Introduction - Models are not All Wet: Models, Water and Temperature (2) I am starting a series of blogs on models, water, and temperature.
A couple of entries ago, I wrote a somewhat muddled blog,
Difference Between Night and Day. My major goal in that blog was to look at how water, especially water vapor, enters into the climate and climate change problem. I used some regional differences in climate, say Florida's and Arizona's, with the hope of suggesting that we have some intuition of how water vapor modifies regional climate. For example, due to the absence of water vapor, Arizona's extremes of daily temperature are larger than in a much wetter Florida.
This simple intuitive notion, however, quickly falls into complexity. It is the typical complexity of climate science, where the members of a set of simple physical processes combine in many different ways to produce a difficult-to-untangle knot of observations. I will come back to this later, but first, here are some of the other ideas I had in mind in that first blog.
At the end of that blog I referred to the paper by
Kukla and Karl, 1993, Nighttime Warming and the Greenhouse Effect (
from Rood’s Class Website). This paper investigates the observed decrease in the range between nighttime lows and daytime highs. At the writing of that paper in 1993, the models of 20 years ago did not simulate this observation especially well. How does one respond to fact that models don’t represent a particular observation? A common way to respond, sometimes put forward by commenters on this blog, is that the models fail to represent the observations; hence, the model is wrong, and to base any conclusions, actions or behavior on model results is grievous failure of reason.
I, of course, reject this conclusion. When I get the result that the model does not represent an observation especially well, then I take this as a piece of information that motivates further investigation. The scientific investigations of my career have been based on the process that we develop a model from a set of physical laws that are expressed as mathematical expressions. The physical laws and the construction of the original model are based in their most fundamental way on observations. If the model has been developed properly, then it offers an approximation of that observed behavior. If this is the case, then we have an experimental tool that can be used for further investigation. That investigation is motivated both by the shortcomings in the model’s ability to represent observations we already have and by new observations that come along. In this approach models evolve as a tool that help us explore and manage the complexity of the climate system. They also help guide our thinking about the future based on the projections that come from the models. Models are, therefore, devices to help us think; they do not provide the answer.
Another idea that I introduced in the
Difference Between Night and Day was that large changes in the amount of water at the surface, for example, the Dust Bowl and irrigation in the Corn Belt, might have significant regional impacts on climate. The place I am going with this, ultimately, is the
Midwest Warming Hole (2 MB if you click), and that requires thinking about water. The Midwest Warming Hole is an observed feature in the center of the United States that is not warming up as fast as the regions around it or as fast as the models predict. This is not a newly discovered feature, but it is a feature that I think takes on new interest as we think about this hot summer, the last hot summer, and how to use the observations today to think about the climate in the future and how to adapt to a warming climate. The Midwest Warming Hole, and the ability or inability to represent it in models, is also a great example to help people think about how to describe model uncertainty.
The last big theme that I want to follow from the
original blog is the improvement of ways to discuss and understand the role of water – solid, liquid and vapor – in climate and climate change. I did a series Just Temperature (
one,
two,
three) which was motivated by the stunningly warm spring in 2012 in the continental United States and my thinking of extreme events as
climate change case studies. The Just Temperature series used the fact that the warming of the Earth has become large enough that it is possible using temperature observations alone to make a compelling case the Earth is warming. But once we make it beyond that fact, we have to think about water to understand the complexity of both the spatial and temporal structure of the observed trends.
So here are three big themes that I want to organize around:
1) Doing science with models
2) Communicating the role of water in climate and climate change
3) Thinking about changes in land use and its impacts on water
These will be interspersed, of course, with some tangents to interesting subjects here and there. But those who know this blog know that eventually I get there.
r