Friday, June 12, 2015
2015 SPRING EXPERIMENT IS COMPLETE
The 2015 Spring Experiment which took place in the Hazardous Weather Testbed for 5 weeks between May 4 to June 12 is now complete. Feedback from the blog, survey, and daily discussions will be analyzed and organized into a final report to be made available this summer. Please come back in the future to view posts during future GOES-R demonstrations in the HWT!
Thursday, June 11, 2015
Week 5 Summary and Feedback (8-12 June 2015)
The final day of Week 5 and the 2015 Spring Experiment took us to the Boulder, Topeka, and Omaha CWA's. The Boulder group had ample opportunity to evaluate the PGLM total lightning products. As with the previous several days, the 1-min satellite imagery from GOES-14 was available over the areas of interest and utilized extensively by participants.
- Bill Line, SPC/HWT Satellite Liaison
GOES-R LAP
- Best to pay attention to gradients, the absolute values weren’t that accurate (for CAPE)
- Higher temporal frequency will be appreciated in goesr era
- Various use cases (convective, flooding, winter, etc.) of how these products can be applied in operations would be helpful to see in training
NUCAPS
- Improve training to present how to use, include more information on editing lower levels, including info on how high to mix
- I prefer to look at rap, even though its model data, because it shows me the fine details that NUCAPS does not capture
- Fusing of all the sources is really the way to go, they should all be blended together, instead of having to use them all
GOES-R CI
- I liked using it this week, but I would want it to work better under thin cirrus
- I remember looking at it a few years ago, and it’s a lot better now than it was then. It provided useful information before event.
- When you get constant rapid scan, it will be available more timely, more often, which will certainly help
- Seeing it overlaid with more in awips would be great. I look at it on webpage now
- For cwsu it will be especially useful, seeing signals will lead me to start the alert.
ProbSevere
- It has its flaws, so people need to know these, must be in training.
- I could see us using that on air, people understand percentages
- I almost view it more as SA type tool; stick with base data as gold standard for warnings,
- On busy day, I could see it being useful, have it up on radar or sat, for SA
- Have something/fields to get at low-topped convection
- Perhaps regionalize probsevere – train dataset on a local area instead of whole country
- Storm motion information could help – eg. acceleration before wind damage
- Someway to prevent the merger with linear storms, perhaps track at higher dbz once storms mature.
- If polygon gets over a certain size, throw in a new threshold, pick out cores within line
- Dbz at a certain temperature level might be useful input
SRSOR and 1-min OTD
- TV viewers would love to look at that.
- Especially useful when debating on issuing another warning downstream, if you could see another updraft coming in, might help to issue or not
- Outflow boundary interaction could clearly be seen
- Low stratus movement, lake and ocean breezes.
- In san fran with stratus – It would help with aviation, timing of amendments knowing exactly where edge of stratus is asap is important
- I would see us using it all the time. West coast wild fires, any hazardous material.
- This could lead to more project ideas, discovering more ways to utilize data
- Want srsor all day every day
- Will always be necessary somewhere in the US
Lightning jump
- We were seeing very large sigma jumps.
- Showed significant jumps with strengthening storms
- More useful over high terrain, where radar coverage isn’t as good.
- I would like it to include a readout of the number of flashes that it jumped.
- Will be good to include this information in probsevere
- Area in cwa with bad radar coverage, something that is nationwide would be great, nice to see in these data void areas
PGLM
- Provides good SA, shows precisely when storms pulse up and down.
- Definitely will be helpful for DSS events. Someone deployed, they can track lightning activity
- Bill Line, SPC/HWT Satellite Liaison
GOES-R LAP
- Best to pay attention to gradients, the absolute values weren’t that accurate (for CAPE)
- Higher temporal frequency will be appreciated in goesr era
- Various use cases (convective, flooding, winter, etc.) of how these products can be applied in operations would be helpful to see in training
NUCAPS
- Improve training to present how to use, include more information on editing lower levels, including info on how high to mix
- I prefer to look at rap, even though its model data, because it shows me the fine details that NUCAPS does not capture
- Fusing of all the sources is really the way to go, they should all be blended together, instead of having to use them all
GOES-R CI
- I liked using it this week, but I would want it to work better under thin cirrus
- I remember looking at it a few years ago, and it’s a lot better now than it was then. It provided useful information before event.
