Skip to main content

How Accurate Are Climate Change Predictions, Really?

Conversation and Challenge

"What would it take,” I asked fellow writer and climate skeptic jackclee, “to convince you that we have a climate problem?”

We’d been having an extended discussion in the comments of my article, Climate Change: How Much Time Do We Have? Jack responded:

Doc, the one evidence I need is for the various climate models to agree with reality. There projections has consistently over estimated the temperature rise. I just don't trust them considering how the models have such variables which are based on assumptions and the small tweak can cause large changes in the model outputs. In a few short years, we will see if these models are for real or they are contrived. Please revisit this in a few years. Take care.

Though marked by Jack’s customary civility, it was a frustrating response for me, as the article in which the conversation was occurring had gone to considerable lengths to examine how much time remains to us to take action on climate, concluding that in certain respects it is already too late. The least misleading answer to the question “How much time do we have?” was, I wrote:

None, really. We are late, and we just need to work not to get any later.

That was true, I had said:

  • Because it’s already too late for the victims of climate change to date.
  • Because climate change is insidious--as with tobacco smoking, the damage is often done before symptoms are evident.
  • Because global carbon emissions still seem to be increasing.
  • Because we are running out of time to avoid what is generally considered ‘dangerous’ warming.

So I tried to address Jack’s concern directly, citing data and discussions that show that, in fact, the observations are consistent with IPCC projections of temperature, as linked in the sidebar below.

Original Model-Observation Comparison Graphs


Updated Comparison Graph, August 2016

Updated observation-model comparison, August 2016.  Graphs courtesy of RealClimate.

Updated observation-model comparison, August 2016. Graphs courtesy of RealClimate.

BEST update, to February 2017

Updated model-observation comparison, courtesy of Dr. Zeke Hausfeather of BEST.

Updated model-observation comparison, courtesy of Dr. Zeke Hausfeather of BEST.

Discussions of Model-Observation Agreement

Although temperatures had been running lower than the central estimate of IPCC projections in recent years, they were, and are, still within the projected ‘envelope,’ as shown in the figure above and discussed at length in the linked articles.

Moreover, I added, there was and is a long track record in the scientific literature of successful predictions by climate models. It was collected and documented by Barton Paul Levenson (also linked in sidebar.)

I quoted Barton as follows below:

Global Climate Models have successfully predicted:

  • That the globe would warm, and about how fast, and about how much.
  • That the troposphere would warm and the stratosphere would cool.
  • That nighttime temperatures would increase more than daytime temperatures.
  • That winter temperatures would increase more than summer temperatures.
  • Polar amplification (greater temperature increase as you move toward the poles).
  • That the Arctic would warm faster than the Antarctic.
  • The magnitude (0.3 K) and duration (two years) of the cooling from the Mt. Pinatubo eruption.
  • They made a retrodiction for Last Glacial Maximum sea surface temperatures which was inconsistent with the paleo evidence, and better paleo evidence showed the models were right.
  • They predicted a trend significantly different and differently signed from UAH satellite temperatures, and then a bug was found in the satellite data.
  • The amount of water vapor feedback due to ENSO.
  • The response of southern ocean winds to the ozone hole.
  • The expansion of the Hadley cells.
  • The poleward movement of storm tracks.
  • The rising of the tropopause and the effective radiating altitude.
  • The clear sky super greenhouse effect from increased water vapor in the tropics.
  • The near constancy of relative humidity on global average.
  • That coastal upwelling of ocean water would increase.

Seventeen correct predictions? Looks like a pretty good track record to me.

Jack's response to that was indirect:

Scroll to Continue

Doc, I came across this web site recently and would like you to comment -

(Jack's site is linked below.)

I will make a pledge to you.

You ask me what it would take to be convinced.

If the items in the forecast for 2015 and 2020 comes true as they projected, I will be convinced.

There were problems with that. The worst for me is that there is simply no point in convincing Jack (or anyone else, for that matter) sometime in 2020 or 2021. We need decisive action on climate, and we need it now.

But there are other issues, too. Some of the ‘predictions’ involve things that are really not all that relevant—global air conditioner sales, for instance. And what would the criteria for predictive success be? Surely it would be unrealistic to expect each and every point to come true precisely? For that matter, some of the projections are not couched very precisely. How could we decide whether or not they should be considered ‘successful’?

Noting all these problems—and, frankly, hoping to split up what looked like a daunting workload—I made a suggestion to Jack:

So, how about this: you and I make a project. We'll sort the predictions for this year (ie., predictions on the 2015 page of the site) that we want to assess--other than what I've done here, no looking ahead! (Full disclosure: I already looked at the case of Lagos, Nigeria, a bit.) Then we'll research them and compare what we find. We each write an article about it.

