GR Earth previews (Gibson Ridge)

  • Thread starter Thread starter Jason Foster
  • Start date Start date

Jason Foster

This is taken from Beau Dodson's post over at EasternUSwx.com

GR Earth will be released soon.

Here are some videos of the software in action - Mike Gibson has done an outstanding job (several years of work) in putting this program together. It is similar to GEMPAK - with its interactive zoom and unzoom (one of the reasons I like GEMPAK). The software will have some analysis features for severe weather that Mike has been working on for some time - along with his winter precip type algorithm (which is already sought after by some NWS offices - other).

He has added a lot of features over the last few months - the list is virtually endless. Expected release date - a few more days.

Here are some screen capture videos of high risk days from this past spring (the pull out images of the Planet Earth with little dots - those are METAR stations)

Videos - posted with Mike's permission

Go here for the links:
http://www.easternuswx.com/bb/index.php?/topic/236100-gr-earth/page__pid__5047799#entry5047799

Screenshots:
http://www.stormreports.org/grearth.htm.



EXCITED????



img9.jpg
 
Last edited by a moderator:
Correction: The link to the screenshots is http://www.stormreports.org/grearth.htm

On the GR Owners forum thread about this, Mike has stated numerous times that this is NOT intended to be a replacement for GR2/GR3. The scope of GR3 is much larger scale -- mesoscale and synoptic-scale, though it does have some radar capabilities (apparently -- I'm not a beta-tester).

Per that same owner's forum, Mike said this will be <$20/mo (yes, it's a monthly charge, and I think I read that it'll be charged annually).
 
I'm not a beta-tester either, but a recent student at ISU here and a member on this forum (at least used to be), Scott Lincoln, is. He has shown me a lot of the GRE test and in my honest opinion, it's not worth it. There are only two things I think GRE has that is better than anything GR2 or GR3 can offer:

1) Viewing radar data from multiple sites simultaneously (mosaics)
2) Looping satellite composited with reflectivity (or other products)

In my opinion, if you want to see analyses of temperature, pressure, wind etc., you can find almost as nice of imagery at other sources such as the rap site, and those are free.

I'll be saving my money for the dual-pol products upgrade in GR2AE when it comes around.
 
Now knowing that GRE isn't intended to be a replacement, I suspect as soon as GRE is rolled out, the rest of the software packages will likely get attention for possible enhancements and upgrades, and that would likely be what we (chasers) are likely to want. Perhaps what was learn by the design team at Gibson on GRE will be very helpful in the next edition of the GRlevelX software packages.
 
One GRE benefit is that you can overlay the RUC analysis output over radar. So you can see if the storm is moving into an environment with better TORH or look for returns along a CAPE gradient.

However i don't think chasers are the target market. More like EMs that don't need to see tilt 4 spectrum width, but do want an easy to use replacement for DTN or Threatnet. And MUCH cheaper that those...
 
One GRE benefit is that you can overlay the RUC analysis output over radar. So you can see if the storm is moving into an environment with better TORH or look for returns along a CAPE gradient.

However i don't think chasers are the target market. More like EMs that don't need to see tilt 4 spectrum width, but do want an easy to use replacement for DTN or Threatnet. And MUCH cheaper that those...

Would it be useful for a newbie? :eek:
 
Absolutely. And non-newbies too. I have it running 24/7 over mt area then use it for a wider view of details. You can also right click over an area and open up gr3/2 to the closest site for the hires look.
 
Jeff Snyder and Rob Dale have it right ... but I'd add that GREarth doesn't really have a target audience. It's more of a program I wrote to help me, a non-meteorologist, understand why the weather is behaving the way it is -- i.e. the big picture. One incident brought this home to me and was one of the main motivators for GREarth: several years ago I was watching a nice storm in western OK in GR2Analyst and posted on the grlevelx.com forums that it looked like a tornado would form. Rdale replied that nothing would happen because the storm was "elevated". Sure enough, nothing happened. It was obvious that no matter how sophisticated your radar app was (xsections, volumetric radar, etc., etc.), it couldn't give the entire picture.

A second motivation for the RUC severe weather fields was Greg Stumpf's presentations on PHI. I wanted to see if the RUC could indicate where there was the potential for severe storms and tornadoes in the short term (+3hrs). This led to combining RUC forecast reflectivity with forecast SCP and STP. The result is somewhat of an automated, graphical MCD. I'm pretty happy with the results even though it overforecasts the extent of the threat.

For hurricane watching, the satellite imagery was a must. GREarth covers all the critical sectors with VIS, IR, and WV. Given the spatial extent of the sat coverage, a spherical Earth projection was needed, hence the "Earth" part of the program.

