In what could prompt another shift in scientific human resources, there are reports (H/T Wired Science) of a robot actually forming and testing hypotheses.  Working in the baker’s yeast genome, the robot Adam worked to fill in gaps in understanding of the yeast’s metabolism.  By scanning a database of similar genes and enzymes from other organisms, Adam utilized algorithms to determine possible genes in the yeast genome that would correspond to orphan enzymes – enzymes that had no known gene coding for them.  After forming an hypothesis, Adam would conduct the experiment, analyze the data, and refine the hypothesis.  The robot’s designers recently confirmed by hand Adam’s discovery of three genes that coded for an orphan enzyme.

At the moment, designing these robots will continue to be a specialized affair – one robot for a particular area of research will likely be different from a robot for a different area of research.  But if there is the potential of standardizing, or at least spreading, such robot development, how and where science can be conducted could change dramatically.

Once again, some of the routine tasks of science will not require people.  There are plenty of potential consequences of this shift, especially since many of those affected would hold Ph.Ds.  Lab heads won’t need as many people on staff, reducing the need to import scientific talent and reducing the positions available in what is the traditional, apprentice-like, guild nature of scientific training.  Given the near obsessive focus on numbers in the decades-long arguments over the scientific pipeline, a potential policy outcome of such a shift would be to claim the problem of scientific talent is solved.  That may not reflect reality, but that wouldn’t be the first time.

On the other hand, this kind of automation has the potential to dramatically decrease the costs of research.  If certain kinds of research no longer need as many people, the costs of their training, salary, and benefits decline or disappear.  Start-up costs for junior researchers and professors decline.  This would allow for research to be conducted at institutions and in places where resources are not as flush.  Running a small lab out of a state school or in an emerging market may not carry as many challenges (or as much stigma) as it does now.

This is, of course, speculative.  But if the policy challenges of emerging science and technology are to be effectively addressed, some speculation will be required to meet the changes as they happen, rather than after it’s been too long to address them.

My first reaction upon seeing this story was that someone was having some fun. But it doesn’t seem like benthic bacteria . . . So this article from the AFP comes as a surprise, and a reminder that forecasting the future remains a perilous business. With news like this, it seems premature to dismiss skepticism about climate science as fading away, far from it, expect skeptics of all sorts to have a bit more bounce in their steps.

DHAKA (AFP) – New data shows that Bangladesh’s landmass is increasing, contradicting forecasts that the South Asian nation will be under the waves by the end of the century, experts say.

Scientists from the Dhaka-based Center for Environment and Geographic Information Services (CEGIS) have studied 32 years of satellite images and say Bangladesh’s landmass has increased by 20 square kilometres (eight square miles) annually.

Maminul Haque Sarker, head of the department at the government-owned centre that looks at boundary changes, told AFP sediment which travelled down the big Himalayan rivers — the Ganges and the Brahmaputra — had caused the landmass to increase.

The rivers, which meet in the centre of Bangladesh, carry more than a billion tonnes of sediment every year and most of it comes to rest on the southern coastline of the country in the Bay of Bengal where new territory is forming, he said in an interview on Tuesday.

The United Nations Intergovernmental Panel on Climate Change (IPCC) has predicted that impoverished Bangladesh, criss-crossed by a network of more than 200 rivers, will lose 17 percent of its land by 2050 because of rising sea levels due to global warming.

The Nobel Peace Prize-winning panel says 20 million Bangladeshis will become environmental refugees by 2050 and the country will lose some 30 percent of its food production.

Director of the US-based NASA Goddard Institute for Space Studies, professor James Hansen, paints an even grimmer picture, predicting the entire country could be under water by the end of the century.

But Sarker said that while rising sea levels and river erosion were both claiming land in Bangladesh, many climate experts had failed to take into account new land being formed from the river sediment.

“Satellite images dating back to 1973 and old maps earlier than that show some 1,000 square kilometres of land have risen from the sea,” Sarker said.

“A rise in sea level will offset this and slow the gains made by new territories, but there will still be an increase in land. We think that in the next 50 years we may get another 1,000 square kilometres of land.”

Mahfuzur Rahman, head of Bangladesh Water Development Board’s Coastal Study and Survey Department, has also been analysing the buildup of land on the coast.

He told AFP findings by the IPCC and other climate change scientists were too general and did not explore the benefits of land accretion.

“For almost a decade we have heard experts saying Bangladesh will be under water, but so far our data has shown nothing like this,” he said.