What do you say?

Jack accepted, and so the present article was born.

Read Jackclee's Article Here!

Sorting The "Predictions"

My first task was to read and sort the predictions on the 2015 ‘predictions’ page. (As of 2018 this site is unfortunately dead, so accessing the original is no longer possible.)

A tedious process of listing, winnowing, consolidation and tabulation eventually produced a more-or-less manageable list of 28 items. Fourteen of them were then eliminated ‘for cause.’ These items (with their original list positions and ‘cause for dismissal’) are listed in Table 1:

Table 1: Rejected 'Predictions'

Table 1: Rejected 'Predictions'

Item DescriptionOriginal list #Reason for rejection

Global air conditioner sales increase


Silly proxy. Sales can be affected by too many things besides climate. (But the projection did apparently come true, FWIW.)

Global emissions projections


Not actually a prediction, and a driver of climate, not a consequence of it.

Lake Mead dry by 2014, 10% chance


Too low a chance to count as a ‘prediction.’

Suna’a, Yemen, to run dry by 2017


Water situation serious, but civil war renders clear outcome relative to prediction impossible.

Various population projections


Not climate predictions, though growing populations do tend to use more energy.

Climate-driven migration in Nigeria


Civil conflict and weak governance make this impossible to assess.

Loss of climate measurement/observation capability


Not a climate prediction, though it makes climate study harder (and has occurred).

Rare earths shortages by 2015


Not a climate prediction.

Worldwide oil supply shortage of 10M barrels/day by 2015


Obviously a bad miss, but still not a climate prediction.

No ‘demand challenge’ to global energy supply in 2015


One more time—not a climate prediction.

Global energy prices to be unstable during 2000-2015.


Certainly, but no, not a climate prediction.

Solar energy predicted to be the least expensive source of electricity by 2016.


Not a climate prediction, but will comment briefly as ‘appendix.’

China to mine 25% more coal; consumption to reach 2.3 billion tons by 2015.


Not a climate prediction, but will comment briefly as ‘appendix.’

US agriculture suffers due to lack of pollinators, leading China to supply up to 40% of US vegetables


CCD—the epidemic of bee deaths—is still quite a problem, but hasn’t undermined US ag quite that badly. And the Chinese economy has grown in ways not well anticipated in 2006.

The Rubric

That leaves 14 predictions to assess. But how to assess them? Not all were precisely quantified, and even when they were, available data aren’t necessarily sorted in such a way that direct comparisons can be made.

I fell back on classroom teaching experience to create a rubric to enable ‘grading’ of each prediction. Here’s what it looked like:

Predictions rubric


  • 4—Prediction within 10%
  • 3—Prediction within 25%
  • 2—Prediction within 50%
  • 1—Correct sign
  • 0—Wrong sign


  • 4—Outcome closely resembles prediction
  • 3—Outcome reasonably resembles prediction
  • 2—Outcome somewhat resembles prediction
  • 1—Outcome points toward possibility of prediction being realized, given enough time
  • 0—No resemblance between outcome and prediction

(1 additional point may be awarded in cases where outcome exceeds prediction--that is, where climate change is worse than predicted.)

With that in hand, I attacked the list of remaining predictions. Here are the results, item by item, and with a discussion of what I see as important points relating to each.

Assessing The Predictions

In all cases, the supporting web links for the prediction and outcomes will be found following the prediction and preceding the section describing the observed outcome.

Prediction #1:

Stanford computer models project a dramatic spike in extreme seasonal temperatures during the period 2010 - 2019. "The Stanford team also forecast a dramatic spike in extreme seasonal temperatures during the current decade [2010 – 2019]. Temperatures equaling the hottest season on record from 1951 to 1999 could occur four times between now [2010] and 2019 over much of the U.S., according to the researchers. The 2020s and 2030s could be even hotter, particularly in the American West."

The reality:

The US experienced significant heat waves in 2011 (“On a national basis, the heat wave was the hottest in 75 years”), 2012 (March brought “a remarkably prolonged period of record setting temperatures”), 2013 (regionally, in the Southwest “46 monthly record high temperatures were reached or broken, and 21 records for the highest overnight temperatures were reached or broken”), and 2015 (“triple-digit heat indices across a large swath of the U.S...”)

Interestingly, consideration of one obscure but telling statistic—the tally of ‘cooling degree days’—the top three hottest US summers occurred during the prediction period so far. In order, they are: 2011, 2010 and 2012.

