In April, the annual European Geosciences Union conference was held in Vienna, Austria. Heleen van Soest, MSc student Climate Studies at Wageningen University, attended the conference, and shares some thoughts and tweets (@Hel1vs).

The opening reception, April 7, reveals that geoscientists are fond of beer. I get to talk to some nice people and hand out my first business cards. Yay! I talk with Walter Schmidt,  President of the Division on Geosciences Instrumentation and Data Systems, about observations and data. Lesson learned: data are important, but never take them for granted. Especially from satellites: they basically measure counts and voltages. To interpret the numbers and get something useful, we already need models, i.e. algorithms. Usually, model skill is tested against data. Disagreement between them is often blamed on model errors, assumptions, etc. Keep in mind that data might be wrong, too. Fortunately, raw data is increasingly archived as such, together with the algorithms used to interpret them. In that way, data can still be used if the algorithms are updated. I dedicate my first #egu2013 tweet to this conversation and go home. I am happy to find a Va Piano (Italian restaurant) in ‘my’ street. Together with Sherlock Holmes (the book, that is), I eat my pasta.

 At #egu2013 opening reception, interesting conversation about models and data: “important, but never take them for granted” (Walter Schmidt)

Monday, 8 April

Permafrost day. An important issue, as permafrost contains about half of the world’s soil carbon. If permafrost thaws, the organic carbon becomes available for microbes to degrade. Greenhouse gas (methane) emissions are a result, further increasing temperatures. This positive feedback is sometimes compared to a time bomb. Modelling studies of permafrost do show it will degrade under further warming. For example, Greenland permafrost south of 76°N will disintegrate this century. However, see RealClimate before you start to worry that this bomb is about to explode.

But today is not only permafrost; I’ve also got something on ice observations.

For ice mass balance estimates, in situ data have sparse spatial coverage, while remote sensing is not very detailed and cannot provide temperature profiles. So both methods will be needed to get a grasp of the system. Many ground or field campaigns are concerned with validation of satellite products. For sea ice thickness, exotic techniques like ultrasound and upward looking sonar can be used. The latter actually measures sea ice draft (the height of the ice above the water), which can be translated to total thickness. New techniques are being developed as well, such as sea ice thickness monitoring by dog sleds with a conductivity probe (using the difference between ice and water). Sounds cool, doesn’t it?

Tuesday, 9 April

There is one session dealing with feedbacks in the Arctic, which I think is very interesting. Many feedbacks play a role, but independent presentations and posters show that some are more important than others (“All animals are equal, but some animals are more equal than others” to quote George Orwell’s Animal Farm). Some of the feedback processes with the most dominant role in determining Arctic temperature change are the ice albedo feedback, the water vapour feedback, and the lapse rate feedback. The first is well known: with (sea) ice loss, more open water or land is exposed to the sun, absorbing more heat and resulting in further ice loss (a positive, or self-amplifying, feedback). The water vapour feedback is not restricted to the Arctic, but a globally important one. With higher temperatures, the air can hold more water vapour, which is a greenhouse gas and as such contributes to further warming (thus also a positive feedback).

The lapse rate feedback is maybe less well-known (see for a good introduction also this “Introduction to climate dynamics and climate modelling“). In the lower part of the atmosphere, called the troposphere, temperature decreases with height (with about 6.5°C per km, the lapse rate). This temperature change is important, as more long wave radiation is emitted at higher temperatures. With global warming,  the upper troposphere in the tropics will warm more than the surface as a result of moist convection, giving a smaller lapse rate. The higher temperatures in the upper troposphere enhance the outgoing radiation to space: a negative lapse rate feedback. At higher latitudes such as the Arctic, however, the surface will warm more than the upper atmosphere. There, more long wave radiation is sent to the surface. So the lapse rate feedback is positive in the Arctic. The global average effect thus depends on a balance between a negative lapse rate feedback in the tropics and a positive lapse rate feedback at high latitudes. For the Arctic itself, all of this means that the lapse rate feedback contributes to Arctic amplification. Arctic amplification is the higher rate of temperature rise in the Arctic compared to the global average (a factor 2). So, if the lapse rate feedback dampens warming in the tropics and amplifies warming in the Arctic, taking the ratio of Arctic over global warming results in a number at least higher than one.

Writing that up refreshed my memory a bit. I could write more about Arctic feedbacks, but let’s just skip to tipping points for now. An important question regarding Arctic sea ice involves tipping points and bifurcations. Some say sea ice has passed a tipping point (see, for example, Livina & Lenton, 2013). Others say that the bifurcation found is just a result of data analysis. Still, amplitude of the seasonal cycle increased while the mean ice extent and the summer minimum decreased. Besides, the Arctic sea ice annual cycle has got into a different regime since 2007, according to Peter Ditlevsen. What does that mean for the future? François Massonnet presents an interesting (sorry, again) modelling study. The five CMIP5 models that best simulated Arctic sea ice predicted a summer ice-free Arctic between 2040 and 2060. This opens up opportunities for exploration and exploitation, and possibly more interesting feedbacks.

Sea ice would have reached minimum even without ‘The Great Arctic Cyclone of August 2012′ -most intense summer storm. ~Irina Rudeva at #egu2013

Wednesday, 10 April

“Prediction is very difficult, especially about the future.” (Niels Bohr)

Moving from sea ice to land, many presentations deal with the Greenland ice sheet. An important concept here is ice sheet mass balance, which compares input (precipitation, snow) with output (e.g. melt water run-off). Until the 90s, Greenland’s ice sheet mass balance was zero, meaning mass loss equalled mass gain. Currently, mass balance is negative due to increasing discharge, so Greenland is losing mass and contributing to sea level rise. Model projections show a further decrease in Greenland’s surface mass balance and significant melt along the ice sheet margins in the future, giving a contribution to sea level rise of 2–20 cm. Although an increase in precipitation might raise the interior of the ice sheet, a decrease in mass balance along the margins also linearly relates to higher temperatures. Linear relations make modelling a little bit easier, but non-linear processes also play a role and should be taken into account. One such non-linear process is the relation between ice sheet topography or height and surface mass balance, the so-called elevation feedback. As temperatures decrease with height in the atmosphere, melting brings the Greenland ice sheet to lower elevations with higher temperatures. The higher temperatures enhance melting, giving rise to the positive elevation feedback. This process can give an additional mass loss of around 10%, compared to projections that do not take the elevation feedback into account. To capture this process well, a regional climate model should be fully coupled with an ice sheet model. And just a small note: Greenland is not the only ice contributing to sea level rise. Over the last decade, the contribution of Northern hemisphere glaciers  to the observed sea level rise increased by 30–40%.

Thursday, 11 April

Today a bit more about Greenland, but first something about snow cover. Snow cover plays a role in boreal forests and permafrost regions, for example. Models can reproduce the mean snow extent, but the variability proves more difficult to capture. Sounds like there are more processes I should take into account in my own model. Maybe I should build a supermodel… Oh wait, such a project even exists! It is called the SUMO project: Super Modelling by combining imperfect models.

Well then, more on Greenland and sea level rise. Model projections using RCP8.5 (the highest emissions scenario) give a contribution by Greenland of 22.3 cm to sea level rise over the next 100 years (Antarctica contributes 7.3 cm). More than 50% of Greenland’s contribution is due to the change in surface mass balance caused by climate change. Atmospheric circulation might still play a role, especially in record events like the widespread ice sheet melt over Greenland in 2012. Antarctica’s contribution to sea level rise, on the other hand, is mostly due to ice shelf basal melting. According to Robert Bindschadler, feedbacks will not be a major factor determining sea level rise during the coming 200 years. Besides, he found that the ice sheet response is surprisingly linear. Going way back in time, mass loss from southern Greenland contributed 17 mm to sea level rise from the little ice age through 2010, which is 9-10% of total sea level rise. Increased thermal expansion also played a role (see e.g. Weather Underground), contributing to the steadily accelerating sea level rise (Jevrejeva et al., 2008).

Friday, 12 April

Found some interesting sessions last-minute. A good way to end my week, with presentations on a topic I would like to look into somewhere in the near future: black carbon, or soot.

Black carbon (BC) is a tiny particle (aerosol) that, when emitted to the air, can absorb solar radiation and warm the atmosphere. When incorporated in more hygroscopic aerosol (e.g. together with sulphate), it can also contribute to cloud formation and affect cloud properties by acting as so-called Cloud Condensation Nuclei (CCN). Clouds can either warm or cool the surface, depending, among other things, on their height in the atmosphere. Black carbon could result in more low clouds in autumn, which warm the surface and could thus contribute to ice melt. The dark BC can also influence snow and ice directly, by lowering the albedo (reflectivity) of these bright surfaces. Including black carbon in models results in snow and ice melt, increased atmospheric absorption of solar UV and more low clouds, contributing to further warming.

But where does that black carbon come from? In summer, the dominant source is biomass burning. Diesel engines also emit black carbon, so filters could help reducing these emissions (see this UNEP report). Another important source turns out to be gas flaring by the oil and petroleum industries. The highest black carbon concentrations in the Arctic occur over Russia and China and are linked to gas flaring there. Globally, flaring contributes only 3% to BC emissions, but the contribution in the Arctic is way larger. One estimate says flaring contributes 42% to Arctic mean BC, another says 2/3 (at 66°N). Either way, black carbon makes a big difference in the Arctic, and flaring in the region will probably increase due to higher industrial activity with sea ice loss. The location of emissions is also important. Emitting black carbon within the Arctic gives a five-fold increase in Arctic surface temperature compared to the same amount of emissions at mid-latitudes (0.24 versus 0.05 K/Tg BC/year). A large part of the difference can be explained by deposition of BC on snow and ice, after emissions within the Arctic.

