The final week of the 2016 GOES-R/JPSS Spring Experiment concluded with our two pairs operating in the Nashville and Huntsville CWAs. Both were able to evaluate the PGLM product via the Huntsville LMA.
LAP
- Convection developed along the moisture and instability gradients in LAP.
- I liked seeing the model data where retrievals were unavailable. In addition to having a continuous field, it often allowed for quick comparisons of retrievals with nearby GFS.
- Our office does look at K-Index for flash flood situations.
- 30-min is a good temporal update frequency. Too frequent of updates would not be that useful, as such fields do not change so rapidly.
- Layer PW was my favorite LAP product as it was most unique, and added value to my analysis. It was particularly useful on days when we had strong low-level moisture advection, tracking the movement of moisture, and dry air aloft.
GOES-R CI
- When I had 1-min imagery, I did not need CI because I could identify areas of imminent CI in the imagery.
- In situations where you are expecting severe thunderstorm activity, you's look more at severe CI. Regular CI was not as useful for severe situations because you could see cb development in the 1-min data.
- When looking for general thunderstorms, I see CI being more helpful, including in the cool season. This would be valuable for DSS purposes.
- I found utility in having both CI products up. If severe CI was pinging on something in addition to regular CI, it helped to focus attention to particular areas of interest.
- It would be helpful to see probability trends for a particular cloud/area.
- We were fine with the display concept
- I like the current instantaneous visualization over a smoothed probability field approach.
ProbSevere
- It would be nice to see a meteogram with a history of ProbSevere probs.
- Everyone is fine with the display and color-scale.
- Similar to VIL of the day, might be helpful to determine "ProbSevere Prob" of the day.
- I think it really well with discrete cells, but later would merge nearby cells.
- I would say this was my favorite product outside of the 10min imagery.
- I thought it performed great this week.
- We would all use this in operations.
- I've worked 5 or 6 severe events in the last month, and I've ProbSevere up for all of them. Usually I have storm relative velocity all tilts, regular velocity in the middle, and the third screen has different fields, including composite reflectivity with ProbSevere. I've also even started putting it on all-tilts. The display does not distract me. In my office, the threshold to warn depends on the day, but I've found with most of our events, especially with severe wind, we can get severe with a threshold of ~60%. Definitely not using it as a yes/no.
SRSOR
- All forecasters loved using it this week!
- 5-min is certainly bettern than 15. But when you are tracking low-end severe situations, subtle boundaries can make all the difference between something going up or not. We get better than 5-min radar data, but 1-min satellite data can fill gaps that we still have. 5-min will be useufl, but 1-min is optimal.
- I think it is certainly time to make the jump to 1-min satellite imagery. There is so much that can be seen, even outside of convection. Forecasters need to use satellite data more in day to day operations. Generally, I think forecasters don't think satellite imagery is as useful as it is, and they have a hard time understanding exactly how much they will see in the 1-min imagery.
- It was helpful to view long loops of the 1-min imagery on the regional scale to get a big picture idea of how the system was evolving.
- It was really helpful for analyzing frontal structure and all the different boundaries.
- Satellite imagery is truly the only visual representation you have of a storm that you can't get with any other product.
- I found it useful to match 1-min lightning data with 1-min satellite data.
SRSOR Winds
- I liked the winds a lot. You could see the vertical structure of a front, and how winds changed with height from the surface. Seeing rapid change over a short vertical distance was intriguing. AMV's could be a big help with our TAFs.
- I felt that the low-level winds were more useful than the upper-level winds. They indicated areas of low-level convergence, moisture transport, veering of winds from the surface, potential for tornadoes.
Lightning Jump
- I liked it more as the week went on. I usually used it in tandem with ProbSevere and PGLM Flash Extent Density. I could see all of these being in a 4-panel and helping with situational awareness for severe operations. Especially on Thursday, I noticed the storms with the biggest LJ's were the ones that strengthened considerably thereafter.
- I'll be interested to use this during cool season events, as I am always looking for more information in these situations.
- I like the way it is now, though I can see others preferring a contoured look.
- I like a 4-panel layout with ProbSevere, lightning jump, Severe CI, Lightning, composite reflectivity, and satellite imagery.
- Forecasters are/will always change to their preferred color tables.
- There will always be a spot for a product like Lightning Jump in my display.
GLM Total Lightning
- The lightning data will be very helpful for DSS - events, fairs, etc. It will be very helpful to have this information updating every 1 minute.
- Especially for cool season events, we are always looking for more data. Lightning from satellite will be helpful.
- I can see this being helpful in EM's decisions to evacuate stadiums.
- This will be big for us during fire weather season in the NW US.
- In the future, with lightning in field offices, there must be very good training on all of this. There is/will be a lot of different lightning data. Generally, forecasters do not know the differences in lightning verbage.
- I will likely overlay it on radar or satellite.
- LMA-1 was the favorite among the group
NUCAPS
- The plan view and cross-section components were my favorite aspect of NUCAPS this week
- The lure is that it is an observation. I think it should remain observationally driven, even though we know there could be a source of error. If so, we know the source of the error. If you add in model data, you don't always know the source of the error.
- Pop-up skew-T will be good to use before and during an event with NUCAPS.
- Modification is not an issue for me. In our office we modify RAP soundings all the time. It takes some time, but it works.
- NUCAPS has a lot of potential, but a lot of bust potential for captivating an office.
- I can't get anyone to look at it in my office in Portland.
- The lack of detail is a killer. That's why I think plan view and cross section displays are more valuable.
- People will use it if they see the value, and it is made clear that this is an observation.
General
- Participants felt that the start of week orientation/familiarization was great.
- It was the perfect amount of products to evaluate.
- It would be nice to have a DRT WES case for slow days.
- I suggest having a group briefing after the groups complete their mesoscale analysis but before CI.
- The broadcaster commented that this was a great experience, and it was wonderful to be able to work directly with NWS forecasters.
- Some of the training material should be put on the CLC so we can go back and look at it in the future.
Showing posts with label SRSORwinds. Show all posts
Showing posts with label SRSORwinds. Show all posts
Friday, May 13, 2016
Thursday, May 12, 2016
Final Thoughts 5/12/16 HUN
Final thoughts on the day. Today was a huge success for the PGLM. Looking at flash densities for storms in the Huntsville area was highly beneficial. I liked that the output was not tied to the cells and also I preferred the smoothing on the product. As I was not really interrogating it like I would radar data, smoothing was preferred and I could see easily exporting this to an emergency manager or social media post.
