Friday, May 25, 2012
Tales from the Testbed Webinar
EWP end of week 3 debrief
Well it's halfway through the experiment and we once again got a chance to sit down with the forecasters and pick their brains about their experience over the past week. We got a lot of really good blog posts this week from the forecasters, but we wanted to ask them some more product specific questions. Below is the feedback we captured...
SATCAST / UW Cloud-top Cooling
- Forecasters asked if they could get the CI products via LