = ˙ δ13C (solid).12 Its mean is negative, consistent with the long-term decline of δ13C in ice cores (Fig. 1.14). However, like emission of CO2, differential emission of 13CO2 varies substantially from one year to the next. It too tracks the evolution of GMT – just out of phase. When GMT increases, emission of 13CO2 decreases and vice versa. The records achieve a correlation of −0.86. Hence the variation of GMT, which accounts for most of the variance in emission of CO2, also accounts for most of the variance in differential emission of 13CO2.
The out-of-phase relationship between rCO2 and δ13C in the instrumental record (Fig. 1.43) is the same one evidenced on longer time scales by ice cores (Fig. 1.14). The out-of-phase relationship in ice cores is regarded as a signature of anthropogenic emission, subject to uncertainties (Sec. 1.2.4). The out-of-phase relationship in the instrumental record, however, is clearly not anthropogenic. Swings of GMT following the eruption of Pinatubo and during the 1997–1998 El Nino were introduced through natural mechanisms (cf. Figs 1.27; 17.19, 17.20). Changes in Fig. 1.43 reveal that net emission of CO2, although 13C lean, is accelerated by increased surface temperature. Outgassing from ocean, which increases with temperature (Sec. 17.3), is consistent with the observed relationship – if the source region has anomalously low δ13C. So is the decomposition of organic matter derived from vegetation. Having δ13C comparable to that of fossil fuel, its decomposition is likewise accelerated by increased surface temperature.
Focusing on CO2 as the greenhouse gas of concern is both myopic and lop-sided: H20 makes 98% of the IR radiative activity in the atmosphere.

Pg. 47
The radiative-equilibrium surface temperature Ts is significantly warmer than that in the absence of an atmosphere.
The discrepancy between Ts and Te follows from the different ways the atmosphere processes SW and LW radiation. Although nearly transparent to SW radiation (wavelengths λ ∼ 0.5 μm), the atmosphere is almost opaque to LW radiation (λ ∼ 10 μm) that is re-emitted by the Earth’s surface. For this reason, SW radiation passes relatively freely to the Earth’s surface, where it can be absorbed. However, LW radiation emitted by the Earth’s surface is captured by the overlying air, chiefly by the major LW absorbers: water vapor and cloud. Energy absorbed in an atmospheric layer is reemitted, half upward and half back downward. The upwelling re-emitted radiation is absorbed again in overlying layers, which subsequently re-emit that energy in similar fashion. This process is repeated until LW energy is eventually radiated beyond all absorbing components of the atmosphere and rejected to space. By inhibiting the transfer of energy from the Earth’s surface, repeated absorption and emission by intermediate layers of the atmosphere traps LW energy, elevating surface temperature over what it would be in the absence of an atmosphere.
Pg. 247ff
The residual, +1.5 Wm−2, represents net warming. It is about 0.5% of the 327 Wm−2 of overall downwelling LW radiation that warms the Earth’s surface (Fig. 1.32). The vast majority of that warming is contributed by water vapor. Together with cloud, it accounts for 98% of the greenhouse effect. How water vapor has changed in relation to changes of the comparatively minor anthropogenic species (Fig. 8.30) is not known. The additional surface warming introduced by anthropogenic increases in greenhouse gases amounts to about 75% of that which would be introduced by a doubling of CO2. Arrhenius’ estimate of 5–6◦ K for the accompanying increase of surface temperature (Sec. 1.2.4) then translates into ∼4◦ K. Yet, the observed change of global-mean temperature since the mid nineteenth century is only about 1◦ K (Sec. 1.6.1). The discrepancy points to changes of the Earth-atmosphere system (notably, involving the major absorbers, water vapor and cloud) that develop in response to imposed perturbations, like anthropogenic emission of CO2.
Focusing on CO2 radiative activity is myopic: The tropospheric heat engine comprises many powerful heat transfer processes.