Given that the prediction period ran from 2010 through 2019, and is thus only about half over, it is tempting to rate this prediction as a ‘5’—that is, the number of observed events matches the predicted number of events, for a ‘4’ on the quantitative rubric, plus a bonus point since there are still several years to run in the prediction period.

However, considering that there are serious definitional issues about just how geographically widespread and how long-lasting a heatwave needs to be to count, and considering my own biases, I reduced that to a ‘3’—“outcome reasonably resembles prediction.”

March 2012 heatwave.  Image courtesy NASA Earth Observatory & Wikimedia Commons.

March 2012 heatwave. Image courtesy NASA Earth Observatory & Wikimedia Commons.

Prediction #2:

Britain’s Met Office projects 2014 temperature likely to be 0.3 degrees Celsius warmer than 2004. “Here is the climate forecast for the next decade [2007-2014]; although global warming will be held in check for a few years, it will come roaring back to send the mercury rising before 2014."

The reality:

Once again, definitional issues cloud the picture a bit. Using the data set associated with Britain’s Meteorology Office, HADCRUT 4, one finds that 2014 temperatures were not 0.3 C warmer than 2004, but rather 0.117 C. (NASA’s data would have made that figure 0.20 C.) Clearly, less warming than forecast. On the other hand, the shape of the temperature curve does match the description given: “ warming will be held in check for a few years [but will] come roaring back.”

Overall, I rate that as a ‘2’—“outcome somewhat matches prediction.”

It’s worth noting, though, that this is more a test of ‘the Met’s’ experimental long-term forecasting ability than of climate modeling; though the 10-year is very long for weather, it is very short for climate. According to Santer et al., the shortest period for which one might expect to see a statistically-significant warming trend is 17 years.

Prediction #3:

By 2015 10 million acres of national forests may be at high risk of uncontrollable, catastrophic wildfires... as much as $12 billion, or about $725 million a year, may be needed to treat the 39 million acres at high risk of uncontrollable wildfire by the end of fiscal year 2015.

The reality:

By September 25 of this year, over 9 million acres had in fact burned. By the end of October (the conventional end of the ‘fire season’, the number had reached 9,407,571 acres. Clearly that is well within the 10% envelope for a ‘4’. There aren’t yet comprehensive numbers on the cost of those fires, but on August 5, a Forest Service Report informed us that “For the first time in its 110-year history, the Forest Service, part of the U.S. Department of Agriculture, is spending more than 50 percent of its budget to suppress the nation's wildfires.” That was not a result of one exceptional year, but rather a consistent trend in fire-fighting costs. The Service called for a change in the funding mechanism to reflect this reality, as ever-increasing proportions of the Service budget were being absorbed by fire-fighting costs, to the detriment of other functions. (The full report link is in the sidebar.)

Rating: 4 “Prediction within 10%.”

Washington State wildfires, 2015.  Image courtesy NASA & Wikimedia Commons.

Washington State wildfires, 2015. Image courtesy NASA & Wikimedia Commons.

Prediction #4:

Water shortages projected for 36 states by 2013. Water managers in most states expect shortages of freshwater in the next decade [2003 – 2013]

The reality:

Unclear. Although the General Accounting Office publishes periodic surveys of state water manager expectations, they do not examine the actual outcomes. And among the responses of the water managers are survey responses that raise real questions about response quality. Apart from answers that were unresponsive—in 2014, the most recent such survey, Indiana and Ohio were both listed as ‘no response or uncertain’—there were instances that were simply not credible.

A notable example is provided by the states of Alabama and Georgia, which both report no concerns about freshwater availability despite the fact that they are, along with Florida, embroiled in a legal and political wrangle over the apportionment of freshwater flowing out of Lake Lanier, the sole source of most of Atlanta’s drinking water. The ‘tri-state water war’ has been before courts since 1990, and was the subject of a closed-doors meeting of all three governors as recently as June 2015.

In my view, if that is not cause for ‘concern’, then something is wrong with the definition of ‘concern’ in use. (To be completely clear, though, water problems in the Southeast are not a climate change issue—regional modeling does not project drought problems to be likely, as overall the region seems likely to become slightly wetter—but a policy and resource versus population issue.)

However, despite such concerns, the 2014 report has the number of ‘concerned’ water managers up by 4 to 40. And in the general media there were very serious water shortages reported for 7 states in 2015. (Of course, the current serious water shortages in California are too well-known to require a citation.)

Considering the information available, the outcome seemed ‘somewhat’ to resemble the prediction, for a rating of ‘2’.

