Climate Models Accurately Predict Warming
Climate models employ piles of data and sophisticated computational techniques to predict what will happen in the future. Sometimes they predict what happened in the past as well. That is important to test the models (because we might know what happened in the past), or to fill in the blanks (we don’t always know exactly what happened in the past) or to understand complex climate systems better.
If you glance at the science denier rhetoric (mainly on blogs, you won’t find much in the peer reviewed literature because it isn’t good science) you’ll see repeated claims that climate models that try to predict global warming don’t match the actual observations of global warming. Most of the time, this claim is simply wrong. Perhaps an improper measurement of warming (like temperatures up higher in the atmosphere where we actually expect cooling rather than warming) is being used, which constitutes a lie. In other cases observed warming is within the model projections, but tracks off to one side (usually the low side) of the average expectation, but remains within the margin of error. This is either a misunderstanding of how the science works, or a willful misrepresentation. (Again, a lie.) But there are actually two legitimate areas where certain climate models seem to overstate observed warming. A recent post by John Abraham at the Guardian explores these areas.
First there is the question of where the warming is observed. We measure warming in several parts of the Earth’s surface. (See “What does “Global Warming” Mean?) One is surface temperatures at about head height, over land, via a gazillion weather stations many of which have been in operation since the 19th century. The other is at the surface of the sea, using a combination of older measurements taken from ships and more recent satellite observations. In addition, we have measurements of the deeper ocean itself, usually averaged out over some depth such as the top 700 meters, or the top 2,000 meters. This combines older and new ship our buoy measurements but tend to not go back in far in time as the land and sea surface measurements.
John Abraham has spent a lot of effort looking at ocean temperatures at depth. He and I recently published this item, and he’s done a lot of additional work on it. The total amount of heat added to the Earth’s surface from anthropogenic warming (caused mainly by greenhouse gasses such as carbon dioxide) is divided between the ocean and the surface, with the ocean taking up much of that heat. I liken the system to a dog with a wagging tail. The ocean is the dog, the surface temperatures, making up a small part of the overall system and being much more variable, is the tail.
In “The oceans are warming faster than climate models predicted,” Abraham notes:
We separated the world’s oceans into the Atlantic, Pacific, and Indian. All three of these oceans are warming with the Atlantic warming the most. We also calculated the ocean heating by using 40 state-of-the-art climate models. Over the period from 1970, the climate models have under-predicted the warming by 15%.
And here you can see a number of climate models superimposed over the observed heating in the top 700 meters of the ocean (the red line):
In a more recent post, Abraham asks the question, “ how have the models done at predicting the changes in air temperatures?”
As noted, global surface temperature is estimated in part from a bunch of thermometers around the globe, but these thermometers were not placed there for this purpose. They are weather stations set up to help track the weather, not to address questions of climate change. They are not evenly distributed, and there are huge gaps in the surface coverage, most notably in the Arctic and interior Africa, both regions where recent warming has probably been greater than elsewhere. In order for these temperature data to be used, they have to be carefully employed in a computational system that helps fill in the gaps. There are other complexities beyond our scope here. Suffice it to say that when a bunch of different groups of scientists (i.e., the British meteorology office, NASA, NOAA, the Japan Meteorological agency, and various university based groups) take on the task of estimating surface temperatures, they all do it a bit differently and thus turn up with slightly different curves.
This is true as well with sea surface temperatures. There is more than one way to measure sea surface temperature, or more exactly, to take existing data and turn it into a useful estimate of sea surface temperatures.
In addition, the data are cleaned up over time as mistakes are found, the basic computational approach used may be updated to make it more precise, and the overall approach to computation may be changed to make it more accurate.
Over time, two clear patterns have emerged. First, if you take all of the different measurements of surface temperature over time spanning from the 19th century to the present, lay them all out in front of you and step back about two meters and squint slightly, you can’t see the difference among them. They all look the same, they all tell the same story. They all have a handful of notable ups and downs along the generally upward march of surface temperatures with industrial pollution. You have to look at the graphs all on the same axes, together, to see the differences, and the differences are minor. This tells us that all the different approaches to processing a largely overlapping set of data end up with the same basic result. So many smart minds working with the best available science all produce the same result. How boring. But also, how reassuring that the science is being done right.
The second pattern emerges when we look at these graphs as they are produced over time. Various groups have said, “hey wait a minute, we’re missing this” or “we’re missing that” factor. Urban heat island effects may change the data! What about the Arctic! Interior Africa! Our satellites were recalibrated what does that do? Etc.
Over time, and honest, well informed, diligent effort by many groups to improve the measurement of the Earth’s surface temperature has resulted in a series of slightly different graphs, and in each and every case of which I’m aware, the resulting, more recent and better done graphs show more warming, and various periods of relative flatness have become steeper (going upwards).
So, what John Abraham has done is to take some of the more recently processed, better quality data and compared it to the usual models to see how well the models have done. They did well.
He based his discussion on a comparison of the most recent climate model simulations with actual global surface temperature measurements as numerically summarized by NASA’s Gavin Schmidt, shown here:
John has superimposed 2015 so far (the star).
This shows the most current computer model results and five current temperature data sets. The dark line is the average of the models, and the various colored lines are the temperature measurements.
The dashed line is slightly above the colored markers in recent years, but the results are quite close. Furthermore, this year’s temperature to date is running hotter than 2014. To date, 2015 is almost exactly at the predicted mean value from the models. Importantly, the measured temperatures are well within the spread of the model projections.
This is the year of the heatwave. We’ve had heat waves off and on for the last few years, but it seems that there are more now than ever before. While some have tried to argue that global warming can’t really cause warming (sometimes expressed as heat waves), it does.
Climate Scientist Stefan Rahmstorf has a blog post, in German, about a current heat wave in Europe. He notes (rough translation):
Europe is currently experiencing the second major heat wave this summer. On 5 July, according to the German Weather Service , a never before measured in Germany temperature reached 40.3 ° C in Kitzingen. But a month later, on August 7, this century record has been revised…One might speculate that a single heat wave could be simply due to chance. If you look at the temperature data, however, in their entirety it is immediately clear that extreme heat has become more common over several decades. (Apologies for errors in translation.)
And he shows this graph:
Percentage of global land area where the temperatures over a month were two or three standard deviations above the average from 1950 to 1980. Two standard deviations (orange) could be described as “very hot”, three standard deviations (dark red) as “extremely hot”. Source: Coumou and Robinson 2013
Meanwhile, in Egypt, there is another heat wave. According to ABC News,
Egypt’s state news agency says 21 more people have died due to a scorching heat wave, raising this week’s death toll to more than 60.
The official MENA news agency said Wednesday that the latest deaths are from the previous day, mostly elderly people. It says 581 people are in hospital for heat exhaustion.
The Mideast has been hit by a heat wave since late July. Egyptian summers are usually hot, but temperatures this week soared to 46 degrees Celsius (114 degrees Fahrenheit) in the south.
At least 40 people had died on Sunday and Monday, including detainees and patients in a psychiatric hospital, according to officials. It wasn’t immediately clear whether Tuesday’s death toll includes a German national living in the southern city of Luxor who died from heatstroke.
Additional items of interest:
It is not just the editorial page: Study Finds WSJ’s Reporting On Climate Change Also Skewed
It is not a hiatus or pause: The alleged hiatus in global warming didn’t happen, new research shows
It is not just the heat, it is also sea level rise: Catastrophic Sea Level Rise: More and sooner
We are expecting a major Carbon Dioxide sink to eventually stop grabbing CO2 and, possibly, to start releasing it: Global Warming: Earth, Wind, Fire, and Ice