WSI Rain/Snow differentiation

[Backscatter]

Does WSI have the ability using RPM to show future rain and snow? I am wanting to show the differentiation. You know... kinda like WXC has.

Can this be done using different RPM contours? If so, anybody know what the contour palettes are called??

Thanks in advance.

[TAFKA wacowx]

Grid algebra lesson begins in 3....2....1

[Backscatter]

Seriously? That's the only way? Is this easy or quite time-consuming?! We are new to WSI!

[TAFKA wacowx]

I'm not an expert, but there are many folks on here that can help you with the grid algebra. As far as I know, that is the only way to do it.

[WeatherSlave]

You might want to use some pre-made USA products and just change the location parameters to your viewing area (base map, map orientation, etc.). If you don't have these to draw from you'll have to go the grid way - which is a "slang term for a female dog",,, that is to say WSI sucks at making this easy!!!

[tvwxguy]

As far as I know, that is the only way to do it.

you are correct old wise one.

[Backscatter]

Why can't you just add another contour for hourly snow going along with hourly rain?

I'm not at work now so I don't even know if that is possible.

I just came across a future wx map at intellicast.com and saw future rain/snow areas and thought that would be cool.

[TAFKA wacowx]

You need to speficy what 'hourly snow' is. Hence using grib algebra. It's not as it was for you on WXC. There is no such variable on RPM.

I know that in the past there were some folks who have posted the entire grid algebra text to help those who had not figured it out on their own. I'm sure someone could offer you one you could tweak.

[MU_WX_Grad]

WSI has pre-built WxPRO RPM products already configured to show the difference between rain/snow/mix.

Just copy those products to your desktop (or your show) and change the map backgrounds. That's all you have to do. It's pretty easy. This is assuming you have the RPM data. No grid algebra involved, too!!

[tvwxguy]

Why can't you just add another contour for hourly snow going along with hourly rain?

I'm not at work now so I don't even know if that is possible.

I just came across a future wx map at intellicast.com and saw future rain/snow areas and thought that would be cool.

It looks bad... it you want it to be accurate, grid algebra is the way to go. I (or MIguell) can help you set up a script if you want. Once your script is built, just put it in Scheduler and all is good to go.

Lemme know.

[aggiezg05]

WSI has pre-built WxPRO RPM products already configured to show the difference between rain/snow/mix.

Just copy those products to your desktop (or your show) and change the map backgrounds. That's all you have to do. It's pretty easy. This is assuming you have the RPM data. No grid algebra involved, too!!

This is our winner. However, it takes quite a while to render. It is definately your best option.

[Backscatter]

Thanks for all the responses.

I am subscribed to my own thread to get email notification (instant) and I don't get emails! Anybody else have that problem?!?!

Back to WSI...

I do everything in ShowFX. Prism renders a lot of stuff that we set up. We don't know any different, since we went with WSI this past summer. These "SEB's" or whatever are there, but we don't ever touch them. I guess because it's all 2D stuff?

Do you mean bmy "scheduler" those things that automatically render SEB's?

I know this sounds stupid, but we picked up on WSI with TVI 3.0, etc. running on Dell 670s.

I will investigate the WSI has pre-built WxPRO RPM products.

Thanks again.

[Mayhem]

I set up grid algebra on our futurecast to draw rain, snow, and mixed precip. What's neat about using the grid algebra is that you can select your parameters. I have one that uses the 850-1000 thickness at the 130 line, another that uses the 540 line, and I can even change to the 541 line or 539 line to get a better initialization. You can also use the GFS, NAM, etc. That's the neat part... you can tweak it yourself, as opposed to the RPM, which you are stuck with whatever they say.

If you would like the equation and how to set up the grid algebra, just PM and I will be glad to help.

[Backscatter]

Thanks Mayhem. PM sent.

[MIguel W foX]

I set up grid algebra on our futurecast to draw rain, snow, and mixed precip. What's neat about using the grid algebra is that you can select your parameters. I have one that uses the 850-1000 thickness at the 130 line, another that uses the 540 line, and I can even change to the 541 line or 539 line to get a better initialization. You can also use the GFS, NAM, etc. That's the neat part... you can tweak it yourself, as opposed to the RPM, which you are stuck with whatever they say.

If you would like the equation and how to set up the grid algebra, just PM and I will be glad to help.


