1 of 1
Offline
Folks. I have route where a coastal high res wind model will cover early stages of the route, with the remainder of the route covered by two global models. Performing Ensemble with all three models throws an error due to the local model not covering the entire course.
Doing an Optimize with the three models does produce a result.
It appears that Ensemble requires all wind models to span the entire route, where Optimize will combine the three to produce a result?
What am I not understanding?
Cheers
Offline
Yes - ensamble routes per wind model loaded. Just Optimise routes with all the currently loaded wind models in the predefined Expedition priority for which model is used at every timestep.
I'm not sure what you mean by what are you not understanding?
If you want a hybrid ensemble route --- what I do is load the models I want use and runt a standard route - then switch to the different combination of models and route --- repeat --- make sure you have a high enough number of routes to save that you don't start lossing routes as you generate new ones.
It isn't as easy as just loading everything and hitting ensemble routing.
The problem is --- how would expedition know how to combine the models in the way YOU want to run an ensamble.
Offline
Exactly. The Ensemble routing runs a route for each model.
The other options merge the data.
Offline
Understood.
What is the rule when I have multiple overlapping models and I do an optimize? Say, HRRR and GFS.
Last edited by WoobaGooba (6/08/2024 6:20 pm)
Offline
Nick will have to give the specifics ---
But from a high level for a given time step - Expedition uses the newest most high resolution data it has. One of the keys there is for a given time step this can create wierd problems if you aren't careful of filpping back and forth between two models if the less favoured model has a higher temporal resolution.
IIRC it favours newer over resolution...
1 of 1