Prediction #5:

Lake Mead’s water levels could drop below its water intake pipes by 2013. "Southern Nevada Water Authority chief Pat Mulroy . . . said the authority is in a race against time to complete a new [third intake] system [or third straw] to draw water from deep in Lake Mead [Hoover Dam]."

The reality:

The Water Authority won their race, but not by much. The ‘third straw’ project is now complete, at an announced cost of $817 million, with another $650 million for a new pumping station. The level didn’t quite reach crisis levels: problems start at a level of 1062 feet, and the system as it was would have been shut down at 1050. This summer saw levels of a little over 1075. That margin of less than 14 feet may not seem small to some, but for context, consider the ‘old normal’: in 1983 Lake Mead stood at 1225 feet.

The outcome reasonably resembles the prediction, for a rating of ‘3’.

Hoover Dam, 2012, with the 'bathtub ring' showing low water level.  Image by Tony Webster, courtesy Wikimedia Commons.

Hoover Dam, 2012, with the 'bathtub ring' showing low water level. Image by Tony Webster, courtesy Wikimedia Commons.

Prediction #6 (related to #5):

Hydroelectric turbines at Hoover Dam could cease generating electricity by 2013. “After 75 years of steadily cranking out electricity for California, Arizona and Nevada, the mighty turbines of the Hoover Dam could cease turning as soon as 2013, if water levels in the lake that feeds the dam don't start to recover, say water and dam experts. Under pressure from the region's growing population and years of drought, Lake Mead was down to 1,087 feet, a 54-year low, as of Wednesday [September 8, 2010]. If the lake loses 10 feet a year, as it has recently, it will soon reach 1,050 feet, the level below which the turbines can no longer run.”

The reality:

Fortunately, the loss rate since 2010 did not continue uniformly, and although there is a small net loss, the turbines still turn—albeit with a 25% power loss. It’s worth noting, though, that hydropower in California is seriously affected by the ongoing drought and water shortage, with reductions of around 60%. As a linked story puts it:

California’s drought is just four years old. But the drop in the state’s hydroelectric production has been precipitous. Hydroelectric sources are projected to contribute just 7 percent of the state’s power this year, down from 23 percent in 2011.

Overall, the outcome was judged as pointing toward a later possibility of realizing prediction, for a rating of ‘1’.

Prediction #7:

Nearly half the world’s population will live in water-stressed countries by 2015. “By 2015 nearly half the world's population — more than 3 billion people — will live in countries that are "water-stressed" — have less than 1,700 cubic meters of water per capita per year, mostly in Africa, the Middle East, South Asia, and northern China."

Note the wording: this does not say that half the world’s total population will be water-stressed; it says that countries accounting for half the world’s population will experience significant water stress.

The reality:

This appears to be a solid ‘hit.’ Though definitive numbers for 2015 are not available, India and China are indeed both experiencing water stress at very significant levels, as has been the case for some time, and together account for close to 50% of global population. The story in the UK's Guardian newspaper, linked, tells the wider tale.

Rating: ‘4’.

Ladakh, India, 2014.  Image by Christopher Michel, courtesy of Wikimedia Commons.

Ladakh, India, 2014. Image by Christopher Michel, courtesy of Wikimedia Commons.

Prediction #8 (related to #7):

By 2015 a number of developing countries will be unable to maintain their levels of irrigated agriculture. “In the developing world, 80 percent of water usage goes into agriculture, a proportion that is not sustainable; and in 2015 a number of developing countries will be unable to maintain their levels of irrigated agriculture...”

(This comes from the same 2000 report as item #7, and is not linked again.)

The reality:

The situation for irrigation is bad and getting worse in both India and China. Additionally, Africa has serious problems, though these arise from a whole network of reasons, from climate change to population growth to poor policy and migration.

Rating: ‘4’.

Prediction #9:

Mt. Kilimanjaro’s remaining ice fields likely to disappear between 2015 and 2020... if current climatological conditions persist, the remaining ice fields are likely to disappear between 2015 and 2020.

The reality:

Thankfully, ‘the snows of Kilimanjaro’ currently seem unlikely to disappear any time soon. This prediction would receive a clear zero, except for one thing: attention to the issue, prompted by the possibility that the prediction could come true, may have been crucial.

Initially, the observed loss of ice mass on Kilimanjaro’s summit was ascribed more or less directly to global warming. But further analysis showed that the loss was probably due to less precipitation falling at the summit, and that in turn this was not so much due to global changes, but to more local ones: deforestation on Kilimanjaro’s massive slopes had altered the local water cycle. Replanting those slopes seems to have helped increase precipitation, slowing (though not halting) ice loss:

...the massive tree planting around the mount Kilimanjaro could have been mitigated the ripple effects of the global warming.

