An observation means nothing unless it can be explained by simulations
The above (quite provocative) statement was recently made by an astrophysicist with 28,561 citations and an h-index of 68 at a Congress of Slovenian Physicists (quote courtesy of Mark Zagar from Vestas).
Of course, the reverse has always been true. At least, since the birth of the Scientific Method. But we modellers are “so nice” that we acknowledge it all the time while observationists within the wind business often blindly trust observations which, at the end of the day, are made by humans and may suffer from unavoidable “human weaknesses”.
At Vortex we are not as radical as that astrophysicist or as Michael White * who says that all observations are wrong, but some are useful… and at least they happened. We take measurements (observations) very seriously but we also have experienced many occasions when they can be wrong and know that models are useful for detecting errors.
As introduced in my last post, Fiction is the new reality, coexistence and complementarity of model results and measurements are now mandatory when approaching a tremendously complex phenomenon such as atmospheric motion.
On the other hand, as a “side dish” to the above course, let me also share here an open question that arose during the recent WindEurope’s Resource Assessment Workshop in Brussels:
If model uncertainty is not related to model error against measurements [and I’m not affirming that it should be], how is uncertainty useful to the user of the model results?
Your comments are very welcome.