So far, models have had difficulties with simulating the magnitude and seasonality of Arctic haze, which is formed by high aerosol concentrations in winter and early spring. Including black carbon from domestic combustion and flaring improves the simulations. Including flaring helps simulating the magnitude of the haze (which is usually underestimated) and better explains the seasonality of the haze, because there is also a strong seasonal cycle in flaring. What might further improve model simulations of (high) black carbon concentrations is a monthly or daily resolution.

Over at Lucia’s, Brandon Shollenberger  found a way to search the results of 12,280 out of 12,465 papers. Based on this search method and the SkS paper rating guidelines, Marcel Crok reports the following breakdown of results:

Category 1 (explicit endorsement with quantification): 65
Category 2 (Explicit Endorsement without quantification): 934
Category 3 (Implicit Endorsement): 2933
Category 4 (Neutral): 7930 [the reported number]
Category 5 (Implicit Rejection): 53
Category 6 (Explicit Rejection without quantification): 15
Category 7 (Explicit Rejection with quantification): 10

The 97% is arrived at by adding up categories 1 to 3 and taking that as a percentage of all categories except 4. This percentage is actually 98% using the numbers above, but these are obtained via a shortcut.

Of course, various other fractions could be calculated from this list, each with a slightly different meaning.  E.g. of those abstracts making a statement about the quantitative contribution of human activity to the warming, 87% (65/75) endorsed dominant human causation. And of those abstracts expressing an explicit position on the cause of global warming, 97.6% (999/1024) endorsed human causation.

Any way you slice it, a strong consensus it is.

It should not come as a surprise that many papers expressed no position on the causes of global warming: Many papers matching the search terms “global warming” or “global climate change” don’t deal with attribution of the warming, but with impacts (48% of the total) or mitigation (28% of the total) of climate change (e.g. the change in the yield of corn as a function of temperature). Other papers may deal with one specific aspect of climate change (e.g. the influence of organic aerosol on the cloud nucleating ability of CCN). Moreover, as certain aspects of science become widely accepted, new research papers have less and less reason to state the obvious, let alone in their abstracts. This is corroborated by the increasing fraction of “no position” abstracts and the simultaneously decreasing fraction of “endorsement” abstracts over time (fig 1b of the paper).

Including the “no position” category into the denominator (as some people seem to be doing) to arrive at a much smaller percentage endorsement makes about as much sense as including all atmospheric science articles in the denominator too, or even all physics articles: it is to be expected that many papers do not state a position on this particular issue. These should not be included as a reference against which to compare the number of endorsement papers. Another example of an apples to oranges comparison is Brandon Shollenberger comparing the number of explicit and quantified endorsements to the sum of explicit and implicit rejections.

Dana Nuccitelli and John Cook explain that they took a conservative approach in their ratings:

For example, a study which takes it for granted that global warming will continue for the foreseeable future could easily be put into the implicit endorsement category; there is no reason to expect global warming to continue indefinitely unless humans are causing it. However, unless an abstract included (either implicit or explicit) language about the cause of the warming, we categorized it as ‘no position’.

Furthermore, the implicit endorsement category includes e.g. abstracts that

mention increased CO2 leading to higher temperatures without including anthropogenic or reference to human influence/activity

What is slightly surprising to me is the small number of abstracts making a quantitative statement about attribution (75 out of 12,465). There are bound to be many more attribution studies in the sample, but apparently many of these did not provide a quantitative statement on the human contribution in their abstract.

Invited expert participants in the discussion include Rasmus Benestad (of RealClimate fame), Demetris Koutsoyiannis and Armin Bunde. The introduction text here slightly differs from that posted on ClimateDialogue.org

The Earth is warmer now than it was 150 years ago. This fact itself is uncontroversial. It’s not trivial though how to interpret this warming. The attribution of this warming to anthropogenic causes relies heavily on an accurate characterization of the natural behavior of the system. Here we will discuss how statistical assumptions influence the interpretation of measured global warming.

Agents of change

Global climate can change (say, on time scales > 10 years) due to a variety of processes. For the sake of this discussion, the following processes are distinguished:

-       natural unforced variability (e.g. oscillations or semi-random processes internal to the climate system)

-       natural forced variability (e.g. changes in the output of the sun or in volcanism)

-       anthropogenic forced variability (e.g. changes in greenhouse gas or aerosol concentrations)

Internal variability

Most experts agree that all three types of processes play a role in changing the Earth’s climate over the past 150 years. It is the relative magnitude of each that is in dispute. The IPCC AR4 report stated that “it is extremely unlikely (<5%) that recent global warming is due to internal variability alone, and very unlikely (< 10 %) that it is due to known natural causes alone.” This conclusion is based on detection and attribution studies of different climate variables and different ‘fingerprints’ which include not only observations but also physical insights in the climate processes.

The IPCC AR4 definitions of detection and attribution are:

“Detection of climate change is the process of demonstrating that climate has changed in some defined statistical sense, without providing a reason for that change.”

“Attribution of causes of climate change is the process of establishing the most likely causes for the detected change with some defined level of confidence.”

The phrase ‘change in some defined statistical sense’ in the definition for detection turns out to be the starting point for our discussion. Because what is the ‘right’ statistical model (assumption) to conclude whether a change is significant or not? And how does our understanding of internal variability enter into this picture?

According to AR4, “An identified change is ‘detected’ in observations if its likelihood of occurrence by chance due to internal variability alone is determined to be small.”  Detection is thus concerned with distinguishing the forced from the unforced component (sometimes referred to as signal and the noise), whereas attribution is concerned with assigning causes to the forced component.

There are different methods for estimating the magnitude of natural climate variability. In one approach control runs (without climate forcing) are performed with GCM’s. Critics wonder whether such control simulations are representative of the real world. In another approach a statistical analysis is performed on the observed climatic time series itself. Here the presence of (natural and anthropogenic) climate forcing forms a complicating factor. Some studies have combined both methods and compared modelled and observed time series, as well as their power spectra as a means to circumvent the influence of climate forcing on the timeseries (cf. AR4 fig 9.7).

Long term persistence

Critics argue though that most if not all changes in the climatological time series are an expression of long-term persistence (LTP). Long-term persistence means there is a long memory in the system, although unlike a random walk it remains bounded in the very long run. There are stochastic /unforced fluctuations on all time scales. More technically, the autocorrelation function goes to zero algebraically (very slowly). These critics argue that by taking LTP into account trend significance is reduced by orders of magnitude compared to statistical models that assume short-term persistence (AR1), as was applied e.g. in the illustrative trend estimates in table 3.2 of AR4. (Cohn and Lins, 2005[i]); Koutsoyiannis and Montanari, 2007[ii]).

This has consequences for attribution as well, since long term persistence is often assumed to be a sign of unforced (internal) variability (e.g. Cohn and Lins, 2005; Rybski et al, 2006). However, LTP can also be a consequence of a deterministic trend (e.g. GCM model output also exhibits LTP). In reaction to Cohn and Lins (2005), Rybski et al. (2006)[iii] concluded that even when LTP is taken into account at least part of the recent warming cannot be solely related to natural factors and that the recent clustering of warm years is very unusual (see also Zorita (2008)[iv]). This translates directly into the question of how important the statistical model used is for determining the significance of the observed trends.

Climate Dialogue
Although the IPCC definition for detection seems to be clear, the phrase ‘change in some defined statistical sense’ leaves a lot of wiggle room. For the sake of a focussed discussion we define here the detection of climate change as showing that some of this change is outside the bounds of internal climate variability. The focus of this discussion is how to best apply statistical methods and physical understanding to address this question of whether the observed changes are outside the bounds of internal variability. Discussions about the physical mechanisms governing the internal variability are also welcome.

Specific questions

1. What exactly is long-term persistence (LTP), and why is it relevant for the detection of climate change?
2. Is “detection” purely a matter of statistics? And how does the statistical model relate to our knowledge of internal variability?
3. What is the ‘right’ statistical model to analyse whether there is a detected change or not? What are your assumptions when using that model?
4. How long should a time series be in order to make a meaningful inference about LTP or other statistical models? How can one be sure that one model is better than the other?
5. Based on your statistical model of preference do you conclude that there is a significant warming trend?
6. Based on your statistical model of preference what is the probability that 11 of the warmest years in a 162 year long time series (HadCrut4) all lie in the last 12 years?
7. If you reject detection of climate change based on your preferred statistical model, would you have a suggestion as to the mechanism(s) that have generated the observed warming?

This was the title of a discussion that was held on the recently launched website ClimateDialogue regarding the possible causes of the decline in Arctic sea ice over the past decades. Three experts participated in this discussion: Walt Meier, Research Scientist at the NSIDC, Judith Curry, professor at Georgia Institute of Technology and Ron Lindsay, Senior Principal Physicist at the Polar Science Center of the University of Washington.

In this blog post I will start off with a description of the observations of the Arctic region, followed by a short overview of the potential causes of the decline in Arctic sea ice, incorporating the views of the three experts as they were expressed on ClimateDialogue. The final parts concern the uniqueness of this decline in a historical perspective and the possibility of having an ice-free Arctic in the summer in the not too distant future.

