EWP Blog

I decided to check out the OUN WRF and I was impressed it picked up the line of convection in Oklahoma very well. It was still a little overdone in the Panhandle but other than that good first impression. ~ Vollmar

While not necessarily in my area of interest today, my attention was drawn to PAH’s CWA at around 22Z when the CI product indicated a 70% probability of convection initiating. Admittedly, I didn’t place too much stock in it as there have been false probabilities showing up this afternoon (plus it is outside of my area of focus). However, within the next half hour, convection developed in the same area.

~Linda

With a significant increase in ProbSvr and good environmental conditions Linda issued a severe thunderstorm warning for a storm in northern TN. ProbSvr at this time was near 90% at this time.

Two volume scans later, while the reflectivity was still fairly significant the ProbSvr dropped to 48 percent. Two more volume scans later the reflectivity looked like the above image where the storm was obviously weak. This downward trend in ProbSvr was great since it was near the southern end of the severe thunderstorm warning and another warning decision was near.

-JB

When looking back at Simulated Satellite today and seeing how it performed, it actually did very well.  It was on point with the timing of initiation of convection and the location of the start of the storm as well as how they moved.  The only thing it really lacked was the strength of the storm.  I would use this product for forecasting purposes! ~Vollmar

…the prob. severe model and SRSOR as well as radar interrogation. The prob. severe model latched on to a storm near the intersection of Sumner, Trousdale, and Macon counties at 2212Z but I wasn’t able to view the actual prob. severe model product until after 2217Z, leading to the warning not being issued with even greater lead time, as I had hoped to be able to do. While waiting for the updated prob. severe model product to come in, I analyzed the super rapid scan imagery and noticed the growing cloud tops between approximately 2207Z through ~2223Z. After the warning was issued, I re-analyzed SRSOR and an OT was detected in the vicinity of the cluster of storms.

~Linda

Another shot at the OT algorithm.  The OTs are placed over the IR satellite in the upper left and are very much bias towards this imagery.

Looking to the right at visible satellite imagery, clearly only one OT is evident in Gloucester and Atlantic County where 4 are showing.

Thinking about how the forecaster IDs OTs on visibility, it appears to be a sharp/rough contrast of colors that draw the eye in… versus a smoothness of color downstream.

If there was a way to add this to the weighting of the algorithm it may help to discourage False Alarms.    On the other hand that could force it to be a daytime only product…. Unless you set in a series of changing variables based on time and angle of the sun.

Here we have a second OT Detect out over the Atlantic at the edge of the domain . This one does pretty good.  This got me thinking… This Algorithm may be best used for someone like the AWC which needs as much data as possible in a data sparce area such as the oceans.

If a greater than 50% chance of severe damage could be directly coorelated to OTs on land…. then they probably would be helpful in identifying Convective SIGMETs areas in a data sparse region over the ocean.

Just Ideas…

Grant H.

I know I wasn’t in this area forecasting today, but I wanted to see how the LAPS Reflectivity did today since I wasn’t as inclined to use it yesterday.  I initially tried to use it 2 hours ago, but it was not working.  However, I was told that it had moved areas so it needed time to play catch up and ingest the new data.  However, when I loaded it now at 22Z the model run that loaded was the 18Z with a snapshot even from 17Z.  Not sure what is going on there, but now it would not be a tool for forecasting but just for verifying because it has loaded up all the way.  As you look it does pick up on the cu-field, but missed the larger storms up in the N&W.  I am not sure what is going on with it unless since I am in DC I can’t use it?  ~Vollmar

Just wanted to stop and say this GOES true colors imagery is awesome.  Interested to see where the data is coming from that makes these false color composite. If indeed the background came out of GOES rather than polar orbiters which i suspect, it could probably be used to generate greenness maps for the fire weather community which is another partner.

Grant H.

Attempted to load CONUS LAPs for the second time today. CONUS data is now available. YAY!

While playing with wind fields, further attempted to load wind fields into streamlines and then arrows on the CONUS domain. Neither was possible while they both appeared on the load bar below. The in progress bar in the lower right “Contouring” continue to move but nothing has appeared on screen other than the load bar.

To check whether this was a size of data problem switched the domain size from CONUS to WFO level and clip the editing area/processing time. there was still no load of streamlines.  there may be something behind the scenes in the code that hasn’t been added.

Grant H.

Convection continues to develop and move southeast out of KY into TN. The strongest storms are moving southeast within the greatest convective instability indicated on the NearCast model. Although the model has holes across the OH and TN Valleys, extrapolation of the southern gradient of the instability axis indicates that the storms on the western side of the line may be in a less favorable environment for strong updrafts. Radar supports that storms are weaker in this area as well.

-JB