Atmospheric vectors did a great job of showing the structure of the pre-frontal trough in the vertical.
1 minute data was vital in tracking the speed of the prefrontal trough, outflow boundaries and the front itself as the afternoon evolved. As always, tracking overshooting tops was beneficial to the warning process and also to monitor storms as they were decaying as well.
NUCAPS did well with the soundings and was also impressed by the theta-e cross section presented across the frontal boundary. While some observational data still needed to be modified, I could see this data being highly useful in the office.
ProbSevere was understandably lower today in this environment but qualitatively the worst cells had the highest probsevere ratings. As a forecaster, this was still a useful product to have today.
CI didn't have a chance to present itself today as convection had already initiated before the session began. - Jason Bourne
Atmospheric vectors did a great job of showing the structure of the pre-frontal trough in the vertical.
1 minute data was vital in tracking the speed of the prefrontal trough, outflow boundaries and the front itself as the afternoon evolved. As always, tracking overshooting tops was beneficial to the warning process and also to monitor storms as they were decaying as well.
NUCAPS did well with the soundings and was also impressed by the theta-e cross section presented across the frontal boundary. While some observational data still needed to be modified, I could see this data being highly useful in the office.
ProbSevere was understandably lower today in this environment but qualitatively the worst cells had the highest probsevere ratings. As a forecaster, this was still a useful product to have today.
CI didn't have a chance to present itself today as convection had already initiated before the session began. - Jason Bourne
AMV depict pre-frontal trough
Atmospheric Motion Vectors are nicely illustrating the pre-frontal trough across Arkansas. Of particular interest is the illustration of the trough with height. For the low level winds (blue) the pre-frontal trough can be seen clearly...particularly in northeast Texas where southwesterly winds are occuring ahead of the trough and northerly winds behind the trough. Looking at upper level winds however (green, red, pink), most remain in a southwesterly direction, indicating that the trough aloft remains to the northwest. This makes sense because the frontal boundary will be tilted with height back towards the northwest. The screenshot is below. - Jason Bourne
Wednesday, May 11, 2016
Wrap up 5/11 PUB, SGF
A bit of a double bust today as convection did not form in Pueblo or Springfield, MO areas today. Also a lot of data issues meant that lots of products were not available for extended amounts of time today. That being said the CI product redeemed itself today and it was the winner of the day as it performed very well in both CWAs and behaved as expected. Also atmospheric motion vectors performed well over the Springfield CWA and showed vertical directional shear just south of the CWA that never made it into the CWA but helped to illustrate where the better environment was. Once data came back in later in the day, we were able to compare LAP CAPE to rap analysis over the SGF cwa and it performed very well and matched up with our thinking very well. With a lack of storms in the area...special attention was paid to a wider angle view of the 1-minute imagery and the various features that it could see. An earlier post highlights the details but many features were discovered that are not usually apparent with the 15 minute imagery. Finally IASI soundings were very helpful in filling data gaps in the data sparse Pueblo area. The soundings seemed to perform very well in the higher terrain and did not need much modification in the lower levels. Not sure if this was because of the higher elevation or not. Will post that imagery below once again along with some of the highlight shots of the day. -Jason Bourne
Wind profile
We had to move to San Angelo, but now we've got a massive supercell
Vertical Shear in SGF Atmospheric Motion Vectors
Looking at Atmospheric Motion Vectors has shown a great illustration of vertical shear just south of the SGF forecast area in Little Rock's forecast area. This shear is better than that depicted in the SGF area but it is possible that some of this better shear could dritf northeastward into the SGF area and provide a boost for convection. One thing this does not show is much convergence in the lowest levels that had shown up on surface observations. By sampling the wind barbs, we could see that the lowest wind barbs were around 900 mb and shows what a difference even 100 mb can make...even in the low levels. -Jason Bourne
Tuesday, May 10, 2016
Daily Summary: Week 4, Day 2 (10 May 2016)
For the second day of week 4, our two groups operated in the Fort Worth/Dallas, TX and San Angelo, TX CWA's. Severe convection developed along the dryline across central Texas.
LAP
- CAPE values were slightly less than those from RAP, but I understand this because GFS was fairly low, which is used as the first guess in the retrievals. Although, RAP may have been high.
- The LAP CAPE gradient was right through our CWA. It was helpful to focus on the location and movement of this gradient, as convection developed right along it.
- Layer PW was helpful for identifying drying aloft. I thought with wind and hail, and TPW values being relatively low, this seemed correct.
- LAP LPW showed really dry air aloft. This was apparent in the RAOBs, and confirmed in LPW.
GOES-R CI
- It was not good yesterday in FWD. We didn't see any probabilities from severe or regular CI despite having cu develop and initiate into severe storms. There did not appear to be cirrus contamination.
- There were instances when severe CI was higher than regular CI
- In San Angelo, things were slow to get going. Early on, there were no probabilities until convection had already began to develop. Later in the day it was better. It keyed in on new development further south when we were focused on the ongoing convection to the north.
ProbSevere
- For the first storm of the day that got a warning yesterday, I was very confident that it was going to go severe. It looked good on radar with a strong reflectivity core. I put a warning on it despite probsevere being only 40-50%. After the warning, the storm weakened pretty quickly. Lightning and shear had both been low.
- Yesterday, if probabilities went over 80, I thought it had a good chance of going severe.
- I went liberal on one warning. Without ProbSevere I would have waited to warn. It definitely added lead time in that situation.
- Through using ProbSevere, I could see warning lead time going up, but also FAR rate increasing.
- Towards the end of the day yesterday, probsevere was in the 90s with a storm, but based on radar reflectivities, it appeared that the storm was weakening. Lightning flash rates were high at this time.
Lightning Jump
-Training needs to be improved for this, as I did not understand the LJ before using it yesterday
- I noticed a lot of fluctuations in LJ in a cell.
- It helped with my confidence, seeing jumps with the rapidly developing storms
- I liked using the 4-panel with ProbSevere and LJ together. The first time I used the two together, ProbSevere gave me more confidence in what I was seeing in radar, and LJ gave me confidence in what I was seeing in Probsevere. I liked this.
