First Experimental SVR for OHX

I issued the first OHX experimental severe thunderstorm warning for the day, based primarily on traditional radar analysis and utilizing the prob. severe model. When two cells merged at the Monroe/Barren/Allen county borders in LMK’s CWA, the prob. severe model indicated an 81% chance of severe and then one volume scan later, the probability jumped up to 97%.

While I didn’t use these products as the basis for my warning, I went back to look at the SRSOR and OTD products. They clearly show the cell merger and significant updraft growth over the course of the last hour.

SRSOR_OTD_LOOP.GIF

~Linda

**Update: Trees were reported down in Clay County, TN.

*** Just a note, the OTD’s appear shifted to the west (and only appear every several minutes) because these detections are from GOES-East, while the satellite imagery in this animation is 1-minute GOES-14. – BL

Lightning Jump Really Helps!

I once again was watching a cluster of storms and the lightning jump turned to red.  I immediately started to monitor that storm and started to see flash extent density increased.  Then Prob Severe started to jump and I actually issued it before it turned hot pink.  I will have to see if this verifies but I think I got a jump by at least 15 minutes or so about issuing a warning! ~ Vollmar

Update it did produce wind damage! WICKED AWESOME!

Using GOES-R product together

It has been beneficial to view multiple GOES-R demonstration products in one AWIPS-II display during the Spring Experiment. This trend was especially evident today as  we were able to use the lightning products in the Sterling CWA. Below is a gif animation of a 4 panel display showing how these products can be used together (sorry about the change in domain in middle of loop).

In the upper left panel, the NearCast (observation-based) analysis theta-e difference (instability) field showed copious amounts of instability in the region. This was due to a plume of high theta-e air below much drier air aloft (see other posts for details). Meanwhile, by 1815 UTC the GOES-R Convective Initiation product (upper-right panel) gave a cloud object an 81% probability of initiating. Given the relative high CI probability in an unstable environment, the forecasters keyed in on this area. At 1830 UTC, it was obvious that this object was indeed growing upscale. At this point, the ProbSevere (lower left panel) product started tracking this cell.

At 1852 UTC, the ProbSevere product gave it's first probability greater than 70%, indicating a relatively high likelihood that this particular radar object would produce severe weather. The probability would actually dip down, before surging up again. This was likely due to an influence of a nearby storm. Shortly after 1900 UTC, total lightning started being detected, seen in the lower right panel as PGLM flash extent density. A 2-sigma lightning jump occurred with this cell at 1959 UTC, indicating an uptick in storm strength. This particular cell went on to produce hail shortly thereafter.

Aside from this 4-panel type of display, forecasters have been finding other unique ways to blend these various datasets together in AWIPS-II (overlays, various 4-panels, etc)

-Bill Line, SPC/HWT Satellite Liaison

NearCast Verifies with Radar

I already blogged about NearCast earlier today, but I wanted to show how it actually verified.  Storms initiated on the boundary and as they moved into the unstable air, they became stronger as the followed the unstable air.  Also, the storms that formed in the lighter blue actually began to weaken as more stable air moved in.  The storms moved like this model showed from NW to SE.  I would use this for storm motion for sure! ~Vollmar

Tracking Tool, Lightning Jumps and Storm Evolution

Above are two examples using the Tracking Meteogram tool and Flash Initiation Density to build flash rates of two individual storms near Washington D.C.    Rapid increases in total lightning signalled the onset of storm intensification in both cases, and multiple lightning jumps were observed as each storm developed and moved across the DC and Baltimore Metro areas. Hail to the size of golfballs was observed in the top storm with the first 1 inch hail reports falling at 1939 UTC.  At the time of this post, the southern storm produced hail to the size of quarters at 2035 UTC.

Chris S.

Storms across the Potomac

Maryland isn't the only state in the mid-Atlantic with potent storms. In Virginia, more storms have had strong ProbSevere signals (with strong satellite growth rates and moderate to large MESH). Severe hail was later recorded northwest of Richmond.

Closer to the Potomac, a storm with moderate to strong satellite growth (~30 minutes prior) increased in probability of severe from 37% to 65% to 85% over the course of six minutes (from 20:26 to 20:32 UTC). This explosively growing storm was warned by the NWS at 20:30 UTC. One inch hail was recorded at 20:36 UTC.

-Cintineo/Sieglaff

An OTD Appears!

At 2030Z, an OTD appears in PAH’s southwestern portion of the CWA. Unfortunately, a jump in the data occurs between 2011Z and 2030Z in the SRSOR imagery. Around the same time as the detection, some hail reports were disseminated from the PAH WFO but nothing over severe criteria.

Some times you just have to sit back and laugh.

So some of this may have been AWIPs 2 some of this may have been alternate issues.

Chris was behind me as we were researching the LJDA components to things. When we noticed the one minute CG data disappeared…. then the 5 minute CG data disappeared. However the LJDA maps/shape files continued to work.  Whalla! a 5 and 6 sigma jump occurred successively over 2 minutes. It was time to warn.

So, I went over to radar base products on another CAVE windows to issue the warning over the base all tilts reflectivity.  Loaded Warngen to find that the old radar display controls from yesterday were still controlling the flow of the storms today 240 when I wanted roughly 320. I immediately attempted to to change the flow and redraw the box. Except now the box refused ro redraw in the southwest direction and immedately insisted on the northwest direction…. fine I’ll attempt to track the ball backward and force a redraw of the box in that direction…. Done. Who Hoo… that wasn’t  hard.

then i went to redraw the verticies….. after clicking on the first vertex the whole screen proceeds to move on me. … “Maybe I missed it” goes through my mind…. and click again…. “nope missed again”… click again…. “okay this is not cool”.  So, I permanently right click down to find a window to edit the vertices. “YAY alternative.”  Next I discover that the one vertex I’m moving is now permanently attached to the mouse icon….. “GAAAHAHHAHAHAHA!!!!!!!”

As i step back and look at it, someone out there of importance could be dead right now. But in this case I’d like to think it was Barney… Because personally I want that fat purple dinosaur to die. Thankfully this is an experiment and i’m happy nobody died… except for that awful singing dinosaur.

Grant H.

When I decided to issue a warning

As the storms to began to break out, I used the prob severe tool in combination with the total lightning initiation and total lightning extent and lightning jump and based on how they all came together, I felt comfortable issuing the warning.  About 15 minutes later, golf ball size hail was reported so with this particular t’storm these products did well together. ~Vollmar

Hagerstown, MD

A storm that developed in southcentral PA had a ProbSevere value of 72%, with moderate satellite growth and nearly 1" of MESH at 18:36 UTC. The probability of severe fluctuated to 45-50%, and then came back up to around 80%. An NWS warning was issued on this storm at 18:57 UTC. Severe hail and wind reports were recorded at 19:02 UTC, in Hagerstown, MD.

When speaking with a forecaster, she noted that she liked to see some sustained signal (several scans of high ProbSevere) before issuing a warning. This underscores that perhaps a forecaster shouldn't necessarily issue a warning as soon as there is a high probability, but that they should perhaps take notice and interrogate the storm further.

-Cintineo