I have an invited op-ed in today’s Financial Post (a Canadian newspaper with a skeptical editorial perspective on climate change). I argue that even though many scientists oversell the predictive capabilities of climate models, action on climate change still makes sense. Here is an excerpt:

So in the debate on what to do about climate change, what are we to make of the overstated claims of predictive accuracy offered by many scientists?
Not surprisingly, the reason for overstated claims lies in the bitter and contested politics of climate change. Myanna Lahsen, an anthropologist who has studied climate modelers, finds that many of these scientists are acutely aware of the fact that any expressed “caveats, qualifications and other acknowledgements of model limitations can become fodder for the anti-environmental movement.” She documents how, more than a decade ago, a prominent climate scientist warned a group of his colleagues at the National Center for Atmospheric Research, home of one of the main U.S. climate modeling efforts that informs the IPCC, to “Choose carefully your adjectives to describe the models. Confidence or lack of confidence in the models is the deciding factor in whether or not there will be policy response on behalf of climate change.”

I witnessed this dynamic in practice while I was waiting to testify on climate policy before the U.S. Congress in 2006. A prominent climate scientist testifying on the panel appearing before mine was asked by a member of Congress about uncertainties in predictions from climate models. The scientist replied, enthusiastically and accurately, that there are a range of important uncertainties coming from scenario inputs and choices in parameterization schemes, instantly overwhelming his congressional audience with technical detail. Much later, and after a long break, the scientist requested an opportunity to clarify his earlier comments, and this time he said, “I would like to give you a little more direct answer to the question on reliability of climate models. I think they are reliable enough to be a very useful guide into the future.”

Lost in the Manichean debate over climate change is the real significance of what climate models really are telling us: We should act on climate mitigation and adaptation not because we are able to predict the future, but because we cannot.

See it all here. Comments and reactions welcomed.

If these guys know what is “wrong” then they must have an idea about what is “right”.

Real Climate writes an entire post responding to Steve McIntyre’s recent discussions of buckets and sea surface temperatures, explaining why the issue doesn’t really matter, but for some weird reason they can’t seem to mention him by name or provide a link to what they are in fact responding to. (If the corrections don’t matter, then one wonders, why do them? Thompson et al. seemed to think that the issue matters.)

Real Climate does seem have mastered a passive voice writing style, however. Since they did have the courtesy to link here, before calling me “uninformed” (in deniable passive voice of course), I though a short response was in order.

Real Climate did not like our use of a proposed correction suggested by He Who Will Not Be Named. So Real Climate proposed another correction based on a graphic printed in The Independent. Never mind that the correction doesn’t seem to jibe with that proposed by Thompson et al., but no matter, we used the one suggested by Mr. Not-To-Be-Named so lets use Real Climate’s as well and see what difference it makes to temperature trends since 1950. Based on what Real Climate asserts (but oddly does not show with numbers), you’d think that their proposed adjustment results in absolutely no change to mid-20th century trends, and indeed anyone suggesting otherwise is an idiot or of ill-will. We’ll lets see what the numbers show.

The graph below shows a first guess at the effects of the Real Climate adjustments (based on a decreasing adjustment from 1950-60) based on the graphic in The Independent.

What difference to trends since 1950 does it make? Instead of the about 50% reduction in the 1950-2007 trend from the first rough guess from you-know-who, Real Climate’s first guess results in a reduction of the trend by about 30%. A 30% reduction in the IPCC’s estimate in temperature trends since 1950 would be just as important as a 50% reduction, and questions of its significance would seem appropriate to ask. But perhaps a 30% reduction in the trend would be viewed as being “consistent with” the original trend

Try again Real Climate. And next time, his name is STEVE MCINTYRE — and his blog is called CLIMATE AUDIT. There is a lot of science and civil discussion there, with a healthy mix of assorted experts and a range of ordinary folks. Questioning scientific conclusions is a lot healthier for science than rote defense, but we all learned that in grad school, didn’t we?

The graph above shows three trend lines.

BLUE: Temperature Trend prediction from the 1990 IPCC report
RED: Temperature Trend prediction from the 2007 IPCC report
GREEN: Observed Trend for 2001-2007 (from average of four obs datasets)

All data is as described in this correspodence (PDF).

Your assignment:

Which IPCC prediction is the trend observed 2001-2007 more consistent with and why? Show your work!

You are free to bring in whatever information and use whatever analysis that you want.

“Politicians seems to think that the science is a done deal,” says Tim Palmer, “I don’t want to undermine the IPCC, but the forecasts, especially for regional climate change, are immensely uncertain”.

