The Small Stuff Matters

There have been several times so far in my short graduate career where I have ended up arguing with one professor or another over something few would think of.  How much does the small stuff matter?  That is, how much does a small change in methods in research matter?  Let me take a moment to talk about why I think that (at least in the context of climate modeling), the small stuff is very important.

Changes in methods reflect changes in the understanding and representation of the physical processes of climate.  Our climate itself includes numerous feedbacks and interactions between the atmosphere, ocean, and land surface.  Say then that there is a small change in the method to represent one physical process.  What does this mean to the representation of the climate system as a whole?  Does it mean only a small change to how climate is represented?  Well, not necessarily.  The numerous feedbacks and interactions considered in a climate model suggests that a small change in method for one process changes how that process interacts with other processes.  Could it potentially cause a ripple effect?

Research in climate and climate modeling has led to the creation of many global climate models and downscaling techniques.  Each of these are different methods, and have different strengths and weaknesses associated with them.  Some global models have a better representation of the El Nino Southern Oscillation, while others have a better representation of aerosols.  In a previous post, I talked a bit about different representations of uncertainty.  One of these representations of uncertainty is called scientific uncertainty.  This type of uncertainty is also called model uncertainty, because it is related to how different physical processes are represented in climate models.  If a small change in modeling methods is implemented into a global climate model, does it increase or decrease the uncertainty?  Depending on why the methods are changing it may increase or decrease the uncertainty.  In an earlier post I commented on an article that indicated that the improvements in modeling in the CMIP5 models would not decrease the uncertainty for the IPCC 5th Assessment Report.    What do you think would happen?

These are a lot of questions that come to mind, and both are related to why the small stuff matters.  What is it about climate that affects what you study?  A small difference in how that is represented can determine if that method better represents the critical process for what you study or a required decision.  Therefore I say, if you don’t know which methods are used to create the climate data you are working with, do you know how well it represents those processes critical to you?  Beware the small stuff then, because it just might be that the small stuff changes what dataset you use and the decision made.

This post originally appeared in July 2013 and is part of our throwback series.

Adrienne Wootten is a PhD Candidate at North Carolina State University.