You can make the rpm give you whatever rain/snow line you want as well.

[Backscatter]

televisionwxguy and smi:

Thanks for your help too. I am going to mess around with some grid algebra stuff sent my Mayhem.

Coming from Galileo, WSI is worlds and worlds better :rockon: :D

[Mayhem]

You can make the rpm give you whatever rain/snow line you want as well.

well kiss my grits. We don't even have the RPM, so I was assuming (you know what happens there). I still would prefer the algebra just because I can mix and match models and make my own ensemble practically.

Things get really weird when you use the NAM 1000mb height and the GFS 850 height to get a thickness! Then plot with NGM moisture fields!

YIKES!

[MattyD.]

Does WSI have the ability using RPM to show future rain and snow? I am wanting to show the differentiation. You know... kinda like WXC has.

Can this be done using different RPM contours? If so, anybody know what the contour palettes are called??

Thanks in advance.

Okay- Hold on... Everything can be done in ShowFX. If you have the RPM there is no real complicated grid Algebra involved. If you don't (then like us) sucks for you and you can spend a lot of time writng Grid Algebra which is just god awful. Some stations do it the very "SH^^^Y" way and pretty much write a very simple Grid Algebra stating that if the surface temps are less than 30 snow... between 31 and 33 mix... and above 34 rain. Others get really involved and work in thickness and temperatures at different heights... None of it is fun.

Trust me- the RPM is NOT perfect but it is better than the long GA way. Any tech support person, should be able to fax you the RPM stuff for it or email it too you.. Or you can click on the help menu on the upper right corner of your Linux box... and it will open up the WSI style Broswer and see it all there.

You will need to make color palletes to make it look really good. It all takes work, especially if you are new to this system... I expect your cursing to begin within 30 seconds after you begin this project.

I will apologize now for the problems that you will encounter. I might suggest A lot of Advil or Tylenol mixed with very stiff drinks.

[Mayhem]

Hey! I take offense! I find grid algebra very fun, once you get to understand it anyway.
The only real downside is that each precip type is it's own seperate contour, so render time can be 15-20 minutes for a 24 hour forecast.

[MI_Wxman]

Ask WSI for the RPM model folder (it should already be on the system).

Look for WSI_RPM_WSM_48hr_Loop
Grid Numeric Algebra already added.

That should give you what you're looking for.

"Turd has spoken" :)

[50 Percent Chance]

Hey folks --

If you prefer to do this in Showfx, I have some prebuilt scenes using the same grid algebra as the prepackaged wxpro products.

You would simply have to go into each contour layer and change the navigation to a view that is just slightly wider than the widest camera viewpoint in your scene. And...make sure the map projection is equidistant cylindric for optimal contouring.

PM me with your name, email, and station call letters if you want these scenes...and I will drop 'em in your system sometime before T-giving.

Gobble--
50%

[forecastguy]

I'm sorry, but I'm an expert level user of both WSI and WXC... Whoever designed/programmed/approved of the WSI contour interface should be ashamed of themselves...

Hmmm.

What do I want to plot, a numeric grid aggregate, a query using a numeric grid, grid algebra... I already have a headache. :frustrated:

Maybe I want to contour wind speeds. Do I pick SA as my source. Or Metar/SA, or Special/Metar, or is it one of the 10 other selections there that look the same... My headache is getting bigger. :frustrated:

When someone showed me how to plot the RPM rain/ice/snow contour using three different layers of grid algebra, I contracted a full-blown migraine. :frustrated:

At worse yet, my dell 690s take 10-15 minutes to render a 48-hour futurecast in showfx... That's way long! :mad:

[Backscatter]

After spending all afternoon, I think we've got something.

We did indeed look at what SEB was creating on the WSM 48 hour and took the grid algebra from that. We had a few problems initially, such as having error messages pop up when trying to validate the expression. Well of course looking back the errors made sense because we didn't create our model data parameters first, A&B!!

Rendering takes about 10 minutes, which is a lot longer than it used to be (for good reason with the extra layers). I am tweaking the resolutions a bit seeing if it helps.

All in all... we have success.

I am getting the color palettes looking good too.

Thanks again everyone... and feel free to add any comments or suggestions if perhaps I'm missing something

[Mayhem]

Happy to hear you got something working!