.
Observations of the Arctic region since 1979

Since 1979 the Arctic region has been extensively monitored by satellites. They detect e.g. the ice surface area, the extent of the area covered with ice and also the total amount or volume of ice. The results of these observations are startling. For example, sea ice area and the amount of perennial (multi-year) ice has decreased dramatically over the past 3 decades, as is visualized by the images in figure 1 and 2, generated by NASA (see here and here).

Fig. 1. Minimum ice extent.
The yellow line indicates the average minimum ice extent of the past 30 years compared to the minimum ice extent on September 16th 2012.

Fig. 2. Perennial ice.
The bright white central mass shows the perennial sea ice. The larger light blue area shows the full extent of the winter sea ice including the average annual sea ice during the months of November, December and January.

There are several institutes that collect data from the Arctic region. For example, NSIDC develops data sets about the ice area and the ice extent (the well-known English term for the outer edge of the area covered with ice) and data on the total volume of sea ice is incorporated in the PIOMAS model. In 2012 the September average ice extent dipped below 4 million km², which is about half of what it was in 1979, see figure 3. Ice volume shows a comparable rapid decrease.

Fig. 3. Average NSIDC sea ice extent for September up to and including 2012.Trend line is in red.

Air temperatures in the Arctic region are measured via surface stations and also – albeit indirectly – with satellites. Figure 4 shows the annually averaged temperature data for the Arctic region from NASA GISTEMP and the satellite data from UAH. The former gives a warming trend of 0.53 °C/decade and the latter of 0.47 °C/decade (1979-2012; Arctic region). This is much higher than the increase in global temperatures, which is 0.16 °C/decade for GISTEMP and 0.14 °C/decade for UAH. Since 1979 the Arctic has warmed about 3.3 times faster than the earth in general. This so-called “Arctic Amplification” is partially caused by the disappearance of sea ice and the effect this has on regional albedo (the so-called ice-albedo feedback).

Fig. 4. Average annual temperature in the Arctic region as compiled by GISTEMP and UAH.

.

Processes affecting the amount and extent of Arctic sea ice

A multitude of processes affects the sea ice, such as wind, ocean currents, the temperature of the atmosphere and the temperature of sea water. The three experts on ClimateDialogue agree that greenhouse gases, through an increase in the surface temperature of the atmosphere and the oceans, play a major role in the sea ice decline in recent decades.

Walt Meier for instance puts it like this in his guest blog:
“.. the multi-decadal decline in all seasons, and in virtually all regions cannot be explained without the long-term warming trend that has been attributed to anthropogenic greenhouse gas (GHG) emissions.”.

Two thirds of the melting of the ice is caused by melting from below (at the edges) – through contact with ocean water – and one third is caused by surface melting (Steele et al. 2010^[P23]). The warming of the oceans is therefore of major importance for the long-term trend in Arctic sea ice.

Besides the long-term (multi-decadal) warming trend, natural variation plays a big role in the shorter term (e.g. a decade or less). This view was in principle shared by all three experts, but it also gave rise to a lot of ambiguities and uncertainties on ClimateDialogue. For example, the Arctic Oscillation (AO) and the North Atlantic Oscillation (NAO) affect the amount of sea-ice^[P9]. The Atlantic Meridional Overturning Circulation (AMOC) affects the sea ice extent in the Greenland Sea, the Barents Sea and the Kara Sea^[P11]. The same applies to the Pacific Decadal Oscillation (PDO) regarding the sea ice in the Bering Sea.

Fig. 5. Map of the Arctic Ocean.

The evidence for the impact of greenhouse gases and other factors on the sea ice comes from modeling studies.

Kay et al. 2011^[P13] show that for the period 1979-2005 approximately 50% of the decline in sea ice is caused by natural variation. Interestingly, this study also shows that the decline of the sea ice can temporarily be stopped:
“The computer simulations suggest that we could see a 10-year period of stable ice or even an increase in the extent of the ice.”
But also that the ice will eventually disappear:
“When you start looking at longer-term trends, 50 or 60 years, there’s no escaping the loss of ice in the summer”.

Day et al. 2012^[P12] show that the variation in the AMO can explain 5-30% of the downward trend in the sea ice extent in the September month and that the influence of the AO thereon is small.

A conclusion from the correlation study of Notz en Marotke 2012^[P8] is:
“We find that the available observations are sufficient to virtually exclude internal variability and self-acceleration as an explanation for the observed long-term trend, clustering, and magnitude of recent sea-ice minima. Instead, the recent retreat is well described by the superposition of an externally forced linear trend and internal variability. For the externally forced trend, we find a physically plausible strong correlation only with increasing atmospheric CO2 concentration. Our results hence show that the observed evolution of Arctic sea-ice extent is consistent with the claim that virtually certainly the impact of an anthropogenic climate change is observable in Arctic sea ice already today.”
They find a strong long-term correlation between the CO₂ forcing and the ice extent in the September month, alongside low correlations with the PDO and the AO.

Fig.6. Correlation between September sea ice extent and CO2 forcing (red), forcing by the sun (blue), PDO index (green) and AO index (yellow) [From Notz & Marotke].

Fig.7. Ice transport through Fram Strait in 2002.

All three experts agree that greenhouse gases played a major role in the long-term decline of the Arctic sea ice in recent decades, although they find it difficult to put a precise number on it.

Keep in mind that their estimates are dividing between human and natural factors, together making 100%. In the discussion greenhouse gases are alluded to as the human influence, but of course (e.g. cooling by aerosols), so a breakdown in greenhouse gases on one hand and natural variability on the other hand is not useful. It is likely that with greenhouse gases the net human influence is meant, since it is implicitly assumed that the total should be 100%.
A range of 50-70% with an uncertainty of 20% is probably a reasonable average of what the three experts think the human contribution to Arctic sea ice melt is, see the quotes below.

Judith Curry:
“My assessment is that it is likely (>66% likelihood) that there is 50-50 split between natural variability and anthropogenic forcing, with +/-20% range.”
“The disagreement seems to arise if we are each forced to pick a single attribution value: mine would be 50%”
Ron Lindsay:
“CCSM4 simulations show about a 50% decline in ice volume since the 1960’s with a typical ensemble spread on the order of 15%, so the CCSM4 runs indicate the decline in ice volume is about 3 times the natural variability, or about 70% of the decline is due to greenhouse gases. The decline in volume seen in the PIOMAS simulations is also very consistent, particularly if one focuses just on the Arctic Ocean since the late 1980’s. So I would go on the high side of the percentage loss due to greenhouse gases for ice volume and less for ice extent, maybe near 50%.”
Walt Meier:
“The 50-70% range for GHGs that Judith mentions is probably a reasonable spread in capturing the potential range.”

However, there is indeed a difference in judgment between Judith Curry and the other two experts on the human influence. In my opinion Meier explains his position most clearly. He refers to Day et al., who arrive at a 70-95% anthropogenic contribution, and gives some strong arguments in his guest blog:
- The decline in sea ice correlates with the increase in global temperature.
- The decline is outside the range of normal variability over the past several decades and probably over the past several centuries.
- The decline is pan-Arctic, with all regions experiencing declines throughout all or most of the year.
- Climate model simulations cannot explain the decline without taking greenhouse gases into account.
- There does not appear to be another mechanism to sufficiently explain the long-term decline.

Curry indicates that the greenhouse gases undoubtedly contribute to the reduction of sea ice, but she highlights the uncertainties therein, especially for the short term. According to her, there is a complex interplay between natural variation and CO₂ forcing together with complex interactions between ocean dynamics, heat transport and sea ice dynamics driven by winds and ocean currents. To support her estimated 50-50 division between anthropogenic and natural causes she highlights the high degree of uncertainty, although uncertainty in itself only causes a wide error margin and is no reason for a given estimate.

Lindsay also refers to Day et al. and says that the evidence must come primarily from modeling studies, only those can help us separate natural climate variations from variations caused by changes in greenhouse gases or other external forcing mechanisms, e.g. the sun or volcanoes.
But above all Lindsay says:
“I believe fundamentally the main process causing the decline in Arctic sea ice is increasing greenhouse gases.”.

.
How unique is the current decline in sea ice in historical perspective?

The most complete data sets are of course the satellite measurements. There is a fairly complete coverage from ice charts dating back to the 1950s and there are Russian and Danish ice charts dating back to the 1930s^[T1]. The 1930s are often cited as a warm period, but this was mainly confined to the Atlantic^[T2]. For the years preceding the 1930s we must rely on proxy data to get an idea of the ups and downs of the Arctic sea ice.

In a recently published paper by Meier et al. ^[P24] the sea ice extent from 1953 onwards is mapped and then linked to satellite data. Figure 8, taken from this paper, clearly shows that the decline of Arctic sea ice has started in the 1970s, not coincidentally at the same time that global surface temperatures started to increase fast.

Fig.8. Arctic sea ice extent starting in January 1953 up to and including December 2011.

A study by Kinnard from 2011^[P21] shows that the decline of the Arctic sea ice extent is unique since at least the last 1450 years and that it corresponds to the surface temperatures in the Arctic.

Fig.9. Sea ice extent (red) and surface temperatures (blue) in the Arctic from Kinnnard et al. 2011.

There are indications that the last time sea ice extent in the Arctic was this low, was during the Holocene Thermal Maximum, about 8000 year BP. See the review paper by Polyak et al. from 2010^[P5].
A good summary of the history of Arctic sea ice can, as so often, be found at SkepticalScience.

.
An ice-free Arctic.

