I'm a professor at U Michigan and lead a course on climate change problem solving. These articles often come from and contribute to the course.
By: Dr. Ricky Rood , 7:47 PM GMT on August 07, 2012
Ledgers, Graphics, and Carvings: Models, Water, and Temperature (4)
This is a series of blogs on models, water, and temperature (see Intro). I am starting with models. In this series, I am trying to develop a way to build a foundation for nonscientists to feel comfortable about models and their use in scientific investigation. I expect to get some feedback on how to do this better from the comments. In order to keep a solid climate theme, I am going to have two sections to the entries. One section will be on models, and the other will be on a research result, new or old, that I think is of particular interest.
Doing Science with Models 1.1: In the previous entry of this series I argued that if one considered the types of models used in design and engineering, then we use models all of the time. In fact, when we build or do just about anything, we use some sort of model to get us started. I ended the previous entry with the example of building a simple picnic bench that would hold three, two-hundred-pound men. Not only do the materials need to be of sufficient strength, but the legs of the bench need to be attached in a way that they form a solid and stable foundation. If the bench wobbles and the legs spread apart, then it will be unsafe. If we have experience of some sort, we construct a model from this experience. For example, if we have built or repaired tables and benches we have some ideas of good and bad construction. If we have no direct experience then we can find or ask about plans. These plans might be a schematic, a graphic model of the bench.
For those who do not build benches, but who, say, balance their checkbooks, there are models as well. The forms in a ledger represent models that have proven usable through practice or that have become standard approaches. Information is collected and organized: the check number, the date, the payee, the amount, the purpose and the category of expenditure.
These graphic, tabular, or touchable models are common enough that we develop intuition about their use. Introductory materials to climate models often use the words “mathematical,” “numerical,” and “computational.” These words take us not only away from our intuitive notions of models, but also into subjects that many of us find difficult and obscure. However, in the past couple of decades we have seen the tabular models of checkbook balancing coded as computational products such as Quicken. Design and architecture move to tools such as Computer-assisted Design. Recently, we have seen this combination of the world of digital models and touchable products come full circle with the advent of three-dimensional printing. In three-dimensional printing, solid objects made of plastic and metal are rendered from mathematical descriptions of the objects. I will return to this idea of mathematical descriptions of objects later. The point that I would like to make now is that using computers as tools to represent the real world has in the last two decades become routine. Therefore, in and of itself, the use of computers to make numerical calculations of the real world is common. It might not be as universally intuitive to people as a ledger or a wooden design of a boat, but there is large body of experience that affirms the value of computer-based modeling.
There are a number of steps that need to be taken from here to climate models. So far, I have been talking about models that are in the spirit of a work or a structure used in testing or perfecting a final product. In climate modeling, the final product of the construction is a model. It is the purpose of that model to provide a credible representation of the climate. That representation has a number of attributes. There is the attribute of representing what we have already observed. There is also the attribute of predicting what we will observe, that is, predicting the future. Therefore, the final product of the whole process is the simulation of and the prediction of the climate.
As with many words, there is more than one definition of model in the dictionary. Another relevant definition from my print edition (third) of the American Heritage Dictionary is “A schematic description of a system, theory, or phenomenon that accounts for its known or inferred properties and may be used for further studies of its characteristics.” (American Heritage Dictionary online) This definition is directly descriptive of a climate model. But like those introductions to climate models that I referred to above, it quickly goes to words like “system” and “theory” that are not quite as intuitive as I would like. This is where I will start next time.
Interesting Research: Attribution of 2011 Extreme Weather to Climate Change - Some might recall in 2011, I wandered into the contentious subject of the attribution of climate change to humans (collected here) and talking about communicating extreme weather events in the media (Shearer and Rood). The paper I highlight in today’s blog is a compilation of efforts to understand the role of planetary warming in some of the extreme events of 2011. The paper is Explaining Extreme Events of 2011 from a Climate Perspective edited by Tom Peterson and others and published in the Bulletin of the American Meteorological Society. This paper looks at six of the extreme events of 2011 and tries to attribute, in a variety of ways, the role played by human-caused global warming. (nice summary in New Scientist)
I want to focus on the part of the paper that discusses the extreme heat and drought in Texas in the summer of 2011. Much of that discussion is based on evaluating the effect of sea surface temperature, and specifically, the role of El Nino and La Nina. El Nino and La Nina are the names given to recurring patterns of sea surface temperature distributions in the eastern, tropical Pacific Ocean. The approach to this problem is to use models to make many simulations with sea surface temperature distributions similar to the La Nina conditions of 2011. Simulations were made for times in the 1960s and for the year 2008. The simulations provide an ensemble of many plausible outcomes, and it is possible to investigate the odds of a drought of similar extreme attributes as the 2011 drought occurring in the 1960s. The authors conclude that the warming climate made the 2011 drought 20 times more likely to occur now than in the 1960s. The authors point out that they cannot make statements about absolute probability. That is, they cannot state that in the absence of carbon dioxide increases and associated warming, that the drought would not have occurred.
This approach of using probability to discuss the impact of warming is an active area of research as well as an emerging way to communicate the relation between extreme weather and global warming. In the Washington Post, Jim Hansen has an op-ed piece that describes a paper which was released on Monday, August 6 (reference at end). In this paper Hansen revisits his metaphor that compares extreme weather in a warming climate with playing a dice game with loaded dice. That is, the dice are loaded in a way such that what used to be “extreme” will more likely occur. Going back to the Texas drought, that result mentioned in the previous paragraph says that the dice are loaded so that the extreme attributes of the 2011 drought are 20 times more likely. The takeaway message from Hansen is that we have, so far, underestimated how much the dice are loaded and that we have underestimated the probability of extreme events such as droughts, floods, heat waves, and yes perhaps, persistent cold snaps.
Hansen, Early Edition, PNAS, Perception of Climate Change
Hansen, Perception of Climate Change, Public Summary
The views of the author are his/her own and do not necessarily represent the position of The Weather Company or its parent, IBM.