[Backscatter]

Now I am noticing big gaps between the rain, mix and snow areas. It looks like pieces of a puzzle not quite fitting together. I also have data missing in the highest elevatations out west.

It looks quite blocky at times, but slightly better when actually rendering out.

Any ideas why the snow and mix areas, or rain and mix areas are not lining up next to each other???

[forecastguy]

Backscatter,

I noticed the same thing tonight.... got some screengrabs and shot them over to WSI... haven't heard anything back yet. :(

Is this what you were seeing?

http://i235.photobucket.com/albums/ee208/forecastguy/rpm3-1.jpg

[Mayhem]

Looks to me like the following... First of all, your grid size seems large, you may want to drop it from a 10 to a 3 or so. Secondly, you will want to change the map definition on the contour page to hone in even tighter.

Lastly, if your equations involve something like A-B > 5400 and the other is A-B<5400, you have a section that is =5400 that has nothing.

Try making the equations >=5400 and <=5400 and see if that fixes it...

[forecastguy]

Mayhem,

Thanks... But a little off base... This script usually works fine... In fact a WSI tech initially set it up...

The problem (as backscatter alluded to) is that today, for some reason the mixed precipitation stopped appearing, and there were sections of data missing.

[Backscatter]

I am trying to emulate the RPM snow/mix/rain found in an SEB, and setting it up on ShowFX.

Forecastguy:

Great snapshot! That is EXACLY what I am seeing, although I do have the mix colors showing up too. If it is a data problem, they picked the "wrong day" since I have been pulling my hair out (as another poster stated I would do) trying to figure it out. You know... Murphy's Law.

If you hear back from WSI, please let me know.

Mahem:

What do you mean by reducing 10 to 3? What am I doing wrong here...
1. I have map definition to equidistant cylindric in my viewing region with resolution set to a whopping 2000X1400 (for now).

2. I have another setting set to Every Frame at 3X.


Since I am brand new to this, I went into the SEB stuff to see the WSI snow/mix/rain script. I copied over everything, including the grid algebra, etc. I don't understand why it looks so darn good in SEB but when I build it in ShowFX, it looks terrible.

Our grid algebra is something like this (from memory, so may not be exact):

A=precip type
B=precip rate

rain~
if(A<2,B,NO_DATA)

mix~
if(A<5&>1,B,NO_DATA)

snow~
if(A>4,B,NO_DATA)

What are the numbers 1,2,3,4 and 5??? Are they parameters set within some RPM code???


For now, I am just wanting to use the SEB's version of snow/mix/rain, get it to work in ShowFX, then progress on our own to using geopotential heights, etc.

UHG!!!

[Mayhem]

I would imagine each number represents a type of precip. That being said, we don't have the RPM at my shop, so that's why I was forced to come up with the method that I use for our futurecast with the grid algebra and the different model data.

Have you been working with the equation that I sent, because I can help troubleshoot that? Another option would be for WSI to copy the scene that I have off of our box and send it to you (I have no problem with that, unless any of you are in an adjacent market, which I don't think either of you are.)

[Backscatter]

No... I haven't used your script yet, but I plan to check that out next week to see if it works. If not, there must be something wrong on our system.

[MattyD.]

And this is what I still can NOT understand. In 1995 or 1996 Kavouras - which became Meteorlogix - then taken over by WXC - sent over the pipeline grib files for the HQL. As part of the HQL data feed were the following: Rain Grib Files, Snow Grib Files, Freezing Rain Grib File and Ice Pellet (Sleet) Grib Files. You could assign different levels, ranges and color banks to each one of them. It might not have been perfect but it was and still is 10x better than anything offered out there today. And you didn't have to try to figure out this "whole" Grid Algebra world. The graphics were clearer and cleaner.

Somehow, Someone needs to explain to us (ALL) how a company 10 years ago figured it out, but the (quote -unquote) best vendor today still can not get it figured out!

Go back in the old Medialine forums, do a search for Grid Algebra and Snow on WSI...You will find posts that go back to 2004! The question really needs to be asked - When will Customer Service listen to the Mets. who operate these machines and tell their data operation people and programmers to get it done!

[Backscatter]

Okay... it has been several days and I think we have something that works.

Whoever said migraines will accompany this procedure from the onslaught was dead-on. Wow. It took some time to figure out, only because I see major flaws in the way RPM handles the data.

