How can industries and engineering adapt for extremes and threshold exceedances if models deviate from observations out of interquantile range?
This is why bias correction of model data is important, to remove systematic difference between model representations and real-world conditions. Having said this, the climate conditions of the past may not be fully represented by estimates derived from currently available, incomplete obervations. It is important to analyse the 'behaviour' of models in the current climate carefully before drawing conclusions regarding exceedence of thresholds.
