A question that's been bugging me

[Chris O’Neill]

Hi, I'm new to the forum, first post, hopefully in the right area.

My question: I've lived in quite a few areas across the United States and have noticed a significant regional difference in weather forecast reliability. Is anyone aware of a database or study looking at the range/ reliability of forecasts on a regional basis? It seems reasonable to me that there could be areas with the right ingredients to make them more chaotic, but my meteorology knowledge is limited.

 

[rdale]

Forecast reliability has much more to do with the forecaster than the region IMO...

 

[Chris O’Neill]

While forecaster skills and tools certainly have an effect, I don't think that's what I'm observing which is what prompted my question about any study or database on regional forecast viability. Any chaotic system will have a point past initial conditions where possible states diverge beyond any ballpark prediction. That point will be a function of how well initial conditions are known and the characteristics of the system. So I wonder if there are persistent regional conditions that significantly shorten that time interval compared to other regions.

Hopefully that makes some sense, my background isn't in meteorology, it's in physics so that's the lens I'm looking through, I appreciate the response.

 

[Jeff Duda]

Orography definitely influences predictability, but on shorter time scales.

As far as I've ever known, synoptic scale predictability is pretty zonally constant across the mid-latitudes. As Lance Bosart, one of the best synopticians of our time has said before, the atmosphere does have certain "erogenous" zones where small perturbations often build up into synoptic scale disturbances. One of those zones is in eastern Russia/Japan/far western north Pacific. But once those disturbances are measured, a pretty reliable wave train of troughs and ridges tends to set up across the rest of the hemisphere (all the way from there eastward over to Europe), which renders predictability pretty even across that band.

The tropics have different synoptic scale predictability because the driving flow pattern is completely different (Hadley cell...Walker circulation featuring easterly trade winds at the surface dominates the pattern...synoptic scale disturbances are exceedingly rare there).

Not many people live close to the poles, so people often look at the stratospheric pattern there, which is more on the S2S (seasonal-to-subseasonal) scale there rather than the synoptic scale.

 

[Chris O’Neill]

Wow I really appreciate the response. What is the disturbance that starts in eastern Russia called?

Also it's possible my newness to the subject is getting in the way of me asking questions better. What are the boundaries on what would be considered synoptic phenomenon?

 

[Jeff Duda]

I think the term "Kamchatka low" may have been used to describe the origin of this North Pacific wavetrain that I described above.

Most of the work on predictability that I have followed has been at the synoptic and convective scales. We will forego the latter here, as that is an entirely different discussion.

The synoptic scale is generally defined as having a length scale of roughly 1000-10000 km and a time scale of about 1 week. That leaves the troughs and ridges associated with Rossby waves as pretty classical synoptic scale phenomena. Larger entities such as ENSO, entire wave trains, and other oceanic oscillations (MJO, PDO, PNA) are generally planetary scale phenomena, whereas hurricanes and fronts are a bit small to be synoptic and are considered mesoscale phenomena.

When you stated "forecast reliability" in your initial post, I interpreted that as asking about synoptic scale predictability, since that is about what is understood to non-meteorologists in relation to the "butterfly effect" and wanting to know if it's going to rain 7 days from now; that would be a synoptic scale forecast field. It is at this scale that Ed Lorenz's discovery that weather has an approximately two-week limit on predictability was first derived.