- When you get constant rapid scan, it will be available more timely, more often, which will certainly help
- Seeing it overlaid with more in awips would be great. I look at it on webpage now
- For cwsu it will be especially useful, seeing signals will lead me to start the alert.
ProbSevere
- It has its flaws, so people need to know these, must be in training.
- I could see us using that on air, people understand percentages
- I almost view it more as SA type tool; stick with base data as gold standard for warnings,
- On busy day, I could see it being useful, have it up on radar or sat, for SA
- Have something/fields to get at low-topped convection
- Perhaps regionalize probsevere – train dataset on a local area instead of whole country
- Storm motion information could help – eg. acceleration before wind damage
- Someway to prevent the merger with linear storms, perhaps track at higher dbz once storms mature.
- If polygon gets over a certain size, throw in a new threshold, pick out cores within line
- Dbz at a certain temperature level might be useful input
SRSOR and 1-min OTD
- TV viewers would love to look at that.
- Especially useful when debating on issuing another warning downstream, if you could see another updraft coming in, might help to issue or not
- Outflow boundary interaction could clearly be seen
- Low stratus movement, lake and ocean breezes.
- In san fran with stratus – It would help with aviation, timing of amendments knowing exactly where edge of stratus is asap is important
- I would see us using it all the time. West coast wild fires, any hazardous material.
- This could lead to more project ideas, discovering more ways to utilize data
- Want srsor all day every day
- Will always be necessary somewhere in the US
Lightning jump
- We were seeing very large sigma jumps.
- Showed significant jumps with strengthening storms
- More useful over high terrain, where radar coverage isn’t as good.
- I would like it to include a readout of the number of flashes that it jumped.
- Will be good to include this information in probsevere
- Area in cwa with bad radar coverage, something that is nationwide would be great, nice to see in these data void areas
PGLM
- Provides good SA, shows precisely when storms pulse up and down.
- Definitely will be helpful for DSS events. Someone deployed, they can track lightning activity
Combining SRSO and Radar Data
The high frequency satellite imagery enables the ability to better match the frequency of radar data and combine them to give a unique view in this case of the radar data near the ground and the top of the storm via satellite. In the following animations, you can see the supercell SE of PUX in feature following zoom mode in AWIPS. Evident are storm structure, outflow boundary, inflow, and anvil development and storm top outflow all in the same image. This helps verify the conceptual model of supercells. MrSnow/Shasta.
Click on the images to see the animation.
Click on the images to see the animation.
Evolution of outflow boundaries in SRSO vs standard visible imagery
Just to compare and contrast what GOES-R and SRSO offers compare to routine visible imagery is shown below. The SRSO loop contains 120 images while the standard imagery contains only 6 . There is so much more evident in the SRSO in terms of convective development and outflow boundaries and interactions with other cells not so easily seen or imagined in the standard imagery. Click on the images to see the animation. MrSnow/Shasta
NUCAPS soundings just in time
A nice clear field of view in was in the development region southeast of Denver. Based on the differences in CAPE analysis in the LAP, GFS, and RAP, the NUCAP sounding was available at the optimal time to check its vertical profile, observed and derived CAPE in the undisturbed environment where the storms were headed. This image below shows that clear area on the visible imagery vicinity KLHX La Junta, CO where the surface temp and dewpoint were 83/59. The unmodified NUCAPS sounding yielded about 1900 J/kg SBCAPE as the surface dewpoint was 60 but the temperature was much cooler than observed. However, modifying the sounding with the observed surface T/Td at KLHX, the resulted CAPE was over 2800 J/kg much higher than just about any other method. MrSnow/Shasta
LAP gets a bad RAP
Looked at the LAP CAPE output again today at 1800 UTC 11 June 2015 southeast of Den ver, CO. It did not seem to capture the magnitude of the CAPE. In the first image below the sampled 460 J/kg grid seems to represent where the satellite added information to the GFS background and agreed nicely with the GFS 6h first guess. The overall gradient and orientation seems pretty good as well.
