Senate calls for NWS consolidation...again

Jeff Duda

site owner, PhD
Staff member
Supporter
This is not the first time I have heard this, even in recent years.

http://www.washingtonpost.com/blogs...-regional-offices/?postshare=6771434473663590

I would really like to hear the opinions of some NWS forecasters about this. Not being one myself, I tend to give the Senate the benefit of the doubt in a situation like this. I think advances in NWP are making human forecasters (at least over a broad region covered by NWS offices) obsolete, and I know I'm not the only person in meteorology who thinks that. I think humans will continue to be useful in decision making for extreme events (severe weather, flooding, tropical systems, winter events etc.) and are still useful for obtaining observations and maintaining equipment. But as far as the need for humans for routine forecasts by shift...that I'm not so sure of. I'm also not convinced that the extra knowledge of the local area that many NWS employees and supporters cite as a reason for not consolidating still makes them indispensable. NWP models now have the complexity and resolution to resolve terrain and land use inhomogeneities (among other things) that humans have probably known about for years and used to make adjustments to previous-generation NWP forecasts that could not resolve mesoscale features. Besides, don't most NWS forecasters just use various blends of operational models in their gridded forecast data nowadays anyways? When I volunteered at DMX in 2008, basically everyone was doing that.
 
I am not an NWS employee, but, any type of human interaction in the forecast process is critical. I also think that total dependence on computer models for routine forecasts, especially those on a daily basis is complete laziness. NWP serves a function of giving the forecasters an idea of a potential scenario outcomes, but the human forecaster has one thing a computer doesn't, human logic and experience, which has been shown time and again to trump the best of computer model output. Models are not now, nor will they ever be so perfect, that they can be used to completely replace a human forecaster. You can automate functions, but as far as scientific knowledge and experience, no way! Also, I think it's ridiculous to discount the local knowledge in forecasting. This adds quite a bit of value when local affects like topography and other subtle mesoscale effects come into play. The models have all kinds of problems parameterizing all kinds of these effects. Local forecasters are crucial for correcting these errors in the final forecast. One final thought, there are 122 local WFO's for a reason. A regional forecast center approach might have advantages by consolidating some functions, but for local forecast & warning support, the local WFO's are critical. The senate bill is a wrong idea.
 
Last edited:
"Besides, don't most NWS forecasters just use various blends of operational models in their gridded forecast data nowadays anyways? When I volunteered at DMX in 2008, basically everyone was doing that."

Only because they're basically forced to. The NWS has been moving towards a centralized regime for a while. Everyone pretty much uses the same damn blends to avoid problems. If you wanna change it, you have to argue with your neighbors.

So yes, they have.. But it doesn't seem like that's what everyone wants.
 
The NWS budget is about $1 billion a year from what I read. That is not that much consider how many billions of dollars each year is lost from wx. Only a small increase in better forecast models would save that much easy per year. They want to re-allocate the $$ saved by this reorg to things like improving modelling, but you can't have it both ways and expect savings. That gain from better modelling could easily be lost due to the issues with messing with the current WFO structure and their mission!

I'm all for making things better and more cost efficient, but robbing Peter to pay Paul is not the answer. Given how wx is so front and center in
the mainstream media and the progressively increasing costs from it each year simply due to increasing population and infrastructure, it is disappointing that
a modest increase in the NWS budget can't be hammered out. You don't look to consolidate things in such a varied and stormy climate the CONUS has.
 
This seems to me to be a flagrant case of politicians trying to micromanage something they don't know much about. I do think local knowledge is important in weather forecasting - i know it is in places like Colorado and New Mexico where terrain and other factors make weather processes quite different from what they are in Oklahoma, Iowa, or Illinois, for example. And I can't think forecasting will be improved when you have people making forecasts for places they don't even know the location of. I know they propose to keep local WCMs, but will that be enough in "all hands on deck" situations such as tornado outbreaks and major flash floods? In short, I do not think Congress should impose organizational structure on the NWS. Nothing about the functionality of Congress suggests to me that this would be a good idea, either.
 
Although the measure mandates centralizing forecasting operations at six regional offices, it would not result in closure of any of the existing 122 forecast offices. Rather, it specifies that these offices maintain a warning coordination meteorologist to serve as a liaison with emergency management for storm preparedness and response activities as well as to conduct media and public outreach. Offices also would continue to maintain radar instrumentation and launch weather balloons.
 
“Likely it would mean the elimination of over 1000 meteorologists jobs,” said Dan Sobien, president of the NWSEO. “It would take a decade for the field of meteorology to recover from a blow like that and those meteorologists to be absorbed back into the enterprise.”
 
I don't understand how local offices are overwhelmed during sever weather outbreaks. My weather forecast comes from Springfield since Missouri is so rural and they've handled outbreaks just fine. I do want to say that, however, it is frustrating sometimes when things are going down and they stay with that one thing with out looking ahead for other potentially hazards weather.