Figure 9.40 Cloud radiative forcing during northern winter derived from ERBE measurements on board the satellites ERBS and NOAA-9 for the (a) LW energy budget, (b) SW energy budget, and (c) net radiative energy budget. Courtesy of D. Hartmann (U. Washington).
Pg. 318ff
A quantitative description of how cloud figures in the global energy budget is complicated by its dependence on microphysical properties and interactions with the surface. These complications are circumvented by comparing radiative fluxes at TOA under cloudy vs clear-sky conditions. Over a given region, the column-integrated radiative heating rate must equal the difference between the energy flux absorbed and that emitted to space.
The components of cloud forcing (9.53) can be evaluated directly from broadband fluxes of outgoing LW and SW radiation that are measured by satellite. Figure 9.40 shows time-averaged distributions of CSW , CLW , and C. Longwave forcing (Fig. 9.40a) is large in centers of deep convection over tropical Africa, South America, and the maritime continent, where CLW approaches 100 W m−2 (cf. Fig. 1.30b). Secondary maxima appear in the maritime ITCZ and in the North Pacific and North Atlantic storm tracks (Sec. 1.2.5). Shortwave forcing (Fig. 9.40b) is strong in the same regions, where CSW < −100 W m−2. Negative SW forcing is also strong over extensive marine stratocumulus in the eastern oceans and over the Southern Ocean, coincident with the storm track of the Southern Hemisphere. Inside the centers of deep tropical convection, SW and LW cloud forcing nearly cancel. They leave small values of C throughout the tropics (Fig. 9.40c). Negative CSW in the storm tracks and over marine stratocumulus then dominates positive CLW , especially over the Southern Ocean. It prevails in the global-mean cloud forcing. Globally averaged values of CLW and CSW are about 30 and −45 W m−2, respectively.
Net cloud forcing is then −15 W m−2. It represents radiative cooling of the Earth-atmosphere system. This is four times as great as the additional warming of the Earth’s surface that would be introduced by a doubling of CO2. Latent heat transfer to the atmosphere (Fig. 1.32) is 90 W m−2. It is an order of magnitude greater. Consequently, the direct radiative effect of increased CO2 would be overshadowed by even a small adjustment of convection (Sec. 8.7).
Trusting climate models driven by CO2 sensitivity is lop-sided: Natural climate factors are poorly quantified but are orders of magnitude larger than estimated CO2 effects.
Pg. 260
Global climate models are sophisticated extensions of the idealized models considered above. Treatments of climate properties in different GCMs are as varied as they are complex. For some properties, like cloud cover, ice, and vegetation, they must resort to empirical relationships or simply ad hoc parameterization. For others, the governing equations cannot even be defined. Together with the ocean simulation, these limitations introduce errors, which can be substantial. Along with discrepancies between GCMs, they leave in question how faithfully climate feedbacks are represented (see, e.g., Tsushima and Manabe, 2001; Lindzen and Choi, 2009).

The accuracy of GCMs is reflected in the skill with which they simulate the TOA energy budget: the driver of climate. By construction, GCMs achieve global-mean energy balance. How faithfully the energy budget is represented locally, however, is another matter. The local energy budget forces regional climate, along with the gamut of weather phenomena that derive from it. This driver of regional conditions is determined internally – through the simulation of local heat flux, water vapor, and cloud. Symbolizing the local energy budget is net radiation (Fig. 1.34c), which represents the local imbalance between the SW and LW fluxes F0 and F ↑(0) in the TOA energy budget (8.82). Local values of those fluxes have been measured around the Earth by the three satellites of ERBE. The observed fluxes, averaged over time, have then been compared against coincident fluxes from climate simulations, likewise averaged. Figure 8.34 plots, for several GCMs, the rms error in simulated fluxes, which have been referenced against those observed by ERBE.
Values represent the regional error in the (time-mean) TOA energy budget. The error in reflected SW flux, Fs 4 − F0 in the global mean (8.82), is of order 20 Wm−2 (Fig. 8.34a). Such error prevails at most latitudes. Differences in error between models (an indication of intermodel discrepancies) are almost as large, 10–20 Wm−2. The picture is much the same for outgoing LW flux (Fig. 8.34b). For F ↑(0), the rms error is of order 10–15 Wm−2. It is larger for all models in the tropics, where the error exceeds 20 Wm−2.
The significance of these discrepancies depends on application. Overall fluxes at TOA are controlled by water vapor and cloud (Fig. 1.32) – the major absorbers that account for the preponderance of downwelling LW flux to the Earth’s surface. Relative to those fluxes, the errors in Fig. 8.34 are manageable: Of order 10% for outgoing LW and 20% for reflected SW. Relative to minor absorbers, however, this is not the case. The entire contribution to the energy budget from CO2 is about 4 Wm−2. Errors in Fig. 8.34 are an order of magnitude greater. Consequently, the simulated change introduced by increased CO2 (2–4 Wm−2), even inclusive of feedback, is overshadowed by error in the simulated change of major absorbers.
Conclusion:
Pg. 262
Discrepancies between GCMs arise from inaccuracies in climate properties and from differences in how those properties are represented. Much of the discrepancy surrounds the representation of convection and its influence on water vapor and cloud, the absorbers that account for most of the downwelling LW flux to the Earth’s surface. The involvement of convection is strongly suggested by models of radiative-convective equilibrium. Those simulations are inherently sensitive to how convection and cloud are prescribed. Cloud is especially significant to radiative considerations because it sharply modifies the atmosphere’s scattering characteristics, which determine albedo, and its absorption characteristics, which determine optical depth.
Footnote:
Best wishes to Dr. Salby and much appreciation for telling it like it is. May you also nail your landing as did Fearless Felix.

h/t malagabay