Ensemble or Deterministic?

Posted Tuesday, March 3rd, 2020 by Andy Park

A lesson on all of your eggs in one basket…

TL;DR:
This forecast is an excellent example of the pitfalls in trusting one model run (deterministic) vs. a group of model runs to compose an average (Ensemble) outcome… Rest of the week: Warming and clearing skies over the area until Friday! There is a large amount of uncertainty for the end of the weekend/start of the workweek in the forecast…

Short Term (Today through Friday):
Left over moisture from an atmospheric river event in the Pacific Northwest could drop a few flakes over the resorts today (but nothing remotely exciting). Semi-Zonal flow will set up over the area later this evening transitioning to lights winds aloft as high pressure builds fully over the Great Basin by Thursday. By Friday both the Euro and GFS Ensemble models point to a large amplifying upper level trough over the Eastern Pacific. (Thanks TropicalTidbits.com!)

Long Term: Saturday to Wednesday
By Sunday, the models show a splitting of the trough but disagree on the intensity; Euro is showing a significant but weaker disturbance than the GFS Ensemble.  GFS IVT Ensemble and deterministic model runs also paint a different picture on locations of the upper level trough and intensity of moisture transport. Extended forecast is definitely hard to pinpoint.

It is safe to say that both models outcomes suggest that the large trough setting up by Friday will begin to impact the West by the end of the day Sunday. However, we see a vastly different picture from both models which have serious implications on what will happen and how much precipitation we will receive. Here is the EPS and GEFS compared side-by-side again for Tuesday evening. Notice the large differences in the 500mb heights over Utah!

I think that this is a great lesson in the use of Ensemble and Deterministic Forecasting. If you were to only use one deterministic model run you would potentially be led down a path of deception. Never put all of your eggs (trust) into one basket (model run). This pattern will need to be analyzed moving forward. Below is the GFS deterministic run for 12Z today, notice the large difference to the ensemble run above (left side). Primary concerns are location of the closed low off the coast of California and differences between the Euro Ensemble…

I’m going surfing! Later taters. (Thanks for reading)

Leave a Reply

Your email address will not be published. Required fields are marked *