Jump to content
  • Member Statistics

    17,585
    Total Members
    7,904
    Most Online
    23Yankee
    Newest Member
    23Yankee
    Joined

Jan 11-12 Model/Forecasting Discussion


Ji

Recommended Posts

  • Replies 6k
  • Created
  • Last Reply

From a modelling perspective, how does the OP become such an outlier within it's ensembles at 72 hrs? If I understand how the ensemble members work that seems nonsensicle that the op varies from all 11 other members at 72 hrs.

I wish I had an answer...in my opinion...the OPS is the model configuration that is run at the highest resolution with what is considered by the respective operational center to have the most optimally performing (or verified) model parameterizations (things like how cumulus convection is handled, boundary layer evolution, radiation and land surface physics..to name a few). Ensembles are run with either different physics or perturbed initial conditions with main goal of assessing the variability of a forecast.

In general, the OPS is probably the most trusted solution on average..statistically speaking...yet nothing precludes the OPS solution from being completely different from an entire suite of ensembles which show a low amount of variability (around their solution). But to that point, either could be right, the GFS OPS solution is just on the flatter, OTS envelope of possible solutions. There is a good physical explanation why, but nothing sticks out from my quick analysis...that's why we leave it to computers to solve all these non-linear differential equations...

Link to comment
Share on other sites

I wish I had an answer...in my opinion...the OPS is the model configuration that is run at the highest resolution with what is considered by the respective operational center to have the most optimally performing (or verified) model parameterizations (things like how cumulus convection is handled, boundary layer evolution, radiation and land surface physics..to name a few). Ensembles are run with either different physics or perturbed initial conditions with main goal of assessing the variability of a forecast.

In general, the OPS is probably the most trusted solution on average..statistically speaking...yet nothing precludes the OPS solution from being completely different from an entire suite of ensembles which show a low amount of variability (around their solution). But to that point, either could be right, the GFS OPS solution is just on the flatter, OTS envelope of possible solutions. There is a good physical explanation why, but nothing sticks out from my quick analysis...that's why we leave it to computers to solve all these non-linear differential equations...

Thanks for the reply.... fascinating stuff

Link to comment
Share on other sites

From a modelling perspective, how does the OP become such an outlier within it's ensembles at 72 hrs? If I understand how the ensemble members work that seems nonsensicle that the op varies from all 11 other members at 72 hrs.

See my post above re: version/resolutoin...also there are actually 21 members.

Link to comment
Share on other sites

I wish I had an answer.....that's why we leave it to computers to solve all these non-linear differential equations...

So here is an "interesting to me question" for ya:

Is this particular pattern relatively easy or relatively difficult for a model such as the GFS to

generate consistent solutions?

My less than informed oninion is that it should be easy--non-challenging--because the pattern

seems to be in stasis--equilibrium--it continues to set up nearly the same with each system.

Sure, each system is unique but they all seem to intensify and weaken in the same regions throughout

the season.

Link to comment
Share on other sites

so are you saying the GFS ensemble mean no longer has a better verification score then the op run?

It still does, but NOT in the short range (I don't know the cross over point, maybe day 5 or something). The OP is usually better than the raw ensemble mean in the short range.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.

×
×
  • Create New...