The common definition of ice-free comes from Wang & Overland^[P18,P19], who indicate that ‘nearly sea ice free’ implies a maximum surface area of less than 1 million square kilometers. It is uncertain when the Arctic will become ice-free for the first time. In their 2012 paper Wang & Overland come up with the following indication, based on CMIP4 models:
“Applying the same technique of model selection and extrapolation approach to CMIP5 as we used in our previous paper, the interval range for a nearly sea ice free Arctic is 14 to 36 years, with a median value of 28 years. Relative to a 2007 baseline, this suggests a nearly sea ice free Arctic in the 2030s.”

The CMIP5 models do a better job in simulating the decline of the sea ice than the older CMIP3 models (as used for IPCC AR4), see Stroeve et al. 2012^[P7]. Figure 10, based on a presentation of Stroeve, indicates the difference between the different models and the observation of Arctic sea ice extent in the September month. The green dot in this figure stands for sea ice extent in September 2012. It is clear that models still partially underestimate the decline of the sea ice, although the observations may not fall significantly outside the range of model calculations.

Fig.10. Observations and modeling of the sea ice extent of the September month.

Of course the three experts on ClimateDialogue do not come up with a certain date upon which the Arctic would become ice-free and they point at the uncertainty which is mainly caused by natural variation. They do indicate that, logically, natural variation will continue to play a role after the Arctic has become ice-free. Therefore, an ice-free year will be followed with years where more or less sea ice will be present. That the Arctic will become ice-free in the summer months is now almost certain, again nicely put into words by Meier:
“The timing is still uncertain but it changes things from a “if the Arctic loses summer sea ice” to “when the Arctic loses summer sea ice”.”

The area covered with sea ice has declined sharply the last 30 years and will continue to decline in the long run. This has consequences for the rest of the world, as can be read in this excellent summary by Neven: “Why Arctic sea ice shouldn’t leave anyone cold”.
This continuous decline of Arctic sea ice certainly has an impact on the Arctic climate and also on the weather in our part of the world, beautifully expressed by Professor Jennifer Francis^[P25]:
The question is not whether sea-ice loss is affecting large-scale atmospheric circulation…
…it’s how can it not?

.
Many thanks to Neven and Bart.
.
References

Data references:
- PIOMAS ice volume
- NSIDC ice extent
- UAH temperature data
- GISS temperature data

Text references Climatedialogue.org:
[T1] Walt Meier – http://www.climatedialogue.org/melting-of-the-arctic-sea-ice/#comment-38
While our most complete dataset, the one we have the highest confidence in, is the passive microwave record, there is fairly complete coverage from operational ice charts back to at least the mid-1950s. And there are Russian ice charts for the Eurasian Arctic back to the early 1930s. Though not complete, these do extend the record and I think provide some sense of the interannual and decadal natural variability of the ice. There are indications of lower ice in the 1930s in the Russian Arctic [Reference 4 in my post], suggesting the influence (AMO?) of a multi-decadal cycle, at least in the Russian Arctic. But the data show a different character in terms of the seasonality and regionality of the lower ice conditions compared to the recent decline.
[T2] Walt Meier – http://www.climatedialogue.org/melting-of-the-arctic-sea-ice/#comment-237
Before the 1950s, the 1930s are often mentioned as a warm period. However, this is primarily in the Atlantic region, where observations were more common. Ice charts from the Denmark (http://nsidc.org/data/docs/noaa/g02203-dmi/) and Russia, indicate some periods of low summer ice, but on a more regional scale than we see now.

Peer-review paper references provided by the experts on ClimateDialogue.org :
[P1] Maslanik, J., J. Stroeve, C. Fowler, and W. Emery (2011), Distribution and trends in Arctic sea ice age through spring 2011, Geophys. Res. Lett., 38, L13502, doi:10.1029/2011GL047735.
[P2] Overland, J.E., M. Wang, and S. Salo (2008), The recent Arctic warm period, Tellus, doi: 10.1111/j.1600-0870.2008.00327.x.
[P3] Mahoney, A. R., R. G. Barry, V. Smolyanitsky, and F. Fetterer (2008), Observed sea ice extent in the Russian Arctic, 1933–2006, J. Geophys. Res., 113, C11005, doi:10.1029/2008JC004830.
[P4] Overland, J.E., M.C. Spillane, D.B. Percival, M. Wang, H.O. Mofjeld (2004), Seasonal and regional variation of pan-Arctic surface air temperature over the instrumental record, J. Climate, 17, 3263-3282.
[P5] Polyak, L., and several others (2010), History of sea ice in the Arctic, Quaternary Sci. Rev., 29, 1757-1778, doi:10.1016/j.quascirev.2010.02.010.
[P6] Stroeve, J., M.M. Holland, W. Meier, T. Scambos, and M. Serreze (2007), Arctic sea ice decline: Faster than forecast, Geophys. Res. Lett., 34, L09501, doi:10.1029/2007GL029703.
[P7] Stroeve, J.C., V. Kattsov, A. Barrett, M. Serreze, T. Pavlova, M. Holland, and W.N. Meier (2012), Trends in Arctic sea ice extent from CMIP5, CMIP3 and observations, Geophys. Res. Lett., 39, L16502, doi:10.1029/2012GL052676.
Presentatie Stroeve
[P8] Notz, D. and J. Marotzke (2012), Observations reveal external driver for Arctic sea-ice retreat, Geophys. Res. Lett., 39, L08502, doi:10.1029/2012GL051094.
[P9] Rigor, I.G., J.M. Wallace, and R.L. Colony (2002), Response of sea ice to the Arctic Oscillation, J. Climate, 15, 2648-2663.
[P10] Stroeve, J. C., J. Maslanik, M. C. Serreze, I. Rigor, W. Meier, and C. Fowler (2011), Sea ice response to an extreme negative phase of the Arctic Oscillation during winter 2009/2010, Geophys. Res. Lett., 38, L02502, doi:10.1029/2010GL045662.
[P11] Mahajan, Salil, Rong Zhang, Thomas L. Delworth (2011), Impact of the Atlantic Meridional Overturning Circulation (AMOC) on Arctic surface air temperature and sea ice variability. J. Climate, 24, 6573–6581, doi: 10.1175/2011JCLI4002.1.
[P12] Day, J.J., J.C Hargreaves, J.D. Annan, and A. Abe-Ouchi (2012), Sources of multi-decadal variability in Arctic sea ice extent, Env. Res. Lett., 7, 034011, doi: 10.1088/1748-9326/7/3/034011.
[P13] Kay, J. E., M. M. Holland, and A. Jahn (2011), Inter-annual to multi-decadal Arctic sea ice extent trends in a warming world, Geophys. Res. Lett., 38, L15708, doi:10.1029/2011GL048008.
Press Release
[P14] Holland, M.M., Bitz, C.M. and Tremblay, B. (2006), Future abrupt reductions in the summer Arctic Sea ice. Geophys. Res. Lett. 33, L23503, doi:10.1029/2006GL028024.
[P15] Tietsche, S., D. Notz, J. H. Jungclaus, and J. Marotzke (2011), Recovery mechanisms of Arctic summer sea ice, Geophys. Res. Lett., 38, L02707, doi:10.1029/2010GL045698.
[P16] Amstrup, S.C., E.T. DeWeaver, D.C. Douglas, B.G. Marcot, G.M. Durner, C.M. Bitz, and D.A. Bailey (2010), Greenhouse gas mitigation can reduce sea-ice loss and increase polar bear persistence, Nature, 468, 955-958, doi: 10.1038/nature09653.
[P17] Maslowski, W., J.C. Kinney, M. Higgins, and A. Roberts (2012), The future of Arctic sea ice, Ann. Rev. Earth and Planetary Sciences, 40, 625-654, doi: 10.1146/annurev-earth-042711-105345.
[P18] Wang, M., and J. E. Overland (2009), A sea ice free summer Arctic within 30 years?, Geophys. Res. Lett., 36, L07502, doi:10.1029/2009GL037820.
[P19] Wang, M. and J. E. Overland (2012), A sea ice free summer Arctic within 30 years: An update from CMIP5 models, Geophys. Res. Lett., 39, L18501, doi:10.1029/2012GL052868.
[P20] Schweiger, A., R. Lindsay, J. Zhang, M. Steele, H. Stern, and R. Kwok. 2011. Uncertainty in Modeled Arctic Sea Ice Volume. J. Geophys. Res., doi:10.1029/2011JC007084
[P21] Kinnard, C., C. Zdanowicz , D Fisher, and E. Isaksson, 2011: Reconstructed changes in Arctic sea ice over the past 1,450 years, Nature, 509-512, doi 10.1038/nature10581.
Figuur 3.2 Kinnard et al
[P22] Schweiger, A., R. Lindsay, J. Zhang, M. Steele, H. Stern, and R. Kwok. 2011. Uncertainty in Modeled Arctic Sea Ice Volume. J. Geophys. Res., doi:10.1029/2011JC007084
[P23] Lindsay, R. W. and J. Zhang, 2005: The thinning of arctic sea ice, 1988-2003: have we passed a tipping point?. J. Climate, 18, 4879–4894.
[P23] M.Steel, J.Zhan, W.Ermold, Mechanisms of summertime upper Arctic Ocean warming and the effect on sea ice melt, Journal of Geophysical Research, VOL. 115, C11004, doi:10.1029/2009JC005849, 2010
[P24] W. N. Meier, J. Stroeve, A. Barrett, and F. Fetterer, A simple approach to providing a more consistent Arctic sea ice extent time series from the 1950s to present, The Cryosphere, 6, 1359-1368, 2012, doi:10.5194/tc-6-1359-2012
[P25] J. Francis, S. Vavrus, Evidence linking Arctic amplification to extreme weather in mid-latitudes, Geophysical Research Letters, Vol. 39, L06801, doi:10.1029/2012GL051000, 2012
Video Weather and Climate Summit – Jennifer Francis.
[P26] D. Kaufmann et al, Recent warming reverses long-term Arctic cooling, Science 4 September 2009, Vol. 325, no. 5945, doi:10.1126/science.1173983
Website 2000 Years of Climate Variablity from Arctic Lakes

The big picture (or as some call it: the Wheelchair): Global average temperature since the last ice age (20,000 BC) up to the not-too distant future (2100) under a middle-of-the-road emission scenario.

