RE: Can we still run PoWT tools individually or will they be going away?

Forums (Forecast Builder)

RSS (Opens New Window)

Joseph Moore, modified 8 Years ago.

Can we still run PoWT tools individually or will they be going away?

Youngling Posts: 72 Join Date: 11/1/13 Recent Posts

It's no secret that ForecastBuilder is mainly a wrapper for the great PoWT methodology. What's in store for the future? Is the idea that all the "manual" PoWT tools would go away and ForecastBuilder would be the only way to edit weather grids? Or will they still be allowed to exist separately in the future?

My two cents: I prefer running each PoWT tool manually because I need time to analyze and collaborate changes as I go through the forecast process. I don't like the ForecastBuilder tool "waiting" for me - personally I feel like I'm being rushed along to just accept the defaults and go go go! (The training makes it pretty clear we shouldn't just be hitting "Ok" a few times to get a forecast out in minutes.) I like being able to run the tools manually because often I have to run PoTCreatePrecipTypes numerous times as I fine tune the forecast and collaborate.

If we can still use the PoWT tools seperately, then the question is... why have the ForecastBuilder wrapper if we're proficient in PoWT? Besides the loading of the same starting dataset (which I am fully on-board with!), it seems like we're having our hands held along using PoWT for the rest of the tool. I realize this is better for those unfamiliar with PoWT, but as someone with experience with PoWT I'd rather run the tools on their own.

Andy Just, modified 8 Years ago.

RE: Can we still run PoWT tools individually or will they be going away? (Answer)

Youngling Posts: 89 Join Date: 6/2/15 Recent Posts

Joseph,

It is true that for those really proficient with PoWT that ForecastBuilder may feel a little like a "step" back. This includes even here at my office ARX. One of the many goals of ForecastBuilder is to get everyone on the same page in building the forecast, in the manner that keeps the forecast internally meteorologically consistent. That was one of the downfalls of PoWT, is remembering what step to do when and truthfully you could still create a forecast that was not meteorologically consistent by following improper step order. Another problem with the PoT_CreatePrecipTypes procedure was a missing integrity check that is now included within the Precip Types / Accumulation Grids step of ForecastBuilder.

With ForecastBuilder you can pause as you go through the process, based on time periods of the forecast. Or, you can jump ahead in the process if you are happy with the grids before. A good example is I need to shift a warm nose north for tonight to help produce more sleet/freezing rain/rain. For this just jump to the top-down step at the beginning GUI.

Inevitably I would like to remove PoT_CreatePrecipTypes and PoT_MergeWx, fully incorporating them within ForecastBuilder vs calling them as ForecastBuilder is doing now. But I'm waiting to see the feedback from the test to see if we can improve the ForecastBuilder GUI framework. I know there is more work that can be done.

Joseph Moore, modified 8 Years ago.

RE: Can we still run PoWT tools individually or will they be going away?

Youngling Posts: 72 Join Date: 11/1/13 Recent Posts

It really does feel like a few steps back, coming from someone who uses PoWT for every imaginable scenario and has been an evangelist for it in our office.

You won't hear any complaints from me about addressing the consistency issues - the fact that even moving from one office to another within CR (let alone outside the region!!!) means re-learning a lot of the forecast process because of the different tools/methods is incredibly frustrating and a serious waste of our time. Of course, I don't want to see this go the other way where R&D is eliminated from the WFO level and only a few select people can make improvements to GFE because there is a strictly enforced method for creating the forecast. I think there's a happy medium, and while I think the current iteration of ForecastBuilder swings too far in one directions, I'm hopeful the team will take into account our feedback and adjust according to our people and sound science.

Regarding PoWT and not knowing what step to do when, I feel like ForecastBuilder suffers the same problem because of the number-labeling issue I've addressed in another Vlab forum post. One of my biggest fears is losing all my hard work because I accidently run ForecastBuilder at the wrong step - our forecast and PoT grids are already being blown out every 12 hours anyway, though.

I would be really disappointed if you removed the base CreatePrecipTypes and MergeWx tools, but as long as I can still run ForecastBuilder in the same manner I guess it's okay - as I've said elsewhere, I'm just afraid it's the default choice to not edit any grids and thus remove the forecaster from the loop as the ForecastBuilder GUI sits and waits, progress bar swinging left and right rushing us along to just accept the default forecast.