The problem is that the CAPE seems very low when compared to the RAP analysis.
Given the strong supercell development, tend to believe the RAP vs the LAP analysis. The second image below shows the same sampled location in the LAP. That is overlaid with the RAP which is showing 1600 J/kg at 1800 UTC.
MrSnow/Shasta
The problem is that the CAPE seems very low when compared to the RAP analysis.
Given the strong supercell development, tend to believe the RAP vs the LAP analysis. The second image below shows the same sampled location in the LAP. That is overlaid with the RAP which is showing 1600 J/kg at 1800 UTC.
MrSnow/Shasta
Downstream SVR for Reno and beyond Charley
Lightning data steady state on time series. DVIL is a little low, but SRS showing another updraft with shadow in the area of the cell. Had a report earlier of pea size hail in Langdon, which was southeast of the core. ProbSevere indicated a mature storm with continued high values.
Integrating SRSO Visible imagery with 1 minute total lightning data
In the images below the SRSO 1 minute imagery and 1 minute total lightning data were integrated to provide a visualization/conceptualization of the evolution of the storms updraft location and distribution of cloud flashes (yellow circles) and cloud to ground strikes (dark blue +/-). We were able to save the editor display, open up the xml file in the localization perspective and change the frame count from 65 to 120 to get a longer loop. You can then open up the editor display for later use.
Also used the SRSO IR images, used the interpolate image on the imaging option to smooth out the IR. The Change Colormap option to adjust the color scale to better enhance the cold top of the storms to fit today’s situation. I also overlaid the GOES-R Pseudo Geostationary Lightning Mapper (PGLM) Surface Flash Extent Density product to monitor lightning along with the 1 minute ENTLN data.
Click on the images to see the animation. MrSnow/Shasta
Also used the SRSO IR images, used the interpolate image on the imaging option to smooth out the IR. The Change Colormap option to adjust the color scale to better enhance the cold top of the storms to fit today’s situation. I also overlaid the GOES-R Pseudo Geostationary Lightning Mapper (PGLM) Surface Flash Extent Density product to monitor lightning along with the 1 minute ENTLN data.
Click on the images to see the animation. MrSnow/Shasta
Strong PGLM surge leads to warning
A strong PGLM surge was noted just before issuing a warning. A strong to severe thunderstorm developed over the southern suburbs of Denver. The cell had a history of nickel sized hail over south Denver. An boost came in the form of an outflow boundary coming in behind the cell. A warning was issued. Unfortunately, it appears that the Lightning Jump algorithm was a no-show just before the warning was issued.
-Shasta/Mr. Snow
-Shasta/Mr. Snow
CI performing very well today in CO
Convective Initiation product (CI) correctly showed increasing values just south of the Denver Boulder area from 1600 to 1800 UTC 11 June 2015. Values increase from 36 to near 60% in the general vicinity of the storms that developed in the Pueblo, CO area. This gave about 30 minutes of lead time to the first lightning strikes around 1715Z. It would be nice to get CI output at 1-5 minute temporal resolution to match RSO/SRSO. We had outflow boundaries evident in the SRSO imagery but had to wait 15 minutes for CI updates. Below are a series of images as the storms were developing showing the CI, the MRMS Composite Reflectivity, Visible satellite, PGLM 6min summary, 0-200km Lighting Jump, 15minute ENI 8km interpolated lighting grid and cloud to ground strikes. The lightning jump showed up to 2 sigma jumps at 1729 UTC. MrSnow/Shasta.
- CI is/was available with RSO! - BL
- CI is/was available with RSO! - BL
Prob severe jumps from 19 to 91%
Prob Severe jumped as storm developed just west of my warning area. Should nicely move into my warning polygon.
To Warn or Not To Warn…
…we didn’t warn. Despite ProbSevere jumping to 75% and decent looking reflectivities, we decided not to warn on the cell in Jackson Co, KS. The MESH estimates were only up to .7″.
But the bigger deciding factor was unimpressive lightning data (seen below). lightningnojump
-BT
But the bigger deciding factor was unimpressive lightning data (seen below). lightningnojump
Some CI points in cirrus
In our northeast CWA as well as south of the Johnson county previously warned storm.