Alarmed by the...Thompson study, way back in 2006, Tanzania President Jakaya Kikwete imposed a total ban on tree harvesting in Kilimanjaro region in a move aimed to halt catastrophic environmental degradation, including melting of ice on Mount Kilimanjaro.

As a result of the measures, the forest cover on the mount Kilimanjaro is slowly, but surely becoming thick.

Experts say the forests on Kilimanjaro's lower slopes absorb moisture from the cloud hovering near the peak, and in turn nourish flora and fauna below...

Given that ice loss has not been completely arrested and that warming continues, the outcome points toward a possibility that the prediction may become true in time, which rates a ‘1’.

Aerial view of Kilimanjaro, showing ice fields on Kibo peak.  Image by clem23, courtesy of Wikimedia Commons.

Aerial view of Kilimanjaro, showing ice fields on Kibo peak. Image by clem23, courtesy of Wikimedia Commons.

Prediction #10:

Computer model forecasts taking into account sea ice thinning and albedo effects project an ice-free summer Arctic Ocean between 2010-2015. “The Arctic Ocean could be free of ice in the summer as soon as 2010 or 2015 -- something that hasn't happened for more than a million years, according to a leading polar researcher. Louis Fortier, scientific director of ArcticNet, a Canadian research network, said the sea ice is melting faster than predicted by models created by international teams of scientists, such as the Intergovernmental Panel on Climate Change. They had forecast the Arctic Ocean could be free of summer ice as early as 2050.”

But this 'prediction' needs more context. Note what is said in this story—(so-called) ‘IPCC models’ at that point (November, 2007) had been estimating that the Arctic sea ice would likely be gone at the annual minimum in September ‘as early as 2050’, but a new regional model by Dr. Maslowski, of the US Naval Observatory, had projected that a much earlier outcome was possible.

Note, too, that the 2007 minimum thoroughly shocked experts; they had been concerned by the record low of 5.6 million square kilometers (mean value for the month of September). Prior to 1990, only once had that value dropped below 7 million square kilometers, and never had it broken through 6 million square kilometers. But in 2007, the disturbing record clocked in 2005 was obliterated by a stunning 4.3 million square kilometer mean September extent--a full 1.3 million kilometers less than the 2005 record (roughly 23% lower). Dr. Fortier’s comment that ‘'s probably going to happen even faster than that” should be read in the context of the shock the 2007 minimum provided.

Graph courtesy Dr. Larry Hamilton.

Graph courtesy Dr. Larry Hamilton.

It should also be noted that the newspaper story is almost certainly wrong in one respect. Though the identity of the ‘computer models’ referred to is never given in the story, it is undoubtedly the regional modeling of Dr. Wieslaw Maslowski, of the US Naval Observatory, as reported in the BBC story linked above.

Dr. Maslowski is directly quoted in another story from the same time:

Given the estimated trend and the volume estimate for October–November of 2007 at less than 9,000 km3, one can project that at this rate it would take only 9 more years or until 2016 ± 3 years to reach a nearly ice-free Arctic Ocean in summer. Regardless of high uncertainty associated with such an estimate, it does provide a lower bound of the time range for projections of seasonal sea ice cover.

So the projection, according to the scientist who made it, should be regarded as a “lower bound”, and the time frame is not 2010-2015, as the story had it, but 2013-2019. All that clarified, we're ready to look at the outcome.

The reality:

Dr. Fortier was wrong.

But consider the continuing decline of the sea ice—after 2007, the September mean has never again risen above 2005 levels. And in 2012 the September mean extent crashed to just 3.6 million square kilometers. (September of this month saw the fourth-lowest value in the record, with a mean of just 4.6 million.)

In that context, it is not so clear that Dr. Maslowski was wrong. The window for his ‘lower bound’ estimate runs until 2019.

The IPCC was wrong, too, or so it appears at this juncture. In 2007, they thought that we had until 2050 or so before the first ice-free Arctic summer. The sea ice crash we have seen since then makes that scenario highly unlikely; currently observers such as the National Snow and Ice Center’s Dr. Walt Serreze now think the likely year is sometime around 2030.

Dr. Fortier gets a ‘1’, even though the mainstream science would do better.

Eko-Atlantic City under construction, Lagos, Nigeria, 2011.  Image by omar 180, courtesy Wikimedia Commons.

Eko-Atlantic City under construction, Lagos, Nigeria, 2011. Image by omar 180, courtesy Wikimedia Commons.

