Verifying models
-
I have past experience in forecasting and warmly appreciate what you offer here. The Access model that is available for the Australian region has an uncomfortable tendency to massively increase the 1000hPa dew points each evening local time [effectively 12Z]. A comparison with the other models at the same time can see differences of up to 30 deg! The Access having values in the high 20's and the others slightly below zero. The grid points where this happens are strongly correlated with the corresponding wind fields. When the winds are effectively CALM, the Access model ramps up the moisture in an unacceptable manner. Do you guys have the opportunities to do 'experiments'? By that I mean, archive some data from the models over a period and generate a comparison. You could then forward this to the relevant provider and potentially help with them looking into issues they may have with their models. I can only imagine some astute enthusiasts out there might see other instances that might deserve a similar monitoring. If done correctly, your community could be helpful to their improving aspects of their models. Cheers.