- I didn't feel like I needed to see a new lj to warn, but it added confidence.
- For me, it was more a a SA tool, alerting us to cells that we hadn't really been watching yet but needed to start keeping an eye on. It was an early indication that we need to look at the storm further.
- I actually liked the display, the blob really jumps out. We aren't really interrogating radar in that panel anyway, so it is fine that it gets covered up.
- I could see LJ being a confidence booster in pulse thunderstorm cases. Maybe not as much with a qlcs event dropping wind damage, when we know it is severe.
SRSOR
- With one storm I was watching, a storm split was obvious in the 1-min imagery much earlier than it was in the radar imagery.
- I could see the strengthening of the updraft and OT well in multiple supercells quicker than was apparent in radar.
- I enjoyed looking at IR imagery. The OTs as indicated by areas of very cold temperature were pretty brief, but the 1-min imagery was able to capture them. They weren't long-lived OTs, occuring in-between radar and routine satellite scans. So it was helpful to have this rapidly updating information.
- I could see the Overshooting Tops shoot up before strengthening in radar.
- With a cell that ended up being severe, you could see it forming along the outflow boundary before you could tell what was going on in radar.
- It was easy to see cu clouds feeding into the supercells
- It was informative to observe low clouds moving under the anvil, a depiction of deep layer shear.
SRSOR winds
- Early in the day and in the 18z radiosonde, there was not much shear. But later, the satellite-winds indicated strengthening westerly flow aloft.
- Helped to find areas of low-level convergence.
- The winds indicated convergence along a boundary ahead of the dryline. We did get enhanced cu development in this area. The winds alerted us to areas of concern.
- When upper-level winds came in with a storm anvil, we could see veering with height.
- This will be useful in extracting a pseudo wind profile, and monitoring layer wind shear.
NUCAPS
- Looking at the Plan view and cross section displays in FWD, I think these can be very helpful.
- Mixing ration plan view at 700 mb showed a nice gradient near the CWA boundary. Convection was dying as it crossed this gradient into the drier air.
- The highest mixing ration values were collocated with the storms in Mexico.
- The NUCAPS data is a great check on the models
- NUCAPS matched closely with the nearby Del Rio sounding in terms of CAPE and PW.
- It reinforced my understanding of the present thermodynamic environment.
- Prior to CI, I would look at NUCAPS while gauging the environment.
- If we do an 18Z sounding, then get NUCAPS a couple hours later, the temporal change would be helpful to see.
- I do not want to have to modify a NUCAPS sounding in the middle of warning operations.
- With the plan view, it was easy to pick out areas of bad data. So in this case, I would not want a model adjustment. But I would want an adjustment to the actual profile.
- I like the idea of not using model data for the low-level correction. I think it would be better to do whatever correction you can with available observations.
- Even without the 18Z sounding, I think NUCAPS would be helpful, assuming you have someone to modify the sounding. This would fill the gap of no 18Z special sounding.
- 130 PM timing for NUCAPS is perfect. It is right before CI usually.
-
LAP
- CAPE values were slightly less than those from RAP, but I understand this because GFS was fairly low, which is used as the first guess in the retrievals. Although, RAP may have been high.
- The LAP CAPE gradient was right through our CWA. It was helpful to focus on the location and movement of this gradient, as convection developed right along it.
- Layer PW was helpful for identifying drying aloft. I thought with wind and hail, and TPW values being relatively low, this seemed correct.
- LAP LPW showed really dry air aloft. This was apparent in the RAOBs, and confirmed in LPW.
GOES-R CI
- It was not good yesterday in FWD. We didn't see any probabilities from severe or regular CI despite having cu develop and initiate into severe storms. There did not appear to be cirrus contamination.
- There were instances when severe CI was higher than regular CI
- In San Angelo, things were slow to get going. Early on, there were no probabilities until convection had already began to develop. Later in the day it was better. It keyed in on new development further south when we were focused on the ongoing convection to the north.
ProbSevere
- For the first storm of the day that got a warning yesterday, I was very confident that it was going to go severe. It looked good on radar with a strong reflectivity core. I put a warning on it despite probsevere being only 40-50%. After the warning, the storm weakened pretty quickly. Lightning and shear had both been low.
- Yesterday, if probabilities went over 80, I thought it had a good chance of going severe.
- I went liberal on one warning. Without ProbSevere I would have waited to warn. It definitely added lead time in that situation.
- Through using ProbSevere, I could see warning lead time going up, but also FAR rate increasing.
- Towards the end of the day yesterday, probsevere was in the 90s with a storm, but based on radar reflectivities, it appeared that the storm was weakening. Lightning flash rates were high at this time.
Lightning Jump
-Training needs to be improved for this, as I did not understand the LJ before using it yesterday
- I noticed a lot of fluctuations in LJ in a cell.
- It helped with my confidence, seeing jumps with the rapidly developing storms
- I liked using the 4-panel with ProbSevere and LJ together. The first time I used the two together, ProbSevere gave me more confidence in what I was seeing in radar, and LJ gave me confidence in what I was seeing in Probsevere. I liked this.
- I didn't feel like I needed to see a new lj to warn, but it added confidence.
- For me, it was more a a SA tool, alerting us to cells that we hadn't really been watching yet but needed to start keeping an eye on. It was an early indication that we need to look at the storm further.
- I actually liked the display, the blob really jumps out. We aren't really interrogating radar in that panel anyway, so it is fine that it gets covered up.
- I could see LJ being a confidence booster in pulse thunderstorm cases. Maybe not as much with a qlcs event dropping wind damage, when we know it is severe.
SRSOR
- With one storm I was watching, a storm split was obvious in the 1-min imagery much earlier than it was in the radar imagery.
- I could see the strengthening of the updraft and OT well in multiple supercells quicker than was apparent in radar.
- I enjoyed looking at IR imagery. The OTs as indicated by areas of very cold temperature were pretty brief, but the 1-min imagery was able to capture them. They weren't long-lived OTs, occuring in-between radar and routine satellite scans. So it was helpful to have this rapidly updating information.
- I could see the Overshooting Tops shoot up before strengthening in radar.