Figure 1: The temperature reconstruction of Shakun et al (green – shifted manually by 0.25 degrees), of Marcott et al (blue), combined with the instrumental period data from HadCRUT4 (red) and the model average of IPCC projections for the A1B scenario up to 2100 (orange).

Earlier this month an article was published in Science about a temperature reconstruction regarding the past 11,000 years. The lead author is Shaun Marcott from Oregon State University and the second author Jeremy Shakun, who may be familiar from the interesting study that was published last year on the relationship between CO2 and temperature during the last deglaciation. The temperature reconstruction of Marcott is the first one that covers the entire period of the Holocene. Naturally this reconstruction is not  perfect, and some details will probably change in the future. A normal part of the scientific process.

The temperature reconstruction ends mid-20th century, so the rapid temperature rise since 1850 is clearly visible in the graphs presented in their study.
And what do we see? Again something that looks like a hockey stick as in the graph from Mann et al 2008.

Figure 2: The temperature reconstruction of Marcott 2013 (past 11,000 years) and a collection of reconstructions (past 1800 years) as presented by Mann 2008.

Are the results from Marcott et al surprising?
Not really. The well-known graph of Holocene temperature variations on Global Warming Art, which is often encountered on the internet, is actually a comparable image. One could say that Marcott et al managed to scientifically confirm the average thick black line of the Global Warming Art image. See figure 3.

Figure 3: Holocene temperature variations from Global Warming Art, with the average in black, combined with the reconstruction of Marcott 2013 in red.

Patterns in temperature reconstructions which resemble a hockey stick, are fervently contested on climate skeptic websites. The Marcott et al reconstruction is no exception. For example, it is hard to keep track of the number of posts WUWT dedicated to this study, and the statistical wonderboy McIntyre is also energetically producing blog posts. Otherwise the general public might get the impression that humans are strongly influencing the climate and apparently that is not a desirable impression.

The study of Marcott suggests that the earth is warming rapidly from a historical perspective, though the authors warn that the low time resolution of about 120 years and subsequent smoothing preclude a hard statement on whether it is truly unprecedented. The study is about the Holocene, the geological period of 11,700 years ago until now. From the main image of Marcott 2013 it can be deduced that after the last ice age, earth’s temperature has risen to about 7000 years ago, followed by a slow decline. The cause of the gradual cooling in recent millennia is a change in the distribution of solar radiation over the earth and over the seasons, known as Milankovitch cycles, which are responsible for the initiation and termination of ice ages.

After the year 1850, the influence of man-made emissions  is clearly visible in Marcott’s figure, an unprecedented increase in temperature in terms of speed over more than 100 years. The average temperature of the last decade was higher than the temperatures have been for 72% of the past 11,000 years:

Our results indicate that global mean temperature for the decade 2000–2009 has not yet exceeded the warmest temperatures of the early Holocene (5000 to 10,000 yr B.P.). These temperatures are, however, warmer than 82% of the Holocene distribution as represented by the Standard5×5 stack, or 72% after making plausible corrections for inherent smoothing of the high frequencies in the stack.

Epochs have a beginning and an end. From the main image of Marcott’s study you could deduce that, regarding climate, a new epoch has begun about 150 years ago. A clear break in the trend over the past 11,000 years. The end of the Holocene was reached in 1850 and the Anthropocene has started, the epoch in which man asserts its influence on climate. This leads to disbelief in certain parts of the population, which was predicted at the start of the Anthropocene by Charles Dickens, who wrote in 1859:

It was the best of times, it was the worst of times,
it was the age of wisdom, it was the age of foolishness,
it was the epoch of belief, it was the epoch of incredulity

Figure 1 at the beginning of the blog post clearly shows that mankind is creating a new world with a climate that human civilization has never encountered before. If human greenhouse gas emissions continue unabated, the temperature will go up even further. According to the IPCC 2007 A1B scenario we will probably have temperatures in the year 2100 that are about +3.0 degrees above the average of 1961-1990. The expected jump in the atmospheric temperature from 1850 to 2100 is of the same order of magnitude as the rise in temperature from the last ice age to the Holocene, as derived from the Shakun 2012 data. The difference is that current and future increase in temperature occurs orders of magnitude faster.

Marcott et al also refer to the climate model projections of IPCC 2007:

Climate models project that temperatures are likely to exceed the full distribution of Holocene warmth by 2100 for all versions of the temperature stack, regardless of the greenhouse gas emission scenario considered (excluding the year 2000 constant composition scenario, which has already been exceeded). By 2100, global average temperatures will probably be 5 to 12 standard deviations above the Holocene temperature mean for the A1B scenario based on our Standard5×5 plus high-frequency addition stack.

I.e. unprecedented, as many as 5 to 12 standard deviations above the mean of the temperatures in the Holocene. Welcome to the Anthropocene!

A famous SF series from past times always began with:
“To boldly go where no man has gone before”
Indeed, we are boldly entering a new epoch where no man has gone before. I have some doubts whether our descendants will be so delighted about it.

.
[UPDATE 31 March 2013]
A summary and FAQ related to the study by Marcott et al (2013, Science), prepared by the authors, can be found at RealClimate.

Dr. John Christy wrote an extensive blog post as a response to my Dutch ‘Klotzbach Revisited’ post (English version here), it is published on “Staat van het Klimaat” and WUWT. I would like to thank Dr. Christy for his interest in my writings.

I have some remarks regarding Dr. Christy’s post, which are addressed in this ‘response-post’ and are built upon some quotes taken from Dr. Christy’s response.
For reference, the original Klotzbach et al 2009 paper (K-2009 in the text) can be found here and the correction paper (K-2010) can be found here.

“Klotzbach et al.’s main point was that a direct comparison of the relationship of the magnitude of surface temperature trends vs. temperature trends of the troposphere revealed an inconsistency with model projections of the same quantities.”

This ‘main point’ is not present at all in the K-2009 paper, the only reference to real data coming from a climate model in the paper is the amplification factor, which was ‘sort of obtained’ by Ross McKitrick from the GISS-ER model. In the abstract a short conclusion is given: “These findings strongly suggest that there remain important inconsistencies between surface and satellite records.”. No word about models.

In my opinion the main point of K-2009 is the suggestion that the surface temperature record is biased. One third of the paper is made up by paragraph 2 with the title: “Recent Evidence of Biases in the Surface Temperature Record”. K-2009 explicitly state:
In our current paper, we consider the possible existence of a warm bias in the surface temperature trend analyses …

“It appears Hagelaars’ key point is that when the data from Klotzbach et al. are extended beyond 2008 to include data through 2012, the discrepancies, i.e. the observed difference between surface and tropospheric trends relative to what models project, are reduced somewhat.”

My key point is that if K-2009 were correct, the absolute temperature difference between surface and troposphere would be expected to increase over time (due to the fact that the presumed bias in the surface temperature data has not magically disappeared, see e.g. this paper by Watts et al). In contrast, this temperature difference has decreased about 33% for the ‘NCDC minus UAH’ data (which showed the largest discrepancy). This absolute difference was used by Marcel Crok in his book “De Staat van het Klimaat” to suggest that the surface temperatures is biased.
Why this large difference with only 13% (4 years) more data? If anything, it casts doubt on the robustness of the K-2009 results.

The other point I wanted to make is that the apparent discrepancies could also, perhaps even more likely, be due to biases present in the satellite data, as indicated by Santer et al 2005. Also new biases are constantly being discovered, see the Po-Chedley & Fu paper.
Why is there no attention given to the potential presence of biases in the satellite data in Dr. Christy’s blog post?

“.. there have been many studies which have looked at the relationship between the magnitude of the surface temperature trend relative to that of the tropospheric layer as defined above (e.g. Douglass et al. 2007.)”

Indeed, and I gave some references in “Klotzbach Revisited”, for instance Santer et al 2005, Karl et al 2006 and a review of Thorne et al 2011.
Douglass et al contains serious flaws as indicated by Santer et al 2008. I would recommend this youtube video of a presentation by Santer to people who are interested in the controversy around this Douglass-2007 paper.

“As noted however, several additional calculations confirm the value of 1.1 utilized by Klotzbach et al. 2010.”

The K-2009 paper uses 1.2 as an amplification factor over land and the correction paper K-2010 uses 1.1. I never found this very plausible. Gavin Schmidt, responsible for the GISS-ER model gave a value by e-mail to Phil Klotzbach of 0.95, still K-2010 uses the value of 1.1 as calculated on McIntyre’s blog. On the same blog Gavin Schmidt gives the following amplification factor:

“A range of [0.784,1.234]… and a mean (if you think that is sensible) of 0.9708 . Lest anyone think that volcanoes or something are affecting this, the same calculation for 2010-2100 is a range of [0.914,1.097] and a mean of 0.9897.”
This 0.97 is close to his previous 0.95.

