I understand there are verification scores used to compare predictive accuracy of models. I am wondering if there is data that looks at the relationship between time until storm event and the predicted placement of storm track vs the actual storm track (or placement of low pressure). So the GFS vs NAM vs EURO, etc would be compared at 8 days, 5 days, 3 days, etc in their predicted track for a given event. At each point in time, you would note the difference in mileage (20,50,100, etc) between the forecasted track and what verified as actual. Of course as the time until the event became less, you would expect less of a discrepancy...think somebody asked something similar a few weeks ago.
Does anything like this exist ?
I am not familiar with verification scores-- maybe that is part of the process of generating them ?