Also, I wanted a program that would help with monitoring Winter weather down South. We're always in marginal conditions here. That led to the inclusion of precip-typed radar, LSRs, and METAR current conditions. The precip typing is derived from the RUC using the Canadian Energy algorithm and is combined with the actual reflectivity in a novel way at render time. A method that gives users an idea of the uncertainty in the typing.

It's important to note that GRS isn't really a "business" in the usual sense. It's one guy writing software. I don't do market research, target specific segments, etc. And, as with all the GRS programs, it'll be the users that mostly define where GREarth goes from here.

Mike

PS. I noticed recently that GREarth does not, in fact, have a mesoanalysis field that indicates whether storms are in an elevated environment! I'm working to remedy that situation now but it probably won't make it into 1.00. There's apparently no official definition/algorithm used to determine whether a sounding shows "elevated" conditions or not. After discussions with some mets, I now have a algorithm that should work as a first cut.
 
Last edited by a moderator:
PS. I noticed recently that GREarth does not, in fact, have a mesoanalysis field that indicates whether storms are in an elevated environment! I'm working to remedy that situation now but it probably won't make it into 1.00. There's apparently no official definition/algorithm used to determine whether a sounding shows "elevated" conditions or not. After discussions with some mets, I now have a algorithm that should work as a first cut.

If you can make GRE able to ingest RUC analysis data at all levels so that it can see an entire atmospheric sounding for each point of analysis, you could look at the level of maximum theta-e. If that isn't close to the ground then storms would almost certainly be elevated.
 
If you can make GRE able to ingest RUC analysis data at all levels so that it can see an entire atmospheric sounding for each point of analysis, you could look at the level of maximum theta-e. If that isn't close to the ground then storms would almost certainly be elevated.
All the RUC processing is done on a server, so the server program would need to generate the "elevated" parameter. GREarth itself just downloads the user selected data fields on-demand. That way people don't have to download multiple 20+MB RUC gribs every hour, only relatively small (<100KB) slices of information.

Anyway, a first attempt algorithm for this has been created and will undergo testing in the near future.

Mike
 
Looks great! Are we getting some of the same features (like built in Spotter placefiles) in GR3?

I was working on a program similar to this for my own use when armchair chasing, but the feature set looks nearly identical so I'll have to try this out. I'm assuming the subscription price model is partly due to the RUC mesoanalysis upkeep/feed? That's the only real downside to this for me, but I can understand why it might be necessary.

If you can make GRE able to ingest RUC analysis data at all levels so that it can see an entire atmospheric sounding for each point of analysis, you could look at the level of maximum theta-e. If that isn't close to the ground then storms would almost certainly be elevated.

Feel free to reply in PM if this is derailing too much, but would LCL be an easier (though more simplistic) mesoanalysis parameter to look at? I'm admittedly not well-versed with determining storm elevation by looking at theta-e.
 
For most purposes using the LCL would be a good index parameter as to whether or storm may or may not be elevated. Better yet, a mixed layer LCL parameter would likely be even better to have viewable on GREarth. There isn't going to be a parameter that gets it right all the time, but viewing LCL values will sure do the trick a majority of the time.
 
I'm not sure LCL would be a very effective discriminator of elevated vs. surface-based convection. It's possible to have surface-based convection even with a relatively dry boundary layer (and resultant high LCL's). Perhaps you meant "high-based" rather than "elevated."

When I was first learning about elevated convection, I was told to look at the height for which the maximum CAPE is attained for a lifted parcel. On the SPC Mesoanalysis, this can be seen on the MUCAPE plot as the "lifted parcel height" contours. If the MUCAPE is attained using a parcel height well above ground level (e.g., > 500 m AGL), then convection in that environment is likely elevated in nature. SBCINH is also a good quantity for making a quick judgment: if it's significantly negative, then it stands to reason that convection is likely rooted above a capped layer near the surface.

Of course, the accuracy of the values we compute for these quantities is a whole different matter, thanks to our sparse observational network. That's probably why you'll occasionally see a tornado-producing storm where the RUC analysis indicates SBCINH to be prohibitively negative (like the AMA area storm on April 20 this year), or more often, a primed dryline environment with decent convergence where the model analysis indicates minimal SBCINH and nothing happens.
 
I'm not sure LCL would be a very effective discriminator of elevated vs. surface-based convection. It's possible to have surface-based convection even with a relatively dry boundary layer (and resultant high LCL's). Perhaps you meant "high-based" rather than "elevated."


I do agree. By no means is it a perfect indicator, as the opposite can also be true (you could have elevated convection despite low LCL's). The Lifted Parcel Height that you mention can be a good indicator, however it also has its' flaws. Like many things when using model analysis there are going to be errors and exceptions...

It will be interesting to see what Mike has developed for an algorithm to determine an 'elevated' vs surface-based.
 
Back
Top