The selection of an amplification factor in K-2009 and K-2010 is arbitrarily, why use the factor 1.1 and not the factor calculated by the person responsible for the GISS-ER model?
In my opinion neither of these factors should be used in scientific papers, like the Klotzbach et al papers, since they are obtained from blogs and are not backed up by peer-reviewed science.

“It is true that these differences are a little closer to zero than shown in Klotzbach et al., but that is due to the fact that there has been no warming in the past 10 years in both types of data”.

I’m glad that Dr. Christy endorses my results, but I do not agree with the ‘but’ part. I do not think conclusions regarding a (change in) climatologically relevant trend can be drawn based upon 10 years of data. However, since Dr. Christy explicitly refers to this very short timescale I will present some trend data regarding the last 10 years (Feb. 2003 – Jan. 2013). Values in °C/decade, first global, then over land, followed by the ocean part:
UAH: +0.055 / +0.159 / -0.003
NCDC: -0.038 / +0.048 / -0.068
NCDC minus UAH: -0.092 / -0.111 / -0.064
It is quite clear that over the last 10 years the UAH dataset (the lower troposphere) gives a positive global warming trend and NCDC does not. Both datasets give a positive warming trend over land and that is not the same as ‘no warming’. The trend over land for UAH is 3.3 times higher than the trend over land by NCDC and this certainly does not reflect the amplification factors 0.95 or 1.1 mentioned in the blog posts.
The land area on earth has warmed during the past 10 years and the relationship between the surface data and the satellite data has turned upside down, over 34 years the trend difference (NCDC minus UAH) was +0.10 °C/decade and it is reversed the last 10 years to -0.11 °C/decade. Why?
I did not use amplification factors in calculating the trend differences (see figure 5 in the original blogpost) due to the fact that these factors over land and ocean are, as far as I can tell, not backed up by peer-reviewed science.

“Therefore models, on average, depict the last 34 years as warming about 1.5 times what actually occurred.”

Dr. Christy is comparing trends based upon averages from climate models with the trend in the observations. Averaging model run will also average out natural variability as present in each model simulation. For instance, the simulated influence of ENSO will not be visible anymore in the averaged data (e.g. see this graph and note the absence of any El Nino variability in the averaged model data). It is not realistic to expect that the observations will nicely follow the model average. Comparing a trend of a model average with the observational surface temperature trend and concluding from such a (simplistic and incomplete) comparison that “the climate sensitivity of models is too high”, is, in my opinion, jumping to conclusions.

In the animation below the temperature observations are compared with two simulations from a climate model. The blue simulation matches the observed trend over the most recent decade and the red does not. In the long run the expected warming is roughly the same. The original (created by Ed Hawkins) can be found at: http://www.climate-lab-book.ac.uk/2013/what-will-the-simulations-do-next/

“Since this increased warming in the upper layers is a signature of greenhouse gas forcing in models, and it is not observed, this raises questions about the ability of models to represent the true vertical heat flux processes of the atmosphere and thus to represent the climate impact of the extra greenhouses gases we are putting into the atmosphere.”

The increased warming in the upper layers of the troposphere is due to the lapse rate feedback and is not a signature restricted to the influence of greenhouse gases. Since the lapse rate feedback is a negative feedback, a smaller lapse rate feedback would in fact result in a larger climate sensitivity as obtained from models. More info regarding this subject can be found on Skeptical Science and RealClimate, for Dutch readers the subject is also covered at Klimaatportaal.

This statement of Dr. Christy is quite an extraordinary claim and according to Carl Sagan “extraordinary claims require extraordinary evidence”. Following the blog post of Dr. Christy, the claim is based upon satellite and surface data that potentially could contain biases, on amplification factors taken from blogs and not from peer-reviewed science papers and by comparing observations with model averages. I would not classify this evidence as extraordinary, especially since many other lines of evidence contradict Dr. Christy’s claim (see e.g. here or here).

I’m left with a lot of questions regarding this topic:

• What explains these large differences in the comparison of satellite and surface temperatures with only 13% more data?
• Are there potential biases in the satellite data, e.g. due to a change in satellites or other factors?
• Is the (autocorrelated) noise in the data playing tricks on me/us?
• What is the skill of climate models in simulating the vertical temperature structure and the influence of the planetary boundary layer?
• What are the actual amplification factors for land and ocean over the satellite period?
• Are the amplification factors constant in time or do they vary, e.g. are they influenced by ENSO? If they vary, what is the magnitude and cause of the variation?
• Is it possible to make a strict division between land and ocean for the complete lower troposphere, e.g. is there some averaging out from ocean to land and vice versa at higher altitudes due to convection?

There are several scientific puzzles remaining; hopefully science can resolve some of them in the near future. Until then, let’s not jump to unsubstantiated conclusions.

The average surface temperature of the earth, measured by ‘thermometers’, are released by a number of institutes, the most well-known of these datasets are GISTEMP, HadCRUT and NCDC. Since 1979 temperature data for the lower troposphere are released by the University of Alabama in Huntsville (UAH) and Remote Sensing Systems (RSS), which are measured by satellites.
The temperatures of these two methods of measurement show differences, for instance: the NCDC data indicate a trend over land of 0.27 °C/decade for the period 1979 up to and including 2012, while over the same period, the trend based upon the satellite data by UAH over land is significantly lower at 0.18 °C/decade. In contrast, the trends for global temperatures indicate much smaller differences, for NCDC and UAH these are respectively 0.15 °C/decade and 0.14 °C/decade for the same period.

Big deal? Almost everything related to climate is a ‘big deal’, so it is of no surprise that the same applies to these trend differences. In a warming world it is expected that the temperatures of the upper troposphere increase at a higher rate than at the surface, regardless of the cause of the warming. The satellite data (UAH and RSS) do not reflect this. Why is the upper troposphere expected to warm at a higher rate and what is the cause of these trend differences between the surface  and satellite temperatures?

The temperature gradient in the troposphere / the ‘lapse rate’

When you go up in the troposphere it gets colder. This is caused by the fact that rising air will cool down with increasing altitude due to a decrease in pressure with altitude, by means of so-called adiabatic processes. This temperature gradient is called the lapse rate, a concept one will frequently encounter in papers regarding the atmosphere in relation to climate. When the air is dry, this temperature drop is about 10 °C per km. When the air contains water vapor, this vapor will condense to water upon cooling as a result of the rising of the air, which releases heat of condensation. So in this way, heat is transported to higher altitudes and the temperature drop with height will decrease. For air saturated with water vapor, this vertical temperature drop is approximately 6 °C per km.

When the earth gets warmer, air can contain more water vapor. This also has an impact on the lapse rate, since more water vapor means more heat transfer to higher altitudes. This effect on the lapse rate is called the lapse rate feedback. More heat at higher altitudes implies that there will be more emission of infrared light, a negative feedback. This effect is particularly important in the tropics. At higher latitudes, the increase in temperature at the surface is dominant, therefore the change in the lapse rate will turn into a positive feedback. See figure 1 (adapted from the climate dynamics webpage of the University of Leuven).

Figure 1. A schematic representation of positive and negative lapse-rate feedbacks.

On average it is expected that the negative lapse rate feedback will dominate, leading to an overall value of about -0.8 W/(m²·K) (Soden and Held 2006 or IPCC 2007).

The satellite temperatures are by no means a representation of surface temperatures; they represent some sort of average of the entire lower troposphere. The temperature values are derived from the microwave radiation of oxygen from different heights in the atmosphere. See figure 2 for the weighting functions against altitude of the RSS temperatures for the lower troposphere (TLT). These weighting functions are different for land and ocean areas.

Figure 2: The weighting functions against altitude for the RSS TLT data.

Since it is expected that the average lapse rate feedback will be negative, the temperature aloft should on average increase more than at the surface (because when it gets warmer, more heat will be transported to higher altitudes). As mentioned above, this is however not corroborated by the comparison between surface and satellite temperatures.

These differences have been a regular subject of research. The most well-known investigations are Santer et al 2005, Karl et al 2006 and a review by Thorne et al 2011.

Santer et al concluded that on monthly and annual time scales the temperature observations of the troposphere in the tropics were consistent with the theory and show a greater warming at a greater height than at the surface. However on a time scale of decades, they only encountered one dataset that met these expectations, as can be seen in figure 3 (adapted from Thorne 2011).

Figure 3: Tropical temperature behavior in observations and models.
The black line in the graphs indicates an amplification factor of 1 for the surface temperature (Ts) to the temperature of the complete lower troposphere (2LT);the red line denotes an amplification factor of about 1.3 (slope of model expected tropospheric versus surface temperatures).
The graph on the left is a representation of the month-to-month variability, were the models and the observations (radiosonde data, satellites and surface temperatures) are in agreement.
The graph on the right is a representation of the trends on a multi-decadal scale. Here the models also show an amplification factor of about 1.3, whereas the observations show a smaller amplification factor, in some cases even smaller than 1.

The extensive report of the U.S. Climate Change Science Program (Karl et al) concluded:
These results could arise due to errors common to all models; to significant non-climatic influences remaining within some or all of the observational data sets, leading to biased long-term trend estimates; or a combination of these factors. The new evidence in this Report (model-to-model consistency of amplification results, the large uncertainties in observed tropospheric temperature trends, and independent physical evidence supporting substantial tropospheric warming) favors the second explanation.