And from what I've seen out of the OKC area, those guys seem to be on point when it comes to doing their job as well. They seemed to be well 'staffed' with spotters, and extra crew members on hand in the 'weather lab' to help pass info along.

I guess what I am trying to say is that this bill is a bunch of bull hockey. Maybe if those who want to do this sit back and look into the NWS offices that they control and see what could be done better, maybe, just maybe forecasting could be better.
 
Some sort of reorganization is pretty much inevitable. This is pretty extreme, but I would not be surprised if something like this happens within the next few years.

Also, the robot revolution is coming. If it can be automated, it probably will be. I can't think of a better example of something that lends itself well to automation than weather forecasting, at least to a first approximation. Martin Ford's recent book "Rise of the Robots" is worth a read.
 
I'm going to guess that by "approved", that it's being reconsidered by the committee.

I'm sure I'll get flamed for saying this, but simply saying "increase funding" is not a good answer in my eyes. All the increased funding in the world will do no good if it's not used efficiently. Now, if the ones that say to increase funding can put forth an actual, solidly logical plan on how that extra money can be used, then I can get behind it. But simply saying to increase funding for a very vague set of things is not good enough for me.

Being a veteran, I know first hand how simply throwing money at a problem doesn't do any good. The VA is no less of a mess with more funding now than it was when all the controversy broke out.
 
I've been a computer programmer for over 30 years. You can write the best, most thoroughly debugged and tested application, and still lack confidence in it. I'm talking about business applications, which are far less complicated than the software use to predict the weather. The fact is, humans will have to be involved in the forecasting process. Computer models play an important role in weather forecasting, but only to a certain point. Computers play an important and indispensible role in daily life, but they can't do everything. I say leave the current system in place.
 
Mike - whoever told you the goal of restructuring was to remove humans misunderstood. That's not in question. Our current breakdown of WFOs is clearly broken, and needs to be fixed. This bill however circumvents the work the NWS had just started to find new ways to deliver info so that's why people are not comfortable with it. Changes are coming...
 
I am not part of either the public or private weather enterprise, so this post should be viewed only as questioning and seeking to learn... but isn't SPC a fantastic example of the successful centralization of forecasting? Yes, I realize they do not issue warnings, but I don't think that invalidates the comparison.

Also, many of the private weather forecasting companies focus on offering hyper-local forecasts, and to my knowledge they don't do it by having office locations all over the country...

Anything that involves change and the potential loss of jobs is disconcerting to those in the field and likely to be affected by it. I certainly understand that. But if looked at objectively and unemotionally, might opinions about this be more favorable?

I find the idea of having a local NWSFO comforting in a nostalgic sense but, like the demise of the local hardware store in favor of the more distant Home Depot, I think this sort of thing is inevitable. It is similar to the move in business to the use of shared service centers for back-office and corporate functions.

Having said that, I still think there can be plenty of opportunities in meteorology despite NWP, just like there is still a huge demand for accountants despite automation in that field (which happens to be my field, so I can speak from experience on this point...) The roles and jobs have evolved, but they are still there...
 
I am not part of either the public or private weather enterprise, so this post should be viewed only as questioning and seeking to learn... but isn't SPC a fantastic example of the successful centralization of forecasting? Yes, I realize they do not issue warnings, but I don't think that invalidates the comparison.

It does... SPC is a broad scale outlook. Many times you'll see locals react to a particular watch with a "huh?" It would be comparable to forecasting the state of Oklahoma to have highs in the 85-100 degree range today. Probably correct - but not very usable.

Also, many of the private weather forecasting companies focus on offering hyper-local forecasts, and to my knowledge they don't do it by having office locations all over the country...

Correct, but they don't do it with humans. Most "hyper-local" forecasts are all computerized. And if you have one of those apps you'll see they can be very good during quiet weather periods. And very VERY bad during active weather.

Anything that involves change and the potential loss of jobs is disconcerting to those in the field and likely to be affected by it. I certainly understand that. But if looked at objectively and unemotionally, might opinions about this be more favorable?

No. Remember that forecasting is only half of the equation. You can have the best forecast in the world - but if you can't communicate that to people it doesn't do a lick of good. Having a WFO helps create relationships - and through relationships the locals & mets can better coordinate what they need and how it can be provided.
 
....SPC is a broad scale outlook. Many times you'll see locals react to a particular watch with a "huh?" It would be comparable to forecasting the state of Oklahoma to have highs in the 85-100 degree range today. Probably correct - but not very usable....

True, but isn't that a function of the nature and intention of specific products such as the Convective Outlook and Convective Watches, as opposed to a question of capabilities? SPC's ability to drill down to the local level is evident in Mesoscale Discussions, which often will note individual storms.

Good point in your post about the relationship aspect of local NWSFOs though. That's one of the reasons why the national CPA firm where I work has a strategy of lots of local offices even though much of the work could be done remotely/virtually or by putting people on a plane.


Sent from my iPhone using Tapatalk
 
Back
Top