Warning based on Severe Probability alone
As an experiment…issuing the Severe Thunderstorm Warning based on several severe probability values of +90%.
We’ll see what happens…
-Shasta
We’ll see what happens…
-Shasta
Losing Lightning Jump Data
We seem to be sporadically losing the ENI data that generates the Lightning Jump algorithm. A storm quickly developing northeast of the Denver International Airport showed a strong jump in the lightning density from the PGLM. Meanwhile there wasn’t even a footprint for the Lightning Jump algorithm. When more raw data filtered in, the Lightning Jump information re-appeared.
-Shasta and Mr. Snow
-Shasta and Mr. Snow
Warning edge of Richardson Co.
Prob Severe was increasing with a signfcnt lightning jump indicated on the ENI data decided to warn on it before it crossed into MO and out of our CWA.
ProbSevere lower on Northern Storm Charley
Cell in the north had a DTA, but DVIL never jumped too high and 50 dbZ heights were not overly high. ProbSevere, as in the loop above also never jumped too high. Thus we did not issue a warning here.
NUCAPS Topeka
NUCAPS first guess was not that representative of the atmosphere when sampling a clear air sounding. After altering, we went from 0 to 3500 J/kg of CAPE. Which we thought was a bit high.NUCAPS before.
We did the same thing for an area farther north where we are expecting more instability. First NUCAPS sounding before altering did show a little CAPE (a few hundred), and altering took it back to ~3000 J/kg, a few J/kg lower than the southern, more stable environment. -BT
We did the same thing for an area farther north where we are expecting more instability. First NUCAPS sounding before altering did show a little CAPE (a few hundred), and altering took it back to ~3000 J/kg, a few J/kg lower than the southern, more stable environment. -BT
NUCAPS sounding OAX
Found a clear area just near the point of the KS/MO/NE line. Looked at the NUCAPS sounding and manipulated the boundary layer points some…did have a dry layer near 925mb..left most of that in…came up with a CAPE slightly higher than was showing on the SPC mesoscale analysis…3209 where the analysis was showing between 2500-3000.
Super Rapid Scan
Cell with warning on it now longing looks like it has a strong updraft in loop. Do have some sign of cells breaking through, even in our cooler northeast
SVR Storms Possible over Northeastern KS this Afternoon?
Instability is increasing again this afternoon in the wake of the departing complex of showers and storms from earlier this morning, which is raising the likelihood that we will see convection redevelop along and ahead of an approaching surface cold front. MLCAPE values are expected to climb into the 1000 to 2500 J/kg range this afternoon, highest near the NE/KS border. Given deep layer shear values are in the 30 to 40 kt range with a strong low level jet, expect we will see some scattered storms develop in the mid to late afternoon hours, especially across northern/northeastern portions of northeast KS. Given the strong mean southwesterly flow, any strong storms that develop ahead of the approaching cold front should quickly move northeastward and out of the area. The best tornado threat should remain just to the north of the area closer to the stalled boundary to the north of the region in southern NE. Expect our main threat will be from damaging wind and large hail, as there appears to be good -10 to -30 degree C CAPE and more associated with a developing line of storms late this afternoon into the evening (which could extend southwest to northeast across the area) as the front shifts eastward across the area.
Lincoln County, CO Severe
A couple of storms developed off of the high terrain south of the palmer divide near Pueblo’s forecast area. One even registered a +16 sigma jump. Although looking at the raw data, the drastic jump was probably because the relative sample size was quite small and any strong jump would be portrayed as impressive. Also of note, the +16 sigma occurred just as the updraft split occurred and the cell divided.
The eastern storm did show high sigma gains and the raw data went along with that. The strong lightning jump with 4 sigma occurred just before there was a strong jump in ProbSevere from 75% to 93%. The storm was already warned on based on base reflectivity data, but, it helped confirm that there was, in fact, a warnable storm.
-Shasta and Mr. Snow
The eastern storm did show high sigma gains and the raw data went along with that. The strong lightning jump with 4 sigma occurred just before there was a strong jump in ProbSevere from 75% to 93%. The storm was already warned on based on base reflectivity data, but, it helped confirm that there was, in fact, a warnable storm.