- With a cell that ended up being severe, you could see it forming along the outflow boundary before you could tell what was going on in radar.
- It was easy to see cu clouds feeding into the supercells
- It was informative to observe low clouds moving under the anvil, a depiction of deep layer shear.
SRSOR winds
- Early in the day and in the 18z radiosonde, there was not much shear. But later, the satellite-winds indicated strengthening westerly flow aloft.
- Helped to find areas of low-level convergence.
- The winds indicated convergence along a boundary ahead of the dryline. We did get enhanced cu development in this area. The winds alerted us to areas of concern.
- When upper-level winds came in with a storm anvil, we could see veering with height.
- This will be useful in extracting a pseudo wind profile, and monitoring layer wind shear.
NUCAPS
- Looking at the Plan view and cross section displays in FWD, I think these can be very helpful.
- Mixing ration plan view at 700 mb showed a nice gradient near the CWA boundary. Convection was dying as it crossed this gradient into the drier air.
- The highest mixing ration values were collocated with the storms in Mexico.
- The NUCAPS data is a great check on the models
- NUCAPS matched closely with the nearby Del Rio sounding in terms of CAPE and PW.
- It reinforced my understanding of the present thermodynamic environment.
- Prior to CI, I would look at NUCAPS while gauging the environment.
- If we do an 18Z sounding, then get NUCAPS a couple hours later, the temporal change would be helpful to see.
- I do not want to have to modify a NUCAPS sounding in the middle of warning operations.
- With the plan view, it was easy to pick out areas of bad data. So in this case, I would not want a model adjustment. But I would want an adjustment to the actual profile.
- I like the idea of not using model data for the low-level correction. I think it would be better to do whatever correction you can with available observations.
- Even without the 18Z sounding, I think NUCAPS would be helpful, assuming you have someone to modify the sounding. This would fill the gap of no 18Z special sounding.
- 130 PM timing for NUCAPS is perfect. It is right before CI usually.
-
Final Analysis DFW 5/10/16
Would like to follow up on all the experimental products and how I felt things went today.
LAP CAPE: This product performed well and tried to correct the underperforming GFS. Did a great job of pinpointing that our best instability would be quite a bit to the east of the dryline. This indeed was where we ended up with the best storms of the day. All in all I found this product useful.
LAP Layer PW: Performed well today in depicting the evolution of the dryline throughout the day. This product continues to prove useful.
Atmospheric motion vectors were once again instrumental in identifying boundaries in the convective initiation phase. It was a great situational awareness tool to gague where the next convection would possibly occur.
ProbSevere did a great job in depicting where the best cells were. It also did a very good job of showing early on which storms might not make it to severe criterea. These storms pulsed up quickly but then quickly died. ProbSevere never made it over 60% in these situations. In one case, I issued a warning and it did not verify. ProbSevere won the day.
NUCAPS soundings were marginally helpful in skew-t format, but what I thought was really insightful was the plan view of fields such as mixing ratio.
CI was really the disappointment of the day today. The CI barely caught any of the cells in our area (maybe one). Also data appeared confusing as CI Severe values were often higher than regular CI values. This product was not very helpful at all today.
Super Rapid Scan was a great help today, especially in identifying overshooting tops when multiple storms were firing. It was a great suppliment to radar data, led to confident warnings and even was used to identify splitting cells. Really like this product in my warning operations.
See other posts today for example of all of the above. - Jason Bourne
LAP CAPE: This product performed well and tried to correct the underperforming GFS. Did a great job of pinpointing that our best instability would be quite a bit to the east of the dryline. This indeed was where we ended up with the best storms of the day. All in all I found this product useful.
LAP Layer PW: Performed well today in depicting the evolution of the dryline throughout the day. This product continues to prove useful.
Atmospheric motion vectors were once again instrumental in identifying boundaries in the convective initiation phase. It was a great situational awareness tool to gague where the next convection would possibly occur.
ProbSevere did a great job in depicting where the best cells were. It also did a very good job of showing early on which storms might not make it to severe criterea. These storms pulsed up quickly but then quickly died. ProbSevere never made it over 60% in these situations. In one case, I issued a warning and it did not verify. ProbSevere won the day.
NUCAPS soundings were marginally helpful in skew-t format, but what I thought was really insightful was the plan view of fields such as mixing ratio.
CI was really the disappointment of the day today. The CI barely caught any of the cells in our area (maybe one). Also data appeared confusing as CI Severe values were often higher than regular CI values. This product was not very helpful at all today.
Super Rapid Scan was a great help today, especially in identifying overshooting tops when multiple storms were firing. It was a great suppliment to radar data, led to confident warnings and even was used to identify splitting cells. Really like this product in my warning operations.
See other posts today for example of all of the above. - Jason Bourne
Atmospheric Motion Vectors Identify Boundaries
The Atmospheric Motion Vectors were great for identifying a boundary where low level convergence was occurring. They did a great job of filling in spots between observations and providing situational awareness on where the next cumulus would develop. Sure enough, there was some enhance cumulus right along the boundary that the AMVs were depicting. I find this product very useful during the convective initiation stage as highlighting locations where the first echos of the day may occur.
GOES Derived Winds Utility
GOES derived winds proved to be useful in determining when convective activity was getting deep enough to tap into some better shear.
Based on the KDRT 18Z sounding, winds were southerly up to around 600 mb. Above 600 mb winds become more westerly and stronger.
The GOES derived winds have consistently showed the southerly low level winds over the past couple of hours (blue wind barbs). Finally, between 20:00-20:30Z, convective activity become deep enough for GOES to pick up a wind measurement higher up in the atmosphere. The red wind barb showed 50 KT winds out of the west up around 350 to 300 mb. Within the hour after the convection achieved this better deep layer sheer, we saw thunderstorms develop and several became severe with 1" hail reports.
- JP
Based on the KDRT 18Z sounding, winds were southerly up to around 600 mb. Above 600 mb winds become more westerly and stronger.
- JP
Interesting feature in 1-min imagery
Monitoring the developing cu in the San Angelo CWA, we noticed some interesting evolution in the 1-min imagery. We noted some condensation blowing off of the cu/updrafts, as can be seen in the animation below.