Prediction #11:

Lagos, Nigeria projected to be at risk from sea level rise. "Nigeria will suffer from climate-induced drought, desertification, and sea level rise... Lagos, the capital, is one of the West African coastal megacities [along with Alexandria, Egypt] that the IPCC identifies as at risk from sea level rise by 2015.”

The reality:

Again, solid, comparable information is hard to come by, and the prediction itself is not very specific. But it is clear that Lagos is facing increased flooding, forming a serious threat to its infrastructure:

An increasingly important threat to the high population and large concentration of residential, industrial, commercial and urban infrastructure systems in Africa’s coastal megacity of Lagos is flooding. Over the past decade, flooding in Lagos has increased significantly, drawing increasing attention to the need for flood risk management.

It’s not as clear what proportion of this risk proceeds from sea level rise, as identified in the prediction, and what proportion from extreme precipitation and increasing storm surge (both expected consequences of climate change, in general) or from other causes, such as land subsidence (which can be either natural or man-made, and which results in localized ‘relative sea level rise.’)

However, it is noteworthy that the there’s a mega-project, underway since 2003 and now said to be nearing completion, intended to protect the city from sea-level rise—an 8-kilometer barrier dubbed the ‘Great Wall of Lagos.’ Not only that, an artificial island will be the site of a glittering new city center, financed entirely by private investment, and intended to become the “Hong Kong of Africa”. As usual, that is linked above, together with another, less enthusiastic take on the project. Not yet reality, but perhaps worth noting in passing, is that serious, widespread issues with both desertification and sea level rise continue to be projected for Africa.

Overall rating: ‘3’.

Adelie penguins.  Image by Jerzy Strzelecki, courtesy Wikimedia Commons.

Adelie penguins. Image by Jerzy Strzelecki, courtesy Wikimedia Commons.

Prediction #12:

Projected extinction of Adélie penguin population around Palmer Station, Antarctica. “A small residual population [Adélie penguins] on Humble Island [near Palmer Station, Antarctica] may survive the climatic shift down the peninsula, [seabird ecologist Bill Fraser of the Palmer Long Term Ecological Research (PAL LTER) project] guessed, but the overall prognosis is that in the next decade the Adélies around Palmer will be gone. ‘Their numbers are in catastrophic decline,’ Fraser said.”

(Unfortunately, the original link appears to be dead, and so is not linked.)

The reality:

The Adelies are not gone yet, though the decline in population continues. Chinstrap and Gentoo penguins, which tolerate warmer temperature better, have been moving in in large numbers.’s been a shock to see how drastically Adelie penguins have declined while gentoos have increased in the last 20 years.

I have linked some of the baseline research, from 1998, and including as one author Dr. Bill Fraser, who was mentioned in the prediction.

It would be great to see some hard numbers on the Adelie population of Palmer Station, to get a better feeling for how the trends are playing out. But it appears that the prediction is somewhere in the midrange: 2018 is probably too aggressive, but all sources discussing the population agree that the species is in trouble in the Palmer Station area.

Rating: ‘3’, “Outcome reasonably resembles prediction.”

Update on Adelie penguins, 7/2/16

A recent article on the prognosis for Adelie penguins, not just around Palmer Station, but around the whole Antarctic continent, stated that, as of 2013, the Palmer Station population had been reduced by about 80%.

The outlook for the species generally is not great:

by 2099, our projections suggest 78% to 51% (mean 58%) of colonies could experience declines, containing 64% to 39% (mean 46%) of the current abundance.

Luckily, while Adelies look to be vulnerable to decline, there are areas projected to serve as 'refugia', so complete extinction doesn't appear to be a risk--over the course of this century, at least. Of course, under any 'business as usual' scenario, warming will not stop magically when the 22nd century arrives.

Update on Adelie penguins, 4/21/19

An NPR interview with marine biologist James McClintock provided an update on the status of the Adelie penguins--and not an encouraging one.

The news is a little sad. The population of 15,000 breeding pairs of Adelie penguins has reached a new low. It's down to 1,100 this year. So over 90 percent of them are disappearing. What's happening mainly that's causing the big problem is that they come in at a very predetermined time of year to lay their eggs. And then along comes these unseasonably late snowstorms because it's getting warmer and more humid. Ironically, it's snowing later. And then the snow melts, and the eggs drown. So the Adelie's having a really tough time right now.