-Shasta and Mr. Snow
Prob SEV and DTA product working well
Prob Severe and ENI DTA product are providing a good first look right now as to where the most dangerous/severe TS are located as NMRS clusters have develops over the Southern part of OAX CWA.
Severe Omaha Charley
Have a quickly developing cell in our southern CWA.
Decided to issue a severe based on a quickly growing lightning time series (with little change in areal extent) as well as a high ProbSevere.
Decided to issue a severe based on a quickly growing lightning time series (with little change in areal extent) as well as a high ProbSevere.
Meso Discussion Omaha Charley
Have a strong gradient of temperatures across the region this hour, with upper 50s in the north and mid 80s in our one clear spot in the southeast forecast area.
Surface low is to the southwest with a frontal boundary extending northeast into our southern CWA. Quite a bit of shear aloft, with SRH from the SPC mesoanalysis in the 200-300 range across the south, higher in the southeast and 0-1 km bulk shear 10 in the west and 25 in the east. These would lead to potential for supercells to develop in the warmer air, and perhaps allow for tornado development close to the frontal boundary.
LAP sounder data is not too high for TT, with CAPE and LI more indicative of storm potential in the south and southeast. Have a PW gradient with higher values to the east.
clockwise starting with upper-left panel: CAPE, LI, Total Totals, K-Index
Surface low is to the southwest with a frontal boundary extending northeast into our southern CWA. Quite a bit of shear aloft, with SRH from the SPC mesoanalysis in the 200-300 range across the south, higher in the southeast and 0-1 km bulk shear 10 in the west and 25 in the east. These would lead to potential for supercells to develop in the warmer air, and perhaps allow for tornado development close to the frontal boundary.
LAP sounder data is not too high for TT, with CAPE and LI more indicative of storm potential in the south and southeast. Have a PW gradient with higher values to the east.
clockwise starting with upper-left panel: CAPE, LI, Total Totals, K-Index
clockwise starting with upper-left: LPW .7-.3, TPW, LPW .9-sfc, LPW .9-.7
CI product
CI product already showing a 89% prob in southern OAX…radar already showing storms building with lightning
Wednesday, June 10, 2015
Daily Summary: Week 5, Day 3 (June 10, 2015)
Warning activities for Wednesday of Week 5 took place in the Chicago, Hastings and Rapid City CWA's. Once again, significant convective activity was not present in any of the LMA's, so PGLM products were not able to receive an evaluation.
- Bill Line, SPC/HWT Satellite Liaison
GOES-R LAP
- Absolute values of CAPE were way off when compared to RAP, SPC mesoanalysis, RAOBS
NUCAPS
- Finer details would still be nice to see in the profiles
GOES-R CI
- Mid to high clouds obscured development in Chicago
- Once developing convection broke through the mid-level cloud deck, 80—90 % probabilities appeared
- Didn’t get many signals today in the Hastings CWA, so wasn’t useful. Thin cirrus were present
ProbSevere
- Our storms were so explosive, once it got close to 50, we would warn from radar because it was going so fast
- I like how if you’re focusing in on one cell (with a feeder band), watching outflow interaction, and then another cell developed nearby. It keeps you aware that you should go look at that now
- Especially valuable when you have lots of stuff on radar
- I could see it being on a SA panel
- Storm coordinator could use to make sure all relevant storms are covered
- I was looking at a storm that was weakening, so I focused my attention elsewhere… looked back a little later and saw the weakening storm was back to 80, telling me that I better take another look … good SA
- If you can’t look at all tilts, its seeing column for you
- New storm popped up when we were focused on other storms; this flagged us, quickly put severe on it
- Towards end of day, outflow boundaries collided, immediately flagged it to pink, then collapsed
SRSOR and 1-min OTD
- In Chicago, you could see where the front was and where cu were developing under a mid level cloud deck. Could see CI as it was occurring below those clouds.