We don't want to call them orphan anvils because they aren't reaching the level of an anvil/glaciation yet. We figure the updrafts are being inhibited by the stable capping layer noted in the Del Rio sounding (to the south) between 800 mb and 850 mb, before blowing off at that level. It is indicating to us that uprafts are begining to get there act together, and initation should be coming in the next hour or so as strong isolation continues.
Below is an image of the satellite-derived motion vectors. Note the blue vectors are tracking the cu, located at 925 mb. Additionally, a vector assocated with one of the blow-offs was sampled at 831 mb. This is where the capping inversion was located, confirming our suspisions that these early updrafts are indeed being capped by a stable layer. This also allows us to look at low-level shear. We have the metar at the surface idnicating 10 kts, a slightly veered and stronger (16 kts) sat-derived wind at 925 mb, and then an even stronger (26 kts) and more veered wind at 831 mb.
We don't want to call them orphan anvils because they aren't reaching the level of an anvil/glaciation yet. We figure the updrafts are being inhibited by the stable capping layer noted in the Del Rio sounding (to the south) between 800 mb and 850 mb, before blowing off at that level. It is indicating to us that uprafts are begining to get there act together, and initation should be coming in the next hour or so as strong isolation continues.
Below is an image of the satellite-derived motion vectors. Note the blue vectors are tracking the cu, located at 925 mb. Additionally, a vector assocated with one of the blow-offs was sampled at 831 mb. This is where the capping inversion was located, confirming our suspisions that these early updrafts are indeed being capped by a stable layer. This also allows us to look at low-level shear. We have the metar at the surface idnicating 10 kts, a slightly veered and stronger (16 kts) sat-derived wind at 925 mb, and then an even stronger (26 kts) and more veered wind at 831 mb.
Monday, May 9, 2016
Daily Summary: Week 4, Day 1 (9 May 2016)
The final week of the HWT 2016 GOES-R/JPSS Spring Experiment has begun! The GOES-R/JPSS Proving Ground/User Readiness meeting is taking place here in the National Weather Center in Norman also this week. Attendees to that meeting will be visiting the HWT to get a first hand look at the evaluations currently taking place. Day 1 of the final week was a busy one, with severe convection, including confirmed tornadoes, near Norman. Below are some photos from the day, along with feedback from our daily debrief.
 | ||
| Photo of large tornado that formed near Sulphur, OK (south of Norman) courtesy of our HWT/EWP Operations Coordinator, Gabe Garfield. |
 |
| Looking at a tornado-warned storm east of Norman from the roof of the NWC. Radar image and warning are in figure to right. |
LAP
- I like looknig at the LPW, especially in the lowest level, because you could see where the moisture transport was. It gives you more information than just TPW. I see this being useful for seeing moisture return from the GoM, as different flow regimes often bring moisture up into a region.
- I thought LPW was helpful. I appreciated that the mid and upper layers showed drying aloft, with greater moisture at the low-levels.
- It was nice to see the dry line depicted in the low layer LPW field.
- CAPE values matched up with 18z Norman sounding, and SPC mesoanalysis.
CI
- In Norman CWA, it highlighted the main area of convective development in the western part of the CWA. It was a little slow, however, as I could see the devleopment in visible imagery by the time the CI algorithm . targeted it.
- The strongest storm in the Norman CWA that produced a large tornado had a 70% severe CI on it.
- If a product performs well over time, I have confidence to use it later.
- In the NW part of the Norman CWA, a line of severe storms had a severe CI over 50 prior to development
- In FWD, an area had high CI well ahead of initiation, and severe CI around 50. This was higher than the rest of the region, so it attracted our attention. This area ended up developing and becoming severe.
- This is a product that, if you can trust it, was good in the situation we had yesterday. I was paying so much attention to our one tornadic storm. Then to the south, there was a cluster of high CI that brought my attention down there. Without CI, I would not have looked down there as early as I did.
- I can see this being useful in the southeast US. With pulse severe thunderstorms and large CWA's, storms could develop anywhere on the map in a given day. This product helps a forecaster with where to look next.
ProbSevere
- I noticed large jumps in ProbSevere simultaneous to storms increasing in strength, so it didn't give us much of an increase in lead-time yesterday. We tended to see large jumps at about the same time that we normally would have issued based off of standard data. For us it increased confidence in these warning decisions.
- I think it is especially helpful for the first cells of the day. ProbSevere jumped into the 60s. It was helpful for marginal cells at the beginning of the shift that were on the edge of becoming severe.
- Later, cells were obviously severe based on other datasets, so it wasn't as helpful. But still, it was great for confidence, especially watching trends in the probabilities. It is also helpful to see the predictor output.
- I like this product a lot, I am already using it in my office (RAH)
- I was most confident to issue at the 80% threshold
- Yesterday in FWD, the rapid increases in probabilities helped. The rate of change of probsevere triggered a severe tstorm warning. I wonder if a rate of change of probsevere prob could be a product on its own.
- It was very helpful to see why the probabiltiies changed via the predictors.
- With the first cell, we issued a warning when we saw probs quickly go from 45 to 65. No lightning wit hthis storm.
- In busy situations, it is helpful for discriminating storms that are more likely severe vs those that weren't. I don't remember a high probsevere yesterday that did not warrent a warning.
SRSOR
- Seeing OTs at this temporal resolution was helpful; it allowed us to figure which storms had the strongest updrafts at any given moment.
- For initial development along the dryline, convection went up very fast. Without 1-min data, we wouldn't have been able to recognize so soon that convective initiation was occurring.
- This validate dthe CI product. I'd see soemthing on CI and then go to 1-min data to validate it.
- 1-min imagery changes the game for interrogating storms before they are seen on radar.
- Having 1-min data available to see things earlier allowed me to get the alert out sooner.
- Seeing stuff go up earlier was the greatest benefit to me. I have an earlier idea of where the strong updrafts were.
SRSOR winds
- In FWD, I noticed a difference between the RAP analysis and satellite-derived winds. The RAP had significant LL convergence that was not apparent in the SRSOR winds. Nothing ever developed in that area. My initial discussion mentioned this, and how the satellite-winds made me doubt that covnection would develop since I was not seeing convergence. Also, not much directional shear.
Lightning Jump
- There was a 26-sigma jump on the tornado-producing cell prior to it producing a tornado, but we had already warned on it.