Note to WSI: Please prove I'm wrong.

Here's the deal:

According to WSI documentation, the "Precip Type" is determined according to assigned values from 1 to 5.

1=rain
2=?
3=freezing rain
4=mix/ice
5=snow

Using the grid algebra for rain and snow work fine. It's the "middle stuff, a.k.a MIX causing all the problems.

Technically, the mix expression should be

if(A>1 & A<5,B,NO_DATA).

However, after COUNTLESS HOURS looking into this, there are some MIX values assigned to 1, and some mix values assigned to 5!

How do I know? Because there are data gaps between snow & mix and mix & rain if the above expression is used.

Maybe there is no data there you ask? Not so. If I set mix>0, and toggle off rain and snow in the stack editor, all the precip shows up and looks like mix (just for testing of course).

My solution is to have this for mix:

if(A>0,B,NO_DATA)

**AND** have this top-down order in the stack editor: snow, rain, mix.

If your color tables are 'good', the mix color is buried nicely under snow and rain.

If anybody has a better approach, I'd sure like to know. My conclusions state that "MIX" is not defined solely by the numbers 2,3 and 4.


50%: PLEASE chime in too. I'd love to know your thoughts.

Before I made mix>0 in the expression, WSI tech. support said my scene was set up 100% the way it was supposed to. Again... I wonder if there is a flaw in the data somewhere.

[ejh4isu]

My solution is to have this for mix:
if(A>0,B,NO_DATA)
**AND** have this top-down order in the stack editor: snow, rain, mix.
If your color tables are 'good', the mix color is buried nicely under snow and rain.


I had mine set up exactly like WSI did too, just copied everything from the Producer to fx; but my mix precip wasn't working at all. When I subbed in your equation, it worked fine.

Just an FYI ... :)

[TAFKA wacowx]

So tell me 50% or other WSI-ers. How come on the Energycast WSI page I use at work, the newly added RPM model info shows some terrific rain/ice/snow/sleet 'futureradar' imagery, but it appears to be such an ordeal for the TV folks to display this on-air? I mean, the loop of this RPM imagery looks sweet and I'll be adding this to my short-range arsenal from now on, but from what I have read here, it seems to be such a difficult thing to display on-air? Am I not seeing the same thing?

[50 Percent Chance]

So tell me 50% or other WSI-ers. How come on the Energycast WSI page I use at work, the newly added RPM model info shows some terrific rain/ice/snow/sleet 'futureradar' imagery, but it appears to be such an ordeal for the TV folks to display this on-air? I mean, the loop of this RPM imagery looks sweet and I'll be adding this to my short-range arsenal from now on, but from what I have read here, it seems to be such a difficult thing to display on-air? Am I not seeing the same thing?

You're talking about two totally different pieces of software here, and to be cliche it's apples and oranges.

[MIguel W foX]

So tell me 50% or other WSI-ers. How come on the Energycast WSI page I use at work, the newly added RPM model info shows some terrific rain/ice/snow/sleet 'futureradar' imagery, but it appears to be such an ordeal for the TV folks to display this on-air? I mean, the loop of this RPM imagery looks sweet and I'll be adding this to my short-range arsenal from now on, but from what I have read here, it seems to be such a difficult thing to display on-air? Am I not seeing the same thing?

We are beta testing the 4km stuff and it looks awesome. Once you understand how to set it up it's really easy. It's like any other model though in missing some events. Some of the lack of knowledge on setting it up is WSI's fault but some is the user.

[forecastguy]

We are beta testing the 4km stuff and it looks awesome. Once you understand how to set it up it's really easy. It's like any other model though in missing some events. Some of the lack of knowledge on setting it up is WSI's fault but some is the user.

I'm sorry, but when setting up these damn scripts requires user "programming" and knowledge of data & variables that are not supported with ***ANY*** documentation (e.g. what the numbers mean for PTYPE), it is entirely WSI's fault.

[WAA]

Few people are more annoyed than me about how ridiculously overcomplicated all of this is, but all the RPM parameters (including PTYPE) are laid out in Tutorial 144.

All I ask for is a way to differentiate precip with grid algebra without the 'blockies' and holes in the data *or* the ability to interpolate the VORTEX precip data so it doesn't jump from hour to hour. Please?