Palmer is a leading climate modeller at the European Centre for Medium-Range Weather Forecasts in Reading, UK, and he does not doubt that the Intergovernmental Panel on Climate Change (IPCC) has done a good job alerting the world to the problem of global climate change. But he and his fellow climate scientists are acutely aware that the IPCC’s predictions of how the global change will affect local climates are little more than guesswork. They fear that if the IPCC’s predictions turn out to be wrong, it will provoke a crisis in confidence that undermines the whole climate change debate.

The IPCC’s forecasts could be wrong in many different ways, over different time periods and spatial scales, including underestimating future changes. And it is not even clear that scientists involved with the IPCC have a collective view on what it would even mean for the IPCC to be “wrong”. As we’ve argued here often, action on climate change makes sense even if the predictions of the IPCC are not yet perfect. But this is a hard case to make when defenders of those predictions allow no room for imperfections to be seen, or questions to be asked.

I am happy to report that claims of “inconsistent with” do exist. Here is an example from a paper just out by Knutson et al. in Nature Geoscience:

Our results using the ensemble-mean global model projections (Fig. 4) are inconsistent with the notion of large, upward trends in tropical storm and hurricane frequency over the twentieth century, driven by greenhouse warming.

The climate modelers at Real Climate apparently don’t like the phrase “inconsistent with” in the context of models and try to air brush it away when they write of Knutson et al.:

. . .we know that (i) the warming [of the oceans] is likely in large part anthropogenic, and (ii) that the recent increases in TC frequency are related to that warming. It hardly seems a leap of faith to put two-and-two together and conclude that there is likely a relationship between anthropogenic warming and increased Atlantic TC activity.

Knutson et al. respond in the comments that this in fact is not how to interpret their paper, and — kudos to them — take strong, public issue with the weaselly words implying a connection that they don’t show (emphasis added in the below, and I’ve copied the whole comment for the entire context):

Mike [Mann],

Statement (i), that “the warming [of the tropical Atlantic Ocean] is likely in large part anthropogenic.” is reasonable, taking “anthropogenic” to mean “greenhouse gas”, given the work of Santer et al (2006, PNAS), Knutson et al (2006, J. Clim.), and Gillett et al (2008, G.R.L.). To quote from Gillett et al:

…our results indicate that greenhouse gas increases are indeed likely the dominant cause of [tropical Atlantic] warming…

However, statement (ii), that “the recent increases in [Atlantic] TC (tropical cyclone) frequency are related to that warming” is vague – with “related to” allowing an interpretation that includes anything from a negative relationship, to a minor contribution, to local SST warming being the dominant dynamical control on TC frequency increase. Some might interpret “related to” to mean “are dominantly controlled by”, and we think the evidence does not justify such a strong statement. In particular, the results of Knutson et al (2008) do not support such an attribution statement,if one focuses on the greenhouse gas part of the anthropogenic signal. Quoting from page 5 of the paper:

Our results using the ensemble-mean global model projections (Fig. 4) are inconsistent [emphasis added] with the notion of large, upward trends in tropical storm and hurricane frequency over the twentieth century, driven by greenhouse warming

We agree that TC activity and local Atlantic SSTs are correlated but do not view this correlation as implying causation. The alternative, consistent with our results, is that there is a causal nonlocal relationship between Atlantic TC activity and the tropical SST field. The simplest version uses the difference between Atlantic and Tropical-mean SST changes as the predictor (Swanson 2008, Non-locality of Atlantic tropical cyclone intensities, G-cubed, 9, Q04V01). This picture is also consistent with non-local control on wind shear (e.g. Latif et al 2007, G.RL.), atmospheric stability (e.g., Shen et al 2000, J. Clim.) and maximum potential intensity (e.g., Vecchi and Soden, 2007, Nature).

We view the SST change in the tropical Atlantic relative to the rest of the tropics as the key to these questions. Warming in recent decades has been particularly prominent in the northern tropical Atlantic, but such a pattern is not evident in the consensus of simulations of the response to increasing greenhouse gases. So, whether changes in Atlantic SST relative to the rest of the tropics – that according to our hypothesis have resulted in the changes in hurricane activity – were primarily caused by changes in radiative forcing, or whether they were primarily caused by internal climate variability, or (most likely) whether both were involved, is obviously an important issue, but this is not addressed by our paper

Now a word of caution — Knutson et al. 2008 is by no means the last word on hurricanes and global warming, and the issue remains highly contested, and will remain so for a long time. Of course, you heard that (accurate) assessment of the state of this particular area of climate science here a long time ago (PDF

Knutson et al. is notable because it clearly identifies observations “inconsistent with” what the models report which should give us greater confidence in research focused on generating climate predictions. We should have greater confidence because if practically everything observed is claimed to be “consistent with” model predictions, then climate models are pretty useless tools for decision making.