- I prefer the ProbSevere contour display over the LJ blog display.
NUCAPS
- If we hadn't had the 18Z RAOB, it would have been useful. 18Z soundings told us what we needed. 20Z NUCAPS matched up well.
- NUCAPS didn't resolve the CAP in the FWD CWA; this was a downfall. The 18Z FWD radiosonde showed that the CAP was actually still present.
- Having NUCAPS 1 hour earlier would have helped with the severe weather in the Norman CWA.
- The parameters matched up well with other data sources after I correct it.
- We have NUCAPS set up in our office. We've had success in using it to determine elevated convective potential, seeing moisture coming in around 600-700 mb, closed lows over CA.
- In warning siutations, having to put so much effort into modifying the sounding is a deal breaker.
- I think if you have a mesoanalysis on shift, they would look at NUCAPS during severe weather operations, and provide updates to the radar operators.
- For me, the most useful aspect of NUCAPS was for checking instability. Those values looked on par yesterday, telling us that the environment would support severe storms.
- I would be fine with having an error line visible in NSHARP with NUCAPS. Training on this would be important.
DFW Mesoscale Discussion May 9
Chance for severe storms increasing across the DFW area ahead of a dryline that is moving into the western portions of the CWA at this time. LAP CAPE and LI is honing in on an area of more favorable instability across the north central counties which is for the most part lining up well with RAP SPC mesoanalysis of ML CAPE and this is confimed by an 18z sounding from Ft. Worth. With fat CAPE in the hail growth zone and very strong mid level lapse rates...large hail should be the biggest threat of the day. With shear values into the 40 kt and above range...convective mode should be supercells. Although large hail will be the primary threat...we are not out of the woods as far as tornadoes are concerned either. The threat for tornadoes will increase in the northeastern portion of the cwa and points north as effective helicity and 0-1 km helicity values are in the 200-300 m2s2 range. LCL values are a little bit higher that ideal for tornadoes so it is possible we see a lot of rotating mesocyclones with few making down to the surface.
As far as low level convergence is concerned along the dryline...Satellite derived winds are not showing as much low level convergence as the RAP mesoscale analysis...shown below.
This makes me think that convective initiation may take a little longer to get going and thus pushing the threat slightly eastward. Fully expect the northeastern quadrant of the cwa to see severe wx but not as confident across the north central near the current position of the dryline. As far as straight line winds are concerned...DCAPE values along the dryline are about 1200 J/kg and observed sounding shows good amount of dry air below 700 mb indicating some downbursts are possible.
Convective initiation values nailed the first radar echos in southern Oklahoma and noted that compared to a fleeting higher CI value, a cluster of high CI and Severe CI values seemed to lend more confidence to the product.
-Jason Bourne
Friday, May 6, 2016
Week 3 (2-6 May 2016) Summary and Feedback
Week 3 concluded out West, with participants forecasting for the Boise and Pendleton CWA's. Severe activity was once again marginal, but there were enough storms to keep participants busy evaluating the products.
LAP
- Values (where we had retrievals) matched NUCAPS, which matched with modified observed soudings and observed soundings throughout the week.
- LAP retrievals seemed to correct underdone GFS CAPE throughout the week.
- It was helpful to have PW broken up into a few layers
- Layer PW gives you a good sense of where differential moisture advection is occurring.
- I liked that GFS filled the gaps in retrievals. It was also nice to compare the retrieval data and model data.
CI
- I would load it in my office; the information was useful.
- The regular CI product performed great this week.
- CI trained your eye to focus on areas of cloud development. I used it several times for situational awareness. It gave a lot of lead-time in DC, where it told me to watch that area prior to development.
- I see this being greatly improved with 5-min updates from GOES-R
- Generally we all liked the display.
ProbSevere
- The display made sense, and I adapted to it fast
- Ideas for predictors that might improve probabilities include: low-level divergence signature in velocity data for wind events, DCAPE for wind, surface-based theta-e for derechos, surface cold pool difference for wind.
- All said that it would be helpful to have the probability broken down be severe threat. They would load it as a 4-panel.
- Loved it, great tool
It tends to blend together nearby storms.
- With the storm on the west side of DC, we had a storm split, but ProbSevere grouped the storms together, so it was unclear from which cell the information was coming.
- I would like it to be able to differentiate storms, especially in linear events.
- There is already getting to be a lot of information in the readout, adding more might get to be a lot. It would be a good idea to color the data, or make the output selectable by the user.
- On marginal days, it confirmed what we were thinking about the storms.
- On one of the marginal days, we warned on a storm that reached 50%. It seems like you need to guage the threshold for each particular day
- ProbSevere and lightning jump give you added confidence to issue warnings or continue warnings if on the bubble. Good warning decision tools.
SRSOR
- We would like to have both the regular imagery and Parallax-correct imagery available to us. There were cases when it helped but also a case where it hurt.
- I wish the 1-min imagery was available more often.
- I can't wait to get this from GOES-R
- I think this will be beneficial for tracking the development and evolution of lake effect snow, where you have low-topped, but intense growth. For me this will be the case on Lake Michigan.
- I see myself using the 1-min imagery during warning operations.
- I compared satellite and radar imagery side-by-side. This allowed me to match storms, and compare storm behavior in the two.
- I overlayed IR and visible imagery. This allowed me to see cloud top trends, temperatures, etc plus the detail in one panel.
- I loaded warning polygons on the srsor imagery. This allowed me to match with storms in radar more easily.
- I could see a visible cloud line along the rear flank downdraft.
- We depended on the 1-min imagery even more when we were operating out west.
SRSOR winds
- Participants would like having a gridded analysis of these winds, in addition to the wind barbs themselves. Fields like wind speed, divergence, and vorticity would be helpful. It is sometimes difficult to visualize these fields from just the barbs.
- On Monday in W Maryland, I could see low-level mass flux occuring in the vicinity of a cloud line. It was nice to quantify it with the winds.
- It would be nice to somehow automate the computation of layer shear.
Lightning Jump
- ProbSevere and Lightning Jumps give you added confidence to issue warnings or continue warnings if on the bubble. These are good warning decision tools.
- I would like to see how negative jumps correlate to echo top collapses.
- I like the colors. They grab my attention. The upper threshold should be increased.
