Friday, March 27, 2009

A Layman’s Guide to Climate Modeling – Part I

Don’t trust climate models and don’t trust anyone who offers the results of such models as proof of anything related to climate. Climate models are not evidence. Specifically, predictive climate models have little more than entertainment value, at least based upon their performance to date. In this category fall the twenty-three or so models the Intergovernmental Panel on Climate Change (IPCC) refer to that predict doom and gloom for the planet due to human CO2 emissions.

I propose to look at the most commonly mentioned forcings and feedbacks and how they are treated in GCMs. Keep in mind that our ability to model a complex open system will only be as good as the data fed to the model, the assumptions used to account for things for which we have no empirical data and the modeler’s ability to describe processes in plain, spoken language prior to reducing those processes to mathematical language. All italics in the text below are mine and intended to emphasize the theoretical nature of the discussion.

The IPCC and other proponents of Anthropogenic Climate Change (ACC) use climate model predictions and simulations to state their case. They do not use empirical observations. Even the mean surface temperatures calculated by Goddard Institute of Space Studies (GISS), the National Oceanic and Atmospheric Administration (NOAA), and Hadley Centre Climactic Research Unit (HadCRU) are the output of models. Therefore understanding what GCMs are is important to understanding why they are not evidence of ACC.

About Climate. Climate is the term used to describe weather, averaged over a long period of time, usually about thirty years.1 This is the simple definition most often used by both the Anthropogenic Climate Change (ACC) proponents and Skeptics. However, that definition inadequately describes the complex open system of global energy exchange. That is what climate really boils down to. The term climate is a way to describe a long-term, complex thermodynamic process on a regional or global scale.

About climate models. Commonly called GCMs, or Global Climate Models, they are also referred to more accurately as General Circulation Models, but some literature also uses the term Global Circulation Models.2 Climate models are complex mathematical approximations of the climate system. Their predictive ability is limited by how much we understand about each component of the total system, by our ability to empirically measure change within the climate system, and the extent of those recorded changes (how many years of data we have collected).

Some modeling terms. Two of the most common terms used in the discussion of climate models are Radiative Forcing (RF) and Climate Feedbacks (CF) or Feedback Mechanisms (FM).

Radiative Forcing (RF) is a concept used for quantitative comparisons of the strength of different human and natural agents in causing climate change.3 These values are not observed measured values, but are derived based upon assumptions, indirect observations of climate components and assumed correlation/causal relationships between components. Radiative forcing is reported in the climate change scientific literature as a change in energy flux at the tropopause, calculated in units of watts per square meter (W/m2); model calculations typically report values in which the stratosphere was allowed to adjust thermally to the forcing under an assumption of fixed stratospheric dynamics.4 RF is the dominant feature of climate models. It is these derived, not measured, values that drive the model outputs.

The fundamental assumption underlying the radiative forcing concept is that the surface and the troposphere are strongly coupled by convective heat transfer processes; that is, the earth-troposphere system is in a state of radiative-convective equilibrium. The concept of radiative forcing is based on the hypothesis (not even a theory) that the change in global annual mean surface temperature is proportional to the imposed global annual mean forcing, independent of the nature of the applied forcing.5 The above stated hypothesis has not been proven. RF values, frequently cited in climate discussions, are essentially assumptions.

Feedbacks are processes in the climate system that can either amplify or dampen the system’s response to changed forcings. A substantial part of the uncertainty in projections of future climates is attributed to inadequate understanding of feedback processes internal to the natural climate system. 6 Feedbacks are not well understood and this fact leads to assumptions in modeling that affect model output. Most models minimize or ignore feedbacks and this degrades the model’s predictive ability.

In my next post I will begin discussing the individual forcings and feedbacks used/ignored by GCMs.

3. IPCC Assessment Report Four (AR4), Chapter 2 (Changes in Atmospheric
    Constituents and Radiative Forcings), page 131, (2007).
4. Radiative Forcing of Climate Change: Expanding the Concept and
    Addressing Uncertainties, page 15, (NRC, 2005).
5. ibid, page 19.
6. Understanding Climate Change Feedbacks, p.1, (NRC, 2003).

Thursday, March 19, 2009

Ice Capades in the Arctic

UPDATE 1: They're making progress. Their current rate of travel (as of 27 March 2009) is 3.63km per day. If they maintain this pace, they will reach the end of their journey in about 227 days on 9 November 2009. They've knocked a whole year off the adventure.