“Skill” is a technical term in the forecast verification literature that means the ability to beat a naïve baseline when making forecasts. If your forecasting methodology can’t beat some simple heuristic, then it will likely be of little use.

What are examples of such naïve baselines? In weather forecasting historical climatology is often used. So if the average temperature in Boulder for May 20 is 75 degrees, and my prediction is for 85 degree, then any observed temperature below 80 degrees will mean that my forecast had no skill. In the mutual fund industry stock indexes are examples of naive baselines used to evaluate performance of fund managers. Of course, no forecasting method can always show skill in every forecast, so the appropriate metric is the degree of skill present in your forecasts. Like many other aspects of forecast verification, skill is a matter of degree, and is not black or white.

Skill is preferred to “consistency” if only because the addition of bad forecasts to a forecasting ensemble does not improve skill unless it improves forecast accuracy, which is not the case with certain measures of “consistency,” as we have seen. Skill also provides a clear metric of success for forecasts, once a naïve baseline is agreed upon. As time goes on, forecasts such as those issued by the IPCC should tend toward increasing skill, as the gap between a naive forecast and a prediction grows. If a forecasting methodology shows no skill then it would be appropriate to question the usefulness and/or accuracy of the forecasting methodology.

In this post I use the IPCC forecasts of 1990, 2001, and 2007 to illustrate the concept of skill, and to explain why it is a much better metric that “consistency” to evaluate forecasts of the IPCC.

The first task is to choose a naïve baseline. This choice is subjective and people often argue over it. People making forecasts usually want a baseline that is easy to beat, people using or paying for forecasts often want a more rigorous baseline. For this exercise I will use the observed temperature trend over the 100 years ending in 2005, as reported by the 2007 IPCC, which is 0.076 degrees per decade. So in this exercise the baseline that the IPCC forecasts have to beat is a naïve assumption that future temperature increases will increase by the same rate as has been observed over the past 100 years. Obviously, one could argue for a different naïve baseline, but this is the one I’ve chosen to use.

I will also use the ensemble average “best guess” from the IPCC for the most appropriate emissions scenario as the prediction. And for observations I will use the average value from the four main group tracking global temperature trends. These choices could be made differently, and a more comprehensive analysis would explore different ways to do the analysis.

So then, using these metrics how does the IPCC 1990 best estimate forecast for future increases in temperature compare for 1990-2007? The figure below shows that the IPCC forecast, while over-predicting the observed trend, outperformed this naïve baseline. So the forecast can be claimed to be skillful, but not by very much.

A more definitive example of a skillful forecast is the 2001 IPCC prediction, which the following figure shows demonstrated a high degree of skill.

Similarly, the 2000-2007 forecast of the IPCC 2007 also shows a high degree of skill, as seen in the next figure.

But in 2008 things get interesting. With data from 2008 included, rather than ending in 2007, then the 2007 IPCC forecast is no longer skillful, as shown below.

If one starts the IPCC predictions in 2001, then the lack of skill is even greater, as seen below.

What does all of this mean for the ability of the IPCC to predict longer-term climate change? Perhaps nothing, as many scientists would claim that it makes no sense to discuss IPCC predictions on time scales less than 20 or 30 years. If so, then it would also be inappropriate to claim that IPCC forecasts on the shorter scales are skillful or accurate. One way to interpret the recent Keenlyside et al. paper in Nature is that their analysis suggests that the IPCC predictions of future temperature evolution won’t be skillful unless they account for various factors not included in the IPCC predictions.

The point of this exercise is to show that there are simple, unambiguous alternatives to using the notion of “consistency” as the basis for comparing IPCC forecasts with observations. “Consistency” between models and observations is a misleading, and I would say fairly useless way to talk about climate forecasts. Measures of skill provide an unambiguous way to evaluate how the IPCC is doing over time.

But make no mistake, the longer the IPCC forecasts lie in a zone of “no skill” — which the most recent ones (2007) currently do (for the time of the forecast to present) — the more interest they will receive. This time period may be for only one more month, or perhaps many years. I don’t know. This situation creates interesting incentives for forecasters who want their predictions to show skill.