NUCAPS
- It would be neat to add winds to the profile. Some sources could be surface obs and satellite derived winds.
- We would like to see the surface modifications be automated.
- It was helpful to have the plan view displays. I would like to see fields like theta-e, cape, pw, height of freezing level, height of -20C level
General
- All commented that they enjoyed the experience, and it was just the right amount of products to evaluate.
- It would be nice to have a WES case for slower days.
- After initial assessment before convection starts, have a group briefing to see what the two groups were observing.
- Broadcaster said it was a great experience, and it was beneficial to work alongside the NWS forecasters.
LAP
- Values (where we had retrievals) matched NUCAPS, which matched with modified observed soudings and observed soundings throughout the week.
- LAP retrievals seemed to correct underdone GFS CAPE throughout the week.
- It was helpful to have PW broken up into a few layers
- Layer PW gives you a good sense of where differential moisture advection is occurring.
- I liked that GFS filled the gaps in retrievals. It was also nice to compare the retrieval data and model data.
CI
- I would load it in my office; the information was useful.
- The regular CI product performed great this week.
- CI trained your eye to focus on areas of cloud development. I used it several times for situational awareness. It gave a lot of lead-time in DC, where it told me to watch that area prior to development.
- I see this being greatly improved with 5-min updates from GOES-R
- Generally we all liked the display.
ProbSevere
- The display made sense, and I adapted to it fast
- Ideas for predictors that might improve probabilities include: low-level divergence signature in velocity data for wind events, DCAPE for wind, surface-based theta-e for derechos, surface cold pool difference for wind.
- All said that it would be helpful to have the probability broken down be severe threat. They would load it as a 4-panel.
- Loved it, great tool
It tends to blend together nearby storms.
- With the storm on the west side of DC, we had a storm split, but ProbSevere grouped the storms together, so it was unclear from which cell the information was coming.
- I would like it to be able to differentiate storms, especially in linear events.
- There is already getting to be a lot of information in the readout, adding more might get to be a lot. It would be a good idea to color the data, or make the output selectable by the user.
- On marginal days, it confirmed what we were thinking about the storms.
- On one of the marginal days, we warned on a storm that reached 50%. It seems like you need to guage the threshold for each particular day
- ProbSevere and lightning jump give you added confidence to issue warnings or continue warnings if on the bubble. Good warning decision tools.
SRSOR
- We would like to have both the regular imagery and Parallax-correct imagery available to us. There were cases when it helped but also a case where it hurt.
- I wish the 1-min imagery was available more often.
- I can't wait to get this from GOES-R
- I think this will be beneficial for tracking the development and evolution of lake effect snow, where you have low-topped, but intense growth. For me this will be the case on Lake Michigan.
- I see myself using the 1-min imagery during warning operations.
- I compared satellite and radar imagery side-by-side. This allowed me to match storms, and compare storm behavior in the two.
- I overlayed IR and visible imagery. This allowed me to see cloud top trends, temperatures, etc plus the detail in one panel.
- I loaded warning polygons on the srsor imagery. This allowed me to match with storms in radar more easily.
- I could see a visible cloud line along the rear flank downdraft.
- We depended on the 1-min imagery even more when we were operating out west.
SRSOR winds
- Participants would like having a gridded analysis of these winds, in addition to the wind barbs themselves. Fields like wind speed, divergence, and vorticity would be helpful. It is sometimes difficult to visualize these fields from just the barbs.
- On Monday in W Maryland, I could see low-level mass flux occuring in the vicinity of a cloud line. It was nice to quantify it with the winds.
- It would be nice to somehow automate the computation of layer shear.
Lightning Jump
- ProbSevere and Lightning Jumps give you added confidence to issue warnings or continue warnings if on the bubble. These are good warning decision tools.
- I would like to see how negative jumps correlate to echo top collapses.
- I like the colors. They grab my attention. The upper threshold should be increased.
NUCAPS
- It would be neat to add winds to the profile. Some sources could be surface obs and satellite derived winds.
- We would like to see the surface modifications be automated.
- It was helpful to have the plan view displays. I would like to see fields like theta-e, cape, pw, height of freezing level, height of -20C level
General
- All commented that they enjoyed the experience, and it was just the right amount of products to evaluate.
- It would be nice to have a WES case for slower days.
- After initial assessment before convection starts, have a group briefing to see what the two groups were observing.
- Broadcaster said it was a great experience, and it was beneficial to work alongside the NWS forecasters.
Thursday, May 5, 2016
Overview of Thursday, May 5, 2016.
Today, we were in the Pendleton, OR CWA, for a Marginal Risk from SPC. The main feature to focus on were 500 mb vorticity maxima moving northwest into the region during the afternoon. This would combine with cold temperatures aloft to product ample MUCAPE. Deep layer shear was relatively strong (40 to 50 knots) in the eastern portions of the area, weaker to the west.
The LAP products provided good estimates of CAPE during the afternoon, when sampled with the cloudy sky data. There were not many locations that had this, but was inferred to be over most of the area with the modified NUCAPS soundings. Using these products together helped gain confidence in the amount of CAPE over the area.
The GOES winds also helped with estimating the deep layer (700 mb to 250 mb) shear over the area, as the afternoon evolved. This was handy to assess the environment and storm mode.
The CI product was helpful with showing CI potential, despite middle to high clouds blocking some areas. The CI values were low in these areas, but would have been much higher had there not been any middle to high clouds. Using the 1-minute SRSOR imagery to watch the cumulus development, and radar to confirm cell development, these low CI values were a clue that initiation was likely.
I did not get a chance to use Prob Severe and lightning jump much, as the convection was not strong. Some cells had enough lightning flash rates to bring the Prob Severe values to 30%, but they would weaken within a few radar volume scans. No lightning jumps were detected.
JJW
The LAP products provided good estimates of CAPE during the afternoon, when sampled with the cloudy sky data. There were not many locations that had this, but was inferred to be over most of the area with the modified NUCAPS soundings. Using these products together helped gain confidence in the amount of CAPE over the area.
The GOES winds also helped with estimating the deep layer (700 mb to 250 mb) shear over the area, as the afternoon evolved. This was handy to assess the environment and storm mode.