- With new cells forming under cirrus canopy, you could see it bubble right through the anvil blowoff
- Saw outflow on w edge of convection
- Better anticipate where new stuff will develop
- Draws me to an area of new updrafts, where should I look
- No question – I will have it loaded when in warning situations, keeping an eye on it
- Good situational awareness display to have the 1min loop loaded
- For training, it would be most helpful to see examples of various phenomenon from the folks who have already been using the data
- As storms were developing, a feeder band was being ingested into the main cell, we were watching it for new development, and stuff did develop along it. Wouldn’t have seen it quite as well with 15 min data
Lightning jump
- Used with probsevere
- Stuff going from nothing to severe fast. 8sigma jump noted
- Sometimes I miss significant jumps
- We would like this to be incorporated into probsevere
- Bill Line, SPC/HWT Satellite Liaison
GOES-R LAP
- Absolute values of CAPE were way off when compared to RAP, SPC mesoanalysis, RAOBS
NUCAPS
- Finer details would still be nice to see in the profiles
GOES-R CI
- Mid to high clouds obscured development in Chicago
- Once developing convection broke through the mid-level cloud deck, 80—90 % probabilities appeared
- Didn’t get many signals today in the Hastings CWA, so wasn’t useful. Thin cirrus were present
ProbSevere
- Our storms were so explosive, once it got close to 50, we would warn from radar because it was going so fast
- I like how if you’re focusing in on one cell (with a feeder band), watching outflow interaction, and then another cell developed nearby. It keeps you aware that you should go look at that now
- Especially valuable when you have lots of stuff on radar
- I could see it being on a SA panel
- Storm coordinator could use to make sure all relevant storms are covered
- I was looking at a storm that was weakening, so I focused my attention elsewhere… looked back a little later and saw the weakening storm was back to 80, telling me that I better take another look … good SA
- If you can’t look at all tilts, its seeing column for you
- New storm popped up when we were focused on other storms; this flagged us, quickly put severe on it
- Towards end of day, outflow boundaries collided, immediately flagged it to pink, then collapsed
SRSOR and 1-min OTD
- In Chicago, you could see where the front was and where cu were developing under a mid level cloud deck. Could see CI as it was occurring below those clouds.
- With new cells forming under cirrus canopy, you could see it bubble right through the anvil blowoff
- Saw outflow on w edge of convection
- Better anticipate where new stuff will develop
- Draws me to an area of new updrafts, where should I look
- No question – I will have it loaded when in warning situations, keeping an eye on it
- Good situational awareness display to have the 1min loop loaded
- For training, it would be most helpful to see examples of various phenomenon from the folks who have already been using the data
- As storms were developing, a feeder band was being ingested into the main cell, we were watching it for new development, and stuff did develop along it. Wouldn’t have seen it quite as well with 15 min data
Lightning jump
- Used with probsevere
- Stuff going from nothing to severe fast. 8sigma jump noted
- Sometimes I miss significant jumps
- We would like this to be incorporated into probsevere
A very strong cornfed storm
Just after 7pm a monster storm developed southwest of Hastings, NE. At 732pm the storm intensified to near 70 dbz up to 32kft! The MESH product showed just over 2″ hail and the RSO showed nice bubbling of the overshooting top on visible satellite. This was definitely picked up by the overshooting top algorithm. Ironically, the lightning was never really all that impressive, but did hold steady. The lightning level really dropped by the time the maximum storm strength was reached.
(New)CAPS post
Pardon my puns. Mr Snow and I were testing out the NUCAPS product now that things have settled down in our DMA. We selected different locations and got different results.
Compared to surface obs, the NUCAPS temperature did not need to be altered, but the moisture did, by about 6 degrees Celsius. This change took my MU parcel cape from 15J/kg to ~2000J/kg. SPC meso had it closer to 1500J/kg. The EL also doubled from 6 to 12 km.
We did have storms just SW of this area, however, as they lifted NE, they weakened.
-Brick Tamland
Compared to surface obs, the NUCAPS temperature did not need to be altered, but the moisture did, by about 6 degrees Celsius. This change took my MU parcel cape from 15J/kg to ~2000J/kg. SPC meso had it closer to 1500J/kg. The EL also doubled from 6 to 12 km.
We did have storms just SW of this area, however, as they lifted NE, they weakened.
-Brick Tamland
Subscribe to:
Posts (Atom)