Using the HRRR the day of chases

[rdale]

Unfortunately that is only for a winter experiment, so don't expect it to run through the spring severe weather season.

We don't get severe weather in the spring in the Great Lakes anymore so we might be okay

 

[Warren Faidley]

I used the HRRR for forecasting dust storms in Arizona a few times last year with some success. The initiation time, placement of cells and movement seemed to be quite helpful although I did not take notes. I plan to experiment (pay attention) a little more this year.

 

[Dan Robinson]

Like Jason said, sometimes the HRRR is spot on, but sometimes it's way off. You never know what day it's going to blow it, so if you rely on it 100% it's going to burn you occasionally. On the Dodge City 2016 day, the HRRR had a long, monster helicity track in NE Colorado with a few lighter ones in Kansas. That storm in Colorado did produce, but it wasn't the most prolific one of the day.

Sometimes an indication the HRRR (or any model for that matter) is not accurate is if it's initial conditions don't match what is currently happening (radar and surface obs).

For me the HRRR is like the Day 1 outlook - good for confirming a forecast or at least partially deriving one, but not reliable enough yet to use as a primary source. I think one day the short term models might get to that point, but we're not there yet.

 

[James Wilson]

On the Dodge City 2016 day, the HRRR had a long, monster helicity track in NE Colorado with a few lighter ones in Kansas. That storm in Colorado did produce, but it wasn't the most prolific one of the day.

Yeah the HRRR doesn't see that amazing OFB along the DL that created all the magic.

 

[Chris Dickerson]

I’ll use the HRRR and other models the morning of a chase and sometimes up until initiation for guidance. But on chase days, I’ve isually narrowed down my target to a certain degree and focus more on surface observations for fine tuning my chase.

 

[StephenHenry]

Hey @Jeff Duda, could you give us a quick summary (or link to some materials) that outline the differences between the Operational HRRR, HRRRX, and HRRRdev3?

Also I'd love to see your SLS poster on the HRRRv4 changes coming next year!

 

[JamesCaruso]

I use the HRRR for validation of my forecast, or to highlight something I might have missed. I look at it last, after making my own forecast. Similar to SPC, as someone else said above. I try to treat it as just one data point without being unduly tempted by it. When out in the field, I try not to look at it, because it’s very hard to resist when you feel confident about your target area yet you see the HRRR blow up a supercell somewhere else. You can’t chase after these ghostly echoes of the future. I remember last year in Nebraska, we were near Broken Bow along with many other chasers on a day that was supposed to be “big”; the HRRR kept showing a couple supercells further south near I-80, and with nothing happening near us we were lured down that way, only to find ourselves under a stronger cap. I find it easier to ignore the lack of a HRRR signal in sticking with a target (assuming there is only one area of interest and no signal anywhere else either), but harder to ignore a HRRR signal outside of my target. Especially as a non-meteorologist, it’s hard to ignore the underlying physics of the model and think that I know better.

I like to use the HRRR model at home in the spring and summer for an idea of the placement and timing of local storms so that I can figure out whether I have time to get a BBQ in. [emoji57]

If the HRRR ever gets relabel enough to show the exact time, place and track of a supercell or tornado, storm chasing would lose a lot of its appeal to me. But that’s a subject for another thread.

 

[Jeff Duda]

Hey @Jeff Duda, could you give us a quick summary (or link to some materials) that outline the differences between the Operational HRRR, HRRRX, and HRRRdev3?

Also I'd love to see your SLS poster on the HRRRv4 changes coming next year!

No, honestly I can't because the configurations of HRRRX and HRRRdev3 change frequently (as often as every few weeks).

The major forecast system changes we are experimenting with include:

• the use of 12-hour partial cycles (compared to the 1-hour spin-up "pre-forecast" currently being used in the operational HRRR)
• More advanced radar data assimilation techniques (direct reflectivity assimilation vs. the operational cloud analysis)
• Use of filtering in the forecast process (currently none is used)
• Using storm-scale ensemble information in the data assimilation process (currently GDAS information are used)

 

[Ethan Lang]

Now that I have this thread going what do you guys think currently of the HRRR as far as forecasting winter weather, particularly snow fall totals?

 

[Dan Robinson]

It's about the same. The HRRR has seemed to be a few degrees too warm on several of our events here, IE, placement of the surface freezing line. The 4K NAM has done better, in my experience.

 

[Ethan Lang]

@Dan Robinson I picked up on this pretty quickly this year so I have been using the NAM 4km more exclusively for winter weather. Another thing I noticed was that the NAM has picked up on dry air inversions more than the HRRR with many of our snow events. In one case back in December if you would have looked at the HRRR 15 hours out you would have thought blizzard conditions with 12+ inches of snow. NAM picked up on dry air inverting and while we got blizzard-like winds, we got less than an inch of snow.

 

[Paul Knightley]

Like others, I like to look at the CAMs and look for consistency in their output - of course, on a marginal severe weather day, this will probably be hard to find as subsequent updates take *some* account of the latest trends. However, a good hand-analysis of the early day situation is the first thing I do every morning when out chasing. I enjoy the process, with a cup of coffee, to get my mind thinking about the state of the atmosphere.

I like to use the CAMs as something of a guide as to weather convection will form, more than where the helicity swaths are. Of course, they can give you a bit of confidence (sometimes mis-placed, of course!)...I recall May 27th last year - CAMs showing some decent enough helicity swathes in S parts of WY - my hand-analysis showed a slow-moving front in the area, which was enough to get us up there - and so we witnessed the several tornadoes which occurred NW and N of Cheyenne.

 

[Clarence Bennett]

I don't look at nor make any decision based off the HRRR or any CAM for that matter. I stick to the meteorology and mesoanalysis the day of.

 

[Ethan Lang]

@Clarence Bennett you are the first one in this thread that had said you don't use the HRRR the day of. Could you explain?