So, a deviation in the long-term trend in the observations would be the most likely explanation for the trend differences. Looking specifically at the satellites, this is not illogical given the difficulties inherent to satellite measurements (see e.g. the various UAH versions), such as the drift in the satellites’ orbit, sensors which deteriorate over time, calibration problems when changing from one satellite to another, or temperature effects in the instruments themselves. Even last year Po-Chedley & Fu found another bias in these satellite temperatures, and one of their conclusions was:
“Creating climate-quality satellite temperature datasets is a challenging process that requires constant attention as new biases are discovered.”

Klotzbach 2009

In 2009 a paper was published in the Journal of Geophysical Research by Klotzbach, Pielke Jr. and Sr., Christy and McNider, with the title:
“An alternative explanation for differential temperature trends at the surface and in the lower troposphere”.
Their general conclusion was:
“The differences between trends observed in the surface and lower-tropospheric satellite data sets are statistically significant in most comparisons, with much greater differences over land areas than over ocean areas. These findings strongly suggest that there remain important inconsistencies between surface and satellite records.”
The Klotzbach-2009 paper implies that these ‘inconsistencies’ are caused by biases in the surface temperatures, see their paragraph 2 with the title: “Recent Evidence of Biases in the Surface Temperature Record”.

To assess the trend differences between NCDC/HadCRUT3 and the UAH/RSS data, they took the monthly values of the different datasets, subtracted these from each other and calculated a trend over these monthly differences. This trend should be 0 when there is no difference between the datasets. Figure 4 contains the results from the Klotzbach-2009 paper (their table 2).

Figure 4: Differences in trends between NCDC/HadCRUT3 and UAH/RSS over 1979-2008.

The biggest difference was obtained for NCDC minus UAH with a trend of +0.15 °C/decade over land and +0.04 °C/decade for the globe as a whole, for which a negative ratio would be expected. This is consistent with the conclusions drawn by Santer et al that the multi-decadal trend is anomalous, whereas the same anomaly is not apparent on shorter time scales. This led them to state the following: “The real conundrum is the complex behavior of the observations”.

The same calculations were performed in Klotzbach-2009 with an amplification factor for the warming at higher altitudes, which obviously lead to even greater differences. They used an average amplification factor of 1.2 which the authors had obtained from Ross McKitrick, via data from a GISS-ER climate model study (copied from an FTP server). This trick caused some turmoil, see e.g. this RealClimate blog post. Gavin Schmidt (responsible for the GISS climate model) came to entirely different conclusions based on his model, namely an amplification factor of 0.95 on average over land and 1.4 over oceans (instead of the 1.2 for the globe as a whole used by Klotzbach-2009). In 2010 a correction was published with respect to the original paper, in which the calculations were repeated with an amplification factor of 1.1 above land and 1.4 above the oceans. Oddly, the value of 0.95 as calculated by Schmidt (see this e-mail exchange) was not used.

Sometimes Klotzbach-2009 is used in climate skeptic texts to indicate that there are problems with the measurements of surface temperatures and that these measurements cannot be trusted. For instance, the skeptical Dutch journalist Marcel Crok used the trend of 0.15 °C/decade from figure 4 in his book “De staat van het klimaat”. For Dutch readers, see this review of this part of Crok’s book on ‘PCCC klimaatportaal’, which contains the following quote [my translation]:
“If there were a ‘bias’ in the surface measurements, then satellite and surface measurements should deviate more and more in the course of time, shouldn’t they? […] and indeed this appears to be the case. Over land, the difference between the temperature measurements and the satellite measurements has risen to 0.5 degrees over the past thirty years. Whilst climate researchers expect the opposite.”
His 0.5 degrees is presumably based on three decades times 0.15, rounded up to 0.5 °C; note that only the highest number in the table of Klotzbach-2009 (the above Fig. 4) has been used. The two types of measurements over land deviate, and according to Gavin Schmidt, surface temperatures over land are expected to be slightly larger than tropospheric temperatures. Moreover, Crok omits the possibility that there could also be a ‘bias’ in the satellite measurements.

Anthony Watts also references to Klotzbach-2009, he writes in his 2012 paper based on his Surface Stations Project:
By way of comparison, the University of Alabama Huntsville (UAH) Lower Troposphere CONUS trend over this period is 0.25°C/decade and Remote Sensing Systems (RSS) has 0.23°C/decade, the average being 0.24°C/decade. This provides an upper bound for the surface temperature since the upper air is supposed to have larger trends than the surface (e.g. see Klotzbach et al (2011). Therefore, the surface temperatures should display some fraction of that 0.24°C/decade trend. Depending on the amplification factor used, which for some models ranges from 1.1 to 1.4, the surface trend would calculate to be in the range of 0.17 to 0.22, which is close to the 0.155°C/decade trend seen in the compliant Class 1&2 stations.
Here an amplification factor of 1.1 to 1.4 (from the Klotzbach-2010 correction paper) is used, while for over land this should be a bit lower than 1. After all, the USA should be categorized under ‘land’ and not under ‘ocean’. Phrases like these, not supported by valid research, should, in my humble opinion, not be published.

4 years later in 2013

2012 has passed and now we have 4 more years of data than back in 2009. Not much has changed regarding the surface temperatures, except some small improvements in the homogenization and the integration of more measurement stations (e.g. HadCRUT4 instead of HadCRUT3). These changes have not resulted in significantly different trends for the global temperatures. The transition from GHCN-M (Global Historical Climatology Network-Monthly as used for e.g. the NCDC temperatures) from version 3.1 to version 3.2 has even led to a higher trend for the land temperatures (having changed from 0.94 °C/century to 1.11°C/century for NCDC for example), with the largest differences before 1970. In addition, the surface temperatures over land of the three aforementioned institutes,  are consistent with the temperature results from the BEST project led by Richard Muller.

So, the trends in surface temperatures have not changed dramatically, though may have slightly increased in some cases. If the climate skeptic fans of the Klotzbach-2009 article are correct regarding there being strong biases in the surface temperatures, these temperature differences should therefore have increased even further.

Anyone with a spreadsheet program and an internet connection is able to redo the calculations of Klotzbach-2009 and verify whether the expected temperature difference has increased. This also applies to me and my results are shown in figure 5. The uncertainty is calculated according to the method as described in Foster & Rahmstorf 2011.

Figure 5: Differences in trends between NCDC/HadCRUT4 and UAH/RSS over 1979-2012.

We now have 13% more data and the trend difference over land between NCDC and UAH has decreased with approximately 33% from 0.15 to 0.10 °C/decade. This would translate in a divergence between surface and tropospheric trends of 0.34 °C over 34 years; less than Crok’s estimate of a few years ago.
Still, over land the trend in surface temperatures is significantly larger than in satellite temperatures. Over the oceans, the GISS model results indicate that the surface should warm significantly more than the troposphere (by a factor of 1.4); this is not seen in the observations, so also for those areas an interesting puzzle remains.
The HadCRUT4 data indicate greater differences than the old HadCRUT3 data and are, as expected, much more in line with the NCDC data.
The math in the book of Marcel Crok now generates a difference in surface minus tropospheric trends of 0.24-0.34 °C over 34 years. The average difference is 0.3 °C and is significantly lower than the 0.5 °C he mentioned in his book.

Conclusion

The expected increase in the differences between the surface temperatures and the satellite temperatures over land has not occurred. To the contrary, 13% more data show that the trend difference over land has decreased by 18% for NCDC/RSS and by 33% for NCDC/UAH.
The trend difference has thus decreased rather than increased, and whereas Klotzbach-2009 and its fans champion the argument that this difference is due to a ‘bias’ in the surface temperature record (some 1700 words were spent on this argument in K-2009), a ‘bias’ in the satellite record may be equally, or perhaps even more plausible.

Be warned if in blog pieces or papers the amplification factors between the temperatures at the surface and higher up in the troposphere are (improperly) used.

Many thanks to Bart for helping me out with the translation and for constructive comments. A reply to this blog post (or rather, to a google translation of the original Dutch post) by Klotzbach’s co-author John Christy has been posted by Marcel Crok and reposted at WUWT. A reply to John Christy is forthcoming.

For more information, see:

- The Idso’s on CO2Science
- Blog Roger Pielke Jr.: here, here and here
- ClimateAudit
- SkepticalScience: here and here
- RealClimate: here and here
- Press release Po-Chedly en Fu
- MET Office HadAT2 vs MSU
- Critiques and discussions: here, here, here and here
- The IPCC about water vapor and lapse rate

ClimateDialogue.org
Exploring different views on climate change

Goal of ClimateDialogue.org
ClimateDialogue.org offers a platform for discussions between invited climate scientists on important climate topics that have been subject to scientific and public debate. The goal of the platform is to explore the full range of views currently held by scientists by inviting experts with different views on the topic of discussion. We encourage the invited scientists to formulate their own personal scientific views; they are not asked to act as representatives for any particular group in the climate debate.

Obviously, there are many excellent blogs that facilitate discussions between climate experts, but as the climate debate is highly polarized and politicized, blog discussions between experts with opposing views are rare.

Background
The discovery, early 2010, of a number of errors in the Fourth IPCC Assessment Report on climate impacts (Working Group II), led to a review of the processes and procedures of the IPCC by the InterAcademy Council (IAC). The IAC-report triggered a debate in the Dutch Parliament about the reliability of climate science in general. Based on the IAC-recommendation that ‘the full range of views’ should be covered in the IPCC-reports, Parliament asked the Dutch government ‘to also involve climate skeptics in future studies on climate change’.

In response, the Ministry of Infrastructure and the Environment announced a number of projects that are aimed to increase this involvement. Climate Dialogue is one of these projects.

