Doesn't have to be constant, if your aim is simply to estimate whether or not there is an acceleration factor. A quadratic should qualitatively reveal the presence of an acceleration factor, if it does a better job of fitting data than a linear trend. If the aim is to accurately estimate the acceleration, then yes, a quadratic would only work for this when the acceleration is constant.A_Gupta wrote:If you assume that the acceleration is a constant.sudarshan wrote: If you can fit it with piecewise linear curves, a quadratic would still do a decent job of showing the acceleration trend, right?
OK, so the claims *are* based on model predictions. I think it's simply a question of - how much do you trust the computer models, then?Most of the sea level rise depends on what you predict for the fates of the Greenland and Antarctic ice sheets.Now my query is - where are these claims of 1 to 2 meters of sea level rise per century coming from?
The issue is this:
* When it comes to experimental data, everybody believes them, except the person who collected the data
* When it comes to computer models, nobody believes them, except the person who built the model
While the lay public is often in awe of computer modeling, scientists who aren't directly in the field of interest, but in related fields, tend to be more skeptical - "what did you feed into your model? how did you validate it? what are your governing equations?" It comes down to this trust deficit, I think. I know the kind of approximations and assumptions that go into computer models (having built many such models myself), so I inherently trust measured data (having also done plenty of experimental investigations myself), such as the NOAA tidal data, much more than what I read about model predictions.
But my basic premise was this. Given that there are so many claims and counter-claims regarding climate change or sea levels or global warming, it should be possible for somebody who is not directly in the field, to tell who is right or who is wrong, by doing a few simple "truth tests" with available data and basic reasoning. That is where I was going, and that's why I was trying out the most basic tests with tidal data or other such available data to figure out which side of the debate was "truthful" and which one was propaganda.
I think my premise was flawed to begin with. There's some truth on both sides, and some propaganda on both sides. Simple "truth tests" aren't going to cut it here. They might help with revealing media bias and propaganda, but sorting out the presence or absence of propaganda in the science part of it with such simple tests isn't going to work.