Model validation: Issues regarding comparisons of point measurements and high-resolution modeling results

• Comparison between a high resolution model and temperature observations.
• Temporal variability not resolved in observations.
• Suggested acceptance level from modeled variability.

In this study we compare a high resolution model of waters on the Norwegian Shelf with hydrographic observations obtained during 2009 at Ingøy, a fixed coastal station off northwestern Norway operated by the Institute of Marine Research. The observations comprise snapshots from Ingøy every two weeks, whereas the model represents an average over a certain volume and is continuous in time. We suggest that bias is the best way to compare the modeled and observed times series, while acknowledging the short-term variability (within a day) it is recommended to use the modeled range to estimate an acceptable deviation between single points in the series. Using the suggested method we conclude that an acceptable deviation between the modeled and observed surface temperatures at Ingøy is 0.6 °C. With such an acceptance level the model is correct in 27 out of 33 points for the time series considered.