There’s a great adventure afoot up in the Arctic. Three intrepid explorers in search of future grant money are braving the elements and the dangers of shifting ice on a 1,000km 100-day publicity stunt, complete with NASCAR-like sponsor logos plastered to clothing and equipment, in order to reach the North Pole. Oh, and along the way they are supposedly measuring ice thickness. We’ll get back to that a little later.

So far they have discovered the following things in 18 days of moving almost nowhere:
It’s frigging cold up there!
The ice moves – sometimes many miles overnight while they are sleeping – resulting in being drifted further from their goal and having to make up that lost ground…er…ice the next day.

So far, due to the above two items, they are averaging 1.5km per day, instead of the 10km per day they intended. If they can keep up this pace, they should complete their adventure, oh around 28 December 2010. That’s just a bit longer than 100 days, so I suspect that either funding or their endurance will fail long before then.

According to the expedition web site Catlin Arctic Survey, all three are experienced Arctic explorers who have previously been to the North Pole. And yet, when reading their daily posts, one notes that they are surprised to wake up in a different place than when they bedded down for the night. They are surprised to find that their tents fill with ice crystals and it’s hard to get warm. One must wonder if, on their previous expeditions, they were simply hanger’s on rather than integral members of the team. One must also wonder if they made their previous trips on foot, as they are doing now, or via some automated conveyance.

Among the top three sponsors for this publicity stunt are:

Catlin Group Limited, a leading global specialty insurer and reinsurer writing more than 30 classes of business. We provide creative risk management solutions and excellent financial security to clients worldwide. (Any bets about their “Climate Change Mitigation” lines of insurance?)

ECX, the premier marketplace for trading carbon emissions, providing the focal point for the majority of trading in the recently developed emissions or carbon markets.

Unlike the AGW crowd who likes to claim that sponsorship taints the science, I’m simply pointing out the hypocrisy of such claims by the AGW crowd. The top two sponsors for this expedition have a vested interest in AGW hysteria. By the AGW line of argument, whatever this survey finds is tainted due to the funding. Most of the gadget sponsors can be excused because their sponsorship of such an expedition is more along the lines of proving their equipment in an extreme environment.

In reality the value, or lack thereof, of this expedition should be in the data collection, but even before they left the starting blocks their results were likely biased. As one looks at their web site, it becomes readily apparent that they are not following the scientific method.

Let us review that method (the steps are listed in the order in which they should occur):

1. Ask a question. The question should be about something that can be measured.

2. Do background research. No sense reinventing the wheel.

3. Formulate a hypothesis. The hypothesis attempts to explain how things work. It should be stated in way that is measurable, falsifiable and which answers the original question. By falsifiable we mean there are measures that, if observed, will negate our hypothesis.

4. Test the Hypothesis through experimentation and observation. This is where we do our measurements. It is also important to run our tests/observations many times.

5. Analyze the data collected through experimentation and observation.1 This is where we look for patterns that either fit or negate our hypothesis.

6. Draw a conclusion. Having analyzed the data, we conclude one of the following three things:
a. the data supports our hypothesis
b. the data does not support our hypothesis (our hypothesis is falsified)
c. the data is ambiguous - it neither supports nor negates our hypothesis.

If the data is ambiguous, then we must re-examine the assumptions and methodology used in our experiment. In other words, go back to the drawing board. This happens more often in science than most lay people realize.

7. Communicate results. We tell the world what we found out. This is where the fun begins in real science, as opposed to most climate science. In real science, the scientist shares his methodology, data, and computer code so that other interested scientists can attempt to replicate the results. In addition to replication, these other interested scientists look for methodological flaws and erroneous assumptions that might invalidate the study. I’ll explore why many climate studies are not real science in a future post, but for now let’s examine why this particular publicity stunt is bogus science.

From the Catlin Arctic Survey web site, it appears that this is the question they are attempting to answer: “How long will the Arctic Ocean's sea ice cover remain a permanent feature of our planet?”

Inherent in this question, however, is a bold assumption that is already falsifiable. They assume that Arctic Sea ice is a permanent feature of our planet. Evidence shows that for about 96% of the Earth’s 4-plus billion years of existence, there was NO ICE ANYWHERE. In rephrasing their question, I’d drop the word “permanent.”

Their background research seems to be based solely on accepting the Intergovernmental Panel on Climate Change reports, all of which rely heavily on climate models for their findings and Al Gore’s “An Inconvenient Truth.” They even emphasize the plight of polar bears.