Topics
We are starting Climate Dialogue with a discussion on the causes of the decline of the Arctic Sea Ice, and the question to what extent this decline can be explained by global warming. Also, the projected timing of the first year that the Arctic will be ice free will be discussed. With respect to the latter, in its Fourth Assessment Report in 2007, IPCC anticipated that (near) ice free conditions might occur by the end of this century. Since then, several studies have indicated this could be between 2030-2050, or even earlier.

We invited three experts to take part in the discussion: Judith Curry, chair of the School of Earth and Atmospheric Sciences at the Georgia Institute of Technology; Walt Meier, research scientist at the National Snow & Ice Data Center (NSIDC) in Boulder, Colorado; and Ron Lindsay, Senior Principal Physicist at the Polar Science Center of the University of Washington in Seattle.

Future topics that will be discussed  include: climate sensitivity, sea level rise, urban heat island-effects, the value of comprehensive climate models, ocean heat storage, and the warming trend over the past few decades.

Our format
Each discussion will be kicked off by a short introduction written by the editorial staff, followed by a guest blog by two or more invited scientists. The scientists will start the discussion by responding to each other’s arguments. It is not the goal of Climate Dialogue to reach a consensus, but to stimulate the discussion and to make clear what the discussants agree or disagree on and why.
To round off the discussion on a particular topic, the Climate Dialogue editor will write a summary, describing the areas of agreement and disagreement between the discussants. The participants will be asked to approve this final article, the discussion between the experts on that topic will then be closed and the editorial board will open a new discussion on a different topic.

The public (including other climate scientists) is also free to comment, but for practical reasons these comments will be shown separately.

The project organization consists of an editorial staff of three people and an advisory board of seven people, all of whom are based in the Netherlands. The editorial staff is concerned with the day-to-day operation of researching topics, finding participants for the discussion and moderating the discussions between the experts. The main task of the advisory board is to guard the neutrality of the platform and to advise the editorial staff about its activities

Editorial Staff
Project leader is Rob van Dorland of the Royal Netherlands Meteorological Institute (KNMI). Van Dorland is a senior scientist and climate advisor in the Climate Services section and is often operating at the interface between science and society.

The second member is Bart Strengers. He is a climate policy analyst and modeler in the IMAGE-project at the PBL Netherlands Environmental Assessment Agency (PBL) and has been involved in the discussion with climate skeptics for many years.

The third member is Marcel Crok, an investigative science writer, who published a critical book (in Dutch) about the climate debate.

Questions
We welcome comments on this blog and are happy to answer any questions regarding this project. You can send an email to info [at] climatedialogue [dot] org.

Postscript (Bart V):

(Disclaimer: I am involved in this initiative as a member of the advisory board)

I think ClimateDialogue is a unique project in both its organization (people with wildly different views are involved) and in its aim: Facilitating a public discussion between scientists with strongly differing opinions.

Discussion topics are chosen to be relevant and interesting to the general public as well as receiving scientific attention. Discussants are chosen to reflect different stances in the spectrum of scientific opinion, explicitly including ‘sceptical’ voices. Naturally, the ensuing discussion is not necessarily representative of the full spectrum of scientific discussion (painting it as such would likely lead to a ‘false balance’).

The idea is that the discussion can alleviate the polarization between ‘sceptics’ and ‘mainstreamers’ and provide some clarity in background of the (dis)agreements. Moreover, having scientists discuss their scientific disagreements in a public setting can go a long way to increase the public trust in science, which has suffered from the (imho incorrect) impression of being closed-minded. All in all, I think that ClimateDialogue provides a valuable service to both the public and the scientific debate. That doesn’t mean that it’s free of risks, but these are more in the framing and the perception than in the discussions itself. Naturally, the participation of good scientists is a necessary condition to make this experiment a success. Don’t hesitate to contact the editors (or me) if you fit the bill and are not afraid of a public debate!

The objective of this study is to gain insight into how climate scientists perceive the public debate on the physical scientific aspects of climate change. More info about the survey was posted on the PBL website at the time, which has recently been updated to include a link to the survey questionnaire. Please note that the survey is no longer active.

Some confusion has arisen over the status of this survey. I responded at WUWT in an attempt to clarify:

We undertook a survey in March/April of this year (which, as Hans Labohm mentioned in a comment on WUWT, had been previewed by a variety of people with different viewpoints). Some respondents, e.g. Timothy Ball, asked to see the questions again. After internal consultation, we decided to publish the survey questions on the institute’s website, so that they are viewable to all. We contacted the survey respondents to inform them of the questions being available to view. I informed Dr Ball of this as well, to follow-up on my earlier email to him.

Our email to all respondents, informing them of the fact that the survey questions are available on the web, was apparently misunderstood to mean that we were again soliciting responses to a survey; this is however not the case. Roger Pielke Sr had already put a notice about the survey on his blog, which he has since updated after an email clarifying that this is an inactive survey, to which he had previously responded.

Below we (Bart Verheggen and Bart Strengers) reply to some of the more substantive questions regarding the survey questions raised on WUWT. However, we will not discuss results or the survey sample at this point in time. We will do so when our manuscript has been accepted.

Referring to question 1a

1a. What fraction of global warming since the mid-20th century can be attributed to human induced increases in atmospheric greenhouse gas (GHG) concentrations?
More than 100% (i.e. GHG warming has been partly offset by aerosol cooling)
Between 76% and 100%
Between 51% and 76%
Between 26% and 50%
Between 0 and 25%
Less than 0% (i.e. anthropogenic GHG emissions have caused cooling)
There has been no warming
Unknown due to lack of knowledge
I do not know
Other (please specify)

Ken Harvey asks (46):

Let’s say that I believe that the correct answer is 0. If I tick that box I immediately lump my opinion in with those who think that 25% is the appropriate answer, despite the world of difference between our positions. I am tempted to tick the next box down indicating less than 0% and I may, or may not give in to that temptation. Let’s sat that I believe that the correct answer is 25%. I face a similar problem – I don’t want to be lumped in with the fellow who thinks that the answer is 0. I am tempted to tick the higher box.

This is indeed a dilemma. An alternative which we considered is to ask the respondent to provide a percentage themselves (i.e. as an open numeric question). This however would ‘force’ the respondent to provide a number which, according to many, would create the impression of much higher certainty than there is about this estimate. (Can we really distinguish whether this contribution is 81 or 82%? [or 2 or 3% if you wish]). It is clear that both options have pros and cons, but we believe that by making enough –but not too much- ranges available, we obtain relevant information about the respondents’ thoughts.

Robert of Ottawa asks (66):

Question 2A asks: Has the trend in global average temperature changed in the past decade, compared to the preceding decades?
and offers the possible answer: The trend over the past decade is negative (i.e. cooling)
But question2b asks: What is your interpretation of the trend over the past decade with respect to the long term (multi-decadal) trend?
and does not offer the possible answer from 2A: “The trends are of natural cause”. Instead, it offers variants for Warmistas or don’t know.

Question 2 doesn’t go into causation (be it anthropogenic or natural); 2b is only about what the respondent thinks is happening to global temperatures on longer timescales. Eg the answer option to 2b “Long-term warming trend has changed as indicated by my previous answer” could be chosen to reflect the opinion that temps are cooling also on longer timescales.

A. Scott asks (72):

Would you share a bit more detail about what the goal of your survey is – what you hope to find from it.

The objective of this study is to gain insight into how scientists, who have published on global warming, perceive physical science issues, which are frequently debated in the public domain. E.g. by investigating the extent to which scientist agree or disagree about these issues (both the big picture issues and the detailed aspects)? How are these responses related to one another? What can we learn from that?

DaveA asks (79):

But still, this only covers questions pertaining to the direct influence/existence of AGW. Where’s the question asking how many climate refugees will eventuate from climate change? Food shortages, extreme weather fatalities, climate anxiety… so often that’s where the disagreement starts and where the label “denier” is pulled out even for those who have accepted an anthropogenic influence.

As explained on the PBL website regarding this survey, we decided to focus on physical science aspects of the public debate (mostly with a ‘skeptical’ signature): These ‘IPCC Working Group I’ topics are a focal point in the public debate and they form the foundations for further deliberation; for example, regarding impacts or response strategies. We chose to be as complete as realistically possible in covering the physical science aspects, acknowledging that that meant we could not include other aspects such as you mention.

Tim Welham asks (87):

This questionnaire is biased at Q1 and shows little professional attention to detail. I haven’t bothered to go further. As Ken Harvey points out in Q1 (we get no further) we see lower answer categories of:
Between 0 and 25%
Less than 0% (i.e. anthropogenic GHG emissions have caused cooling)
The design of this questionnaire forces anyone replying into categories which may not reflect their real view. There is absolutely no reason why a ‘Percentage: Please write in’ could be used to give the respondent an accurate reply. Then the final category ’caused cooling’ is junk. What can be less than 0%?. If it is 0% why should it cause anything in terms of cooling or heating?

More people may deem the survey too detailed rather than not detailed enough. For deliberations regarding an open question or offering ranges as answer options, please see above in our reply to Ken Harvey. There is no right or wrong there, but multiple options with each their specific pros and cons. However, by offering a wide range of answer options (even providing the option of answering that greenhouse gases cause cooling (i.e. a negative percentage)), we believe we avoided a bias a much as possible. One could actually make the argument that the answer options for question 1a have an inherent bias in the other direction, since the option that GHG by themselves are responsible for more than the observed warming (i.e. >100%; see e.g Huber and Knutti, 2011) has not been subdivided into multiple ranges.

Bart Verheggen and Bart Strengers