The CI product was helpful with showing CI potential, despite middle to high clouds blocking some areas. The CI values were low in these areas, but would have been much higher had there not been any middle to high clouds. Using the 1-minute SRSOR imagery to watch the cumulus development, and radar to confirm cell development, these low CI values were a clue that initiation was likely.
I did not get a chance to use Prob Severe and lightning jump much, as the convection was not strong. Some cells had enough lightning flash rates to bring the Prob Severe values to 30%, but they would weaken within a few radar volume scans. No lightning jumps were detected.
JJW
GOES Winds Detecting Strong Shear Aloft!
The GOES winds in the lower left corner of this 4 panel image shows 718 mb level winds of 14 knots, and 275 mb level winds at 77 knots, in north central Wallowa County at 2150 UTC. This would give approximate deep layer shear (in the 720 mb to 275 mb layer) of 63 knots. Pretty strong shear, and a great way to get the latest shear values across this area! JJW
GOES SRSOR cloud tracer
18Z GOES derived winds denote a lack of low level winds through north central NV. However GOES SRSOR visible imagery aides in tracking cumulus clouds through northern NV and gives a sense of the speed and direction of the flow in an otherwise data void area.
Tornadotod
 |
| 18Z GOES Satellite Derived Winds |
 |
| 1838-1927Z GOES SRSOR Loop (note the movement of developing cumulus over north central Nevada) |
Satellite Winds Showing Upper Level Diffluence
The upper level winds in the western portion of Boise's forecast area are showing some upper level diffluence. This is also where some storms are starting to fire per recent radar imagery...
~Lilly Miller~
GOES Winds Estimating 700-300mb Shear - Need More 700mb Samples.
The lower left image shows the 1-minute SRSOR imagery with the GOES winds overlaid. The cursor readout over the far north central portion of the Pendleton, OR CWA shows one sample of winds at 718 mb of 21 knots. It also shows a sample of 362 mb winds of 44 knots. This gives an approximate 700-350 mb shear of 23 knots, which is rather weak. However, sites to the southeast of that area are showing 250 mb winds in the 70-75 knot range. This would give larger shear values from 700-250mb (100 mb deeper than the sample location), assuming the 700 mb winds are similar to the sample point. The lack of sampling of 700 mb winds suggest that this assumption cannot be made. Even so, GOES wind data at 700 mb would be something to watch for as the afternoon progresses. JJW
Wednesday, May 4, 2016
Daily Summary: Week 3, Day 3 (4 May 2016)
Today was a very quiet day. One group operated in Florida (Miami and Tampa Bay), monitoring ongoing convection coming off the GoM. The other group operated in the Memphis CWA, tracking sub-severe convection associated with a cold core low advancing through the region.
LAP
- Correctly increased CAPE... GFS has been underdone for CAPE each day this week, so it was nice to see LAP correctly increase the values.
- Clear sky and cloudy sky retrieval values were closer to other data sources than were GFS values.
- Convection developed within the max in LAP PW, which was maximized vertically through the atmosphere at that location as seen in layer PW
- Participants commented that they like the inclusion of GFS model data to fill holes in retreivals
CI
- Pinged development in S Florida
- It worked well for us yesterday.
- There were some that it missed, but there were hits too.
ProbSevere
- Never got out of the single digits. It was still useful information, since convection was not severe
- Having estimates of CAPE and shear were useful
- Confirmed our thinking of what the mesoscale environment would be like
- It did pick out cells within the large cloud shield
- Broadcaster: In my office, we use GR, so we could pull in probsevere files
SRSOR
- In the mid-level cloud plume along the pre-frontal trough, it was helpful to see convective clouds poking through.
- It is helpful to compare SRSOR with radar imagery to match up cells
- I liked the sandwich product, this would be helpful to watch in real-time
SRSOR winds
- The winds nicely showed the shift with drier air moving through.
Lightning Jump
- The few jumps we had were low, which fit the situation.
NUCAPS
- I looked at NUCAPS offshore, which showed drier air aloft, which made sense as drier air was moving in.
- The modified sounding over W TN showed around 500 j/kg of CAPE, which looked right, giving we confidence in how the environment had evolved since 12Z.
- With the Plan view displays it was helpful to see the environment at a particular level.
- With the plan view, you can quickly see how the airmass is changing, more or less stable, based on mid-level cooling temperatures which steepens lapse rates
- The Plan view display is most beneficial in the clear sky pre-convective environment
- I would use theta-e to diagnose areas of CSI and strong frontogenesis.
LAP
- Correctly increased CAPE... GFS has been underdone for CAPE each day this week, so it was nice to see LAP correctly increase the values.
- Clear sky and cloudy sky retrieval values were closer to other data sources than were GFS values.
- Convection developed within the max in LAP PW, which was maximized vertically through the atmosphere at that location as seen in layer PW
- Participants commented that they like the inclusion of GFS model data to fill holes in retreivals
CI
- Pinged development in S Florida
- It worked well for us yesterday.
- There were some that it missed, but there were hits too.
ProbSevere
- Never got out of the single digits. It was still useful information, since convection was not severe
- Having estimates of CAPE and shear were useful
- Confirmed our thinking of what the mesoscale environment would be like
- It did pick out cells within the large cloud shield
- Broadcaster: In my office, we use GR, so we could pull in probsevere files
SRSOR
- In the mid-level cloud plume along the pre-frontal trough, it was helpful to see convective clouds poking through.
- It is helpful to compare SRSOR with radar imagery to match up cells
- I liked the sandwich product, this would be helpful to watch in real-time
SRSOR winds
- The winds nicely showed the shift with drier air moving through.
Lightning Jump
- The few jumps we had were low, which fit the situation.
NUCAPS
- I looked at NUCAPS offshore, which showed drier air aloft, which made sense as drier air was moving in.
- The modified sounding over W TN showed around 500 j/kg of CAPE, which looked right, giving we confidence in how the environment had evolved since 12Z.
- With the Plan view displays it was helpful to see the environment at a particular level.
- With the plan view, you can quickly see how the airmass is changing, more or less stable, based on mid-level cooling temperatures which steepens lapse rates
- The Plan view display is most beneficial in the clear sky pre-convective environment
- I would use theta-e to diagnose areas of CSI and strong frontogenesis.
Subscribe to:
Posts (Atom)