NWS Point Forecasts

[Morgan Palmer]

I know someone has the answer.

The NWS point forecasts have gone from "experimental" to non-experimental... at least the "experimental" wording/label is gone.

Are these based off the Nat'l Digital Forecast Database, or the individual zones from the NWS office? I do not have an example handy, but over time, I've been inclined to look at the "points" (or whatever they're called), then look at the zones from the offices. Because of years of habit, I tend to trust the zones more when just glancing and not doing my own forecast.

I have noticed occasions when points just a few miles apart may have very different forecast conditions, particularly when examining points are on one side and then another of a CWA boundary.

Curious. Are these point forecasts now generally accepted, or do folks prefer the zones, like I still do.

MP

 

[Anthony Silver]

If I'm not mistaken, they are based on the zones. I dont think they have anything to do with the National digital fcst database.

I personally like the zone forecasts, especially if I'm just needing local weather information to plan ahead for outdoor recreational activities.

 

[rdale]

Nothing comes from the zones - everything comes from NDFD. The met makes his grids, and from that the ZFP is generated and the data goes into the system for your point-n-click display.

- Rob

 

[Mike Umscheid]

Are these based off the Nat'l Digital Forecast Database, or the individual zones from the NWS office?

Point forecasts on the web are generated from the NDFD. They are generated automatically and forecasters have zero flexibility to go in and make manual edits. It is entirely automatic.

I do not have an example handy, but over time, I've been inclined to look at the \"points\" (or whatever they're called), then look at the zones from the offices. Because of years of habit, I tend to trust the zones more when just glancing and not doing my own forecast.

I have noticed occasions when points just a few miles apart may have very different forecast conditions, particularly when examining points are on one side and then another of a CWA boundary.

NDFD is not perfect. There will be times of interoffice inconsistencies, especially in dynamic mid-latitude weather. This has always been the case, but now with NDFD, it is more visible. We forecasters (at least in central region) must meet passed down guidelines for a more "seamless" NDFD database, but there will be times when disagreements come about. This is why interoffice collaboration is now a very regular aspect of the NWS forecast process.

Also, remember one other thing. You will sometimes see product inconsistencies between the ZFP and the point-forecast. One is a point forecast and the other is an areal forecast. Of course with areal forecasts, you will get a more generalized forecast averaged over the domain of the zone group. The point forecast is (when looking at the point forecast meteograms) essentially a regurgitation of the grid value for that weather element at that exact location and time. There are some different text formatting rules for the point forecast text output versus the ZFP, and I'm not exactly keen on what those are precisely. But yes, you will see differences at times, but usually not overly significant.

Mike U

 

[Anthony Silver]

Whoa...looks like I was misinformed about this.

 

[rdale]

"occasions when points just a few miles apart may have very different forecast conditions, particularly when examining points are on one side and then another of a CWA boundary. "

I still have the email archived somewhere from the SOO list - where NWS HQ suggested that local offices NOT run models on their own, because it could create too dramatic of inconsistencies on the boundaries... Classic!

 

[Morgan Palmer]

All excellent explanations. I now consider myself informed, so I can stop complaining, and start using these products more wisely. Thanks all.