|In Science, Sometimes The Answer Is "No"
||[Nov. 16th, 2011|10:40 pm]
I have been doing the research for my poster at AGU. (Which is in *gulp* a bit less than three weeks? Yikes!)|
Last year I concluded (somewhat to my surprise) that if you're interpolating our model data onto a different grid, it is actually worth using a fancy-pants statistical method based on math complicated enough that you have to use someone else's code, rather than the a boring straightforward method with math you can actually wrap your brain around.
This year, I'm investigating whether it makes a difference if you use the fancy-pants method with the added bells and whistles (separate land/water interpolation and elevation as a covariate (okay, so it's just one bell and one whistle)), versus the plain old vanilla fancy-pants. I mean, it certainly makes the pictures look a lot more detailed, right?
So far the answer is shaping up to be... no, not really.