Prediction #13:

Antarctic ozone hole will continue to expand through 2015. “Some existing agreements, even when implemented, will not be able by 2015 to reverse the targeted environmental damage they were designed to address. The Montreal Protocol is on track to restore the stratospheric ozone layer over the next 50 years. Nevertheless, the seasonal Antarctic ozone hole will expand for the next two decades [2000-2020] — increasing the risk of skin cancer in countries like Australia, Argentina, and Chile—because of the long lag time between emission reductions and atmospheric effects.”

(The source for this prediction is the same as #11, above, and is not re-linked.)

This is not really a climate prediction, either, but I consider it nevertheless because it bears in several ways on the current topic, aside from the fact that it was included on the website. Essentially, it’s an important environmental issue involving science, global policy, and numerical modeling of atmospheric processes, and one in which we can observe the outcome of an international treaty intended to mitigate human-induced damage to the atmosphere.

The reality:

Essentially, ozone loss has gradually stabilized since implementation of the Montreal Protocol. The lowest 30-day extent occurred in 2006, but this year saw the single largest one-day ozone hole on the record. Despite that, some thickening of the ozone layer has been observed, and scientific observers believe that recovery of the layer may have begun.

Rating: ‘4’.

Maps courtesy of DLR (German Aerospace Center.)

Maps courtesy of DLR (German Aerospace Center.)

14. The prediction:

By 2015 the cost in lost income of degraded coral reefs is projected to reach several hundred million dollars annually.

The reality:

Again, one might wish for better numbers. But the worldwide decline of coral reefs is so serious as to merit an entire chapter in Elizabeth Kolbert’s Pulitzer-winning book, The Sixth Extinction, and NOAA officials warned last month of a third-ever global-scale coral-bleaching event:

This bleaching event, which began in the north Pacific in summer 2014 and expanded to the south Pacific and Indian oceans in 2015, is hitting U.S. coral reefs disproportionately hard. NOAA estimates that by the end of 2015, almost 95 percent of U.S. coral reefs will have been exposed to ocean conditions that can cause corals to bleach.

The biggest risk right now is to the Hawaiian Islands, where bleaching is intensifying and is expected to continue for at least another month. Areas at risk in the Caribbean in coming weeks include Haiti, the Dominican Republic and Puerto Rico, and from the U.S. Virgin Islands south into the Leeward and Windward islands.

The next concern is the further impact of the strong El Niño, which climate models indicates will cause bleaching in the Indian and southeastern Pacific Oceans after the new year. This may cause bleaching to spread globally again in 2016.

Earlier, reported estimates put the annual value at risk at about $30 billion, and stated that the Caribbean might have lost 80% of its coral. This estimate dates from 2002, however, and so does not ‘confirm’ the scale of contemporary losses.

Considering the available information, while the value at risk remains uncertain, the estimates of total value imply that the prediction’s losses would amount to a few per cent of the total value. Given that loss rates are very high, it would seem to follow that the outcomes we see are ‘closely resembling the prediction,’ which would merit a rating of ‘4’.

Update, 3/19/17

The "third coral bleaching event" mentioned did indeed continue in 2016, and indeed intensified as the world saw a record-warm year on the strength of the ongoing anthropogenic warming trend in combination with an El Nino nearly as strong as that of 1997-98. Unfortunately, it appears to be continuing in 2017, as global temperatures have remained quite warm even after the El Nino ended.

The world has lost roughly half its coral reefs in the last 30 years. Scientists are now scrambling to ensure that at least a fraction of these unique ecosystems survives beyond the next three decades. The health of the planet depends on it: Coral reefs support a quarter of all marine species, as well as half a billion people around the world.

"This isn't something that's going to happen 100 years from now. We're losing them right now," said marine biologist Julia Baum of Canada's University of Victoria. "We're losing them really quickly, much more quickly than I think any of us ever could have imagined."

Even if the world could halt global warming now, scientists still expect that more than 90 per cent of corals will die by 2050. Without drastic intervention, we risk losing them all.

Israeli corals near Eilat, at the head of the Gulf of Aqaba.  Image by Ludwig14, courtesy Wikimedia Commons.

Israeli corals near Eilat, at the head of the Gulf of Aqaba. Image by Ludwig14, courtesy Wikimedia Commons.

Table of prediction ratings

Item DescriptionPointsComments

US heat weaves, 2010-2019


Rating reduced due to definitional questions.

UK 2014 temperature


Underpredicted; technically weather, not climate

US Wildfire



US water shortages


Poor information on outcomes

Lake Mead water levels


$1.4 billion spent on remediation

Hoover Dam hydro generation



Global water stress


Definitive numbers not yet available

Agricultural irrigation at risk



Snows of Kilimanjaro


Human response to trend altered outcome

Louis Fortier over predicts ice-free Arctic


Mainstream climate science would do much better than Fortier

Lagos at risk for sea level rise


Large expenditures on mitigation of risk

Palmer Station Adelies extinct



Antarctic ozone hole extent


Model predictions appear to be accurate

Coral bleaching costs


Economic costs hard to document, but extent of coral loss is clear

Total points



How do you interpret those numbers? In school, that would likely be a D, or perhaps a C-; a pass, to be sure, but nothing to brag about.

