Need to just point out some things. 1. agreed the vertical scale is in metres, but the horizontal scale is in 1000s of years. I can see that the last 250 (definitely more) odd years (for ease of calc) that the change is almost flat. If the level change was 1000mm (1m) we would have seen it. So the change is definitely less than that. I would not subscribe a value larger than 0.25m in the last 250 years in the most conservative case otherwise we would have seen it on that graph. Now consider that we had maunder minimum occurred around 300 years ago from which we seeing a net increase. So how much per year do you subscribe? whatever you like, its in a millimetre/year or very close to it! Now compare to the GIA correction. Which bit of this extra conservative analysis would you disagree with?
2. about your calculation. I agree with the math but I think you might have lose the purpose for the calc :) What you have done is treat the "c" as a basis. That if two distant years use the same correction then "c" cancels out of correction. Agreed if all uncertainties are ignored. But I am talking about prediction. What I am saying is that the "c" we are applying is of similar magnitude to the observed acceleration over the past century. Let me explain another way.
We have observed an acceleration over the past century. Now over the last decade we have decided to fudge using a GIA. Its in the application of the models, that are predicting these rises, where the mishap will happen.
We will be modelling something similar to sea_level_rate(t+1) = f(sea_level_rate(t)) + acceleration(t)*(1 year) + correction.
This is where the magnitude matters. The prediction will be massively determined by the correction! I would prefer that the past observed acceleration be somewhat important and well represented in the predicted data. Definitely more than a recently discovered fudge factor. What do you think?