Which brings us to their hypothesis: the Arctic Sea Ice is melting and this melting is human-caused.

This hypothesis is based upon computer models, not observation. In real science we formulate a hypothesis to explain empirical observations.

As they state on their web site: “The most frequently cited date for the seasonal disappearance of the Arctic Ocean's sea ice is the UN Intergovernmental Panel on Climate Change’s 2050-2100, based on the known rate of its shrinking surface area, and the IPCC’s long-range global climate forecasts." (computer models)

A problem with these IPCC models is that we really don’t have a long-term “known rate” of shrinking surface area. We have less than thirty years of data and in the last two years, sea ice extent seems to be recovering…growing.

In other words, at the moment empirical observations are at best ambiguous.

Unfortunately, they’ve already reached their conclusion before they even set out on their expedition or taken a single measurement:

“The melting of the sea ice will accelerate climate change, sea level rise and habitat loss on a global scale. Its loss is also a powerful indicator of the effects of human activity on our planet’s natural systems and processes. The Survey’s scientific findings will be taken to the national negotiating teams working to replace the Kyoto Protocol agreement at the UN Climate Change Conference of Parties in Copenhagen in December 2009.”

Their opening sentence is already scientific nonsense. Sea ice floating in water has already displaced all the water it will displace. Melting the sea ice will not raise the sea level.

So they’ve constructed an experiment whereby they will take ice measurements. However, as there is no other data against which to compare their data, there is no way to falsify their hypothesis with the single data set they collect.

“The team will be travelling on foot, hauling sledges from 81°N 130°W, across 1000-km of disintegrating and shifting sea ice…” (bold mine)

“The Catlin Arctic Survey has developed and tested (more likely one of their sponsors) a portable, ice-penetrating radar. This will take continuous and detailed measurements of both the snow and ice layers along the 1000 km route.”

The measurements will be continuous only for the duration that they are on the ice and shifting ice somewhat negates the level of possible detail. Sea ice at a given location changes daily due to wind and ocean current.

On their web site they emphasize that the climate modellers rely on the ice data rather than snow data. This totally ignores the fact that as the snow accumulates it compresses, melts and freezes, and becomes ice. Snow also reflects sunlight, which affects the rate of melt and overall global warming.

“Despite the technological advances of the 20th century, we still only have estimates of the thickness of the sea ice cover on the Arctic Ocean. Travelling across the sea ice, the Catlin Arctic Survey team will take precise measurements of its thickness and density. This will enable the programme’s Science Partners to determine, with a greater degree of accuracy, how long the sea ice will remain. Currently, its predicted meltdown date is anywhere between four and a hundred years from now.”

Nowhere is there any discussion of return trips or continuous measurements over the next several years. They are taking one set of measurements. They are, in effect, creating the cherries they intend to pick later. (Cherry-picking data.) Their "science partners" are the very climate modellers and the IPCC who have a vested interest in AGW hysteria.

“Evidence for the earlier meltdown date would provide fresh impetus to resolve through international agreements the more sustainable and responsible management of the increasingly accessible natural resources, revealed as the ice recedes. The survey will assist scientists in providing policy-makers with higher resolution forecasts than are made to date, which in turn will facilitate decisions where previously indecision has existed.”

Not only did they reach conclusions prior to taking any measurements, they tip their hand as to the purely political component of what they are doing. They are not collecting data to better understand climate processes in the Arctic, they are collecting data to drive policy in a particular direction.

In a nutshell, here is the Catlin Arctic Survey scientific method:

1. Draw Conclusions. Our results will help us formulate carbon limits and trading schemes in developing the Kyoto replacement accord.
2. Formulate hypothesis. Accept and parrot the AGW consensus.
3. Do background research. Accept and parrot computer model projections quoted by a political organization that is promoting a specific political agenda.
4. Ask a question. Phrased in a way that our pre-ordained answer will scare the public and politicians so that we get more money.
5. Test the hypothesis. Gather dubious data absent any falsifiability criteria or multiple iterations.
6. Analyze the data. Feed said data into supercomputer models to “refine” the pre-ordained conclusion of those earlier models.
7. Report Results. The ice will melt sooner rather than later, so give us more money so we can get a firm date of the ice melting.

This is not science.

This is propaganda.

This is a publicity stunt.

1. In most explanations of the scientific method, this step and the following step are combined. I separate them here for clarity.