But those numbers aren't grades. Consider that:

  • The most frequent rating was 4, the highest possible;
  • The least frequent rating was 0, which was never awarded;
  • The highest rating was given the same number of times (5) as the two lowest ratings combined.

Jack picked his predictions on this basis:

Doc, you missed my point about the far reaching projections of this site. The point is they are meant to scare and not based on anything real.

I think this exercise shows that however they may have been meant, they are indeed based on reality.

RatingTimes awardedPoints resulting



















Jack's "Big Three" Predictions--Temperature

I'm going to take the privilege of the tardy--for Jack published his Hub roughly six weeks before I wrote these words--and comment briefly on the three predictions that he addresses there.

1. Temperature increase. In part, I've already addressed this issue above when I cited the various model-observation comparisons that have been made. But let's get to the nitty-gritty.

Jack quotes Jim Hansen (not Michael Mann and Jim Hansen; in 1988, the former was still a humble physics undergrad at Yale):

If the current pace of the buildup of these gases continues, the effect is likely to be a warming of 3 to 9 degrees Fahrenheit from the year 2025 to 2050, according to these projections. This rise in temperature is not expected to be uniform around the globe but to be greater in the higher latitudes, reaching as much as 20 degrees, and lower at the Equator.

How does this stack up against reality? Jack doesn't really examine that, citing only a Daily Caller report on a single study which concluded that observed warming so far did not exceed natural variability over the last 8,000 years. But that says nothing about the prediction that was made.

But it's not a difficult question to answer: since 1988, the GISTEMP temperature record shows a total warming of 0.45 degrees Celsius, according to a standard 'least-squares' regression, or about 0.16 degrees C per decade. If we presume that warming continues at that same rate until 2025, then we would see 0.56 C; for 2050, that would be 0.96 C. What's that in Fahrenheit? Well, rounding up to 1 C for simplicity, that would be 1.67 degrees F, or a little more than half the 1988 estimate.

But before we conclude that climate science and global warming are nothing but bunk, perhaps we should look at what more recent science has to say? After all, 1988 was a long time ago in terms of scientific progress. And what more recent work has to say, too, is not difficult to answer, for the International Panel on Climate Change (IPCC) put forward estimates of future warming in each of its Assessment Reports to date.

The first came in 1990, and in line with the 1988 quote estimated a warming rate of 0.25 C per decade. Multiplying by the six decades to 2050, one reaches 1.5 C total warming, or 2.7 F, reasonably close to Hansen's lower bound. But by the Second AR, in 1995, new work had reduced that number to just 0.14 degrees per decade. Since that is 2.7 decades, we'd expect to see roughly 0.38 C--quite close to the 0.45 we've actually observed. By the time of TAR (2000) the warming rate had crept up to 0.16 C, and by AR4 in 2007, the best estimate had become 0.18 C. Clearly, work that is no more than 2 decades old is pretty close to reality in estimating global mean surface temperature.

GISTEMP record with trend.  Graph by author, using online tool.

GISTEMP record with trend. Graph by author, using online tool.

2015 Update

2015, powered by a strong El Nino, continues to set temperature records. With October data now in for several of the major datasets, the year is almost certain to set new records for warmest on record. As the Washington Post reports:

Earlier this month, Britain’s weather service, the Met Office, and NASA both stated that the Earth’s average temperature is likely to rise 1 degree Celsius above pre-industrial levels for the first time by the end of this year. This milestone is significant since it marks the halfway point to two degrees Celsius, the internationally accepted limit for avoiding the worst consequences of climate change.

More specifically, GISTEMP reported an October anomaly value of 1.04 C, the warmest monthly anomaly in their record, and the first to exceed a degree Celsius. (HADCRUT4 has not yet released their October value; September clocked in at 1.03 C, though their baseline is slightly different from GISTEMP and thus the numbers are not directly comparable.) The Japanese Met Office also reported a record-warm October. Even the UAH satellite record reported their warmest October ever, as reported by the 'skeptic' website Watt's Up With That.

Though there was never any statistical evidence that a true 'pause' in warming was taking place, the slowdown in warming rate seen through much of the 00s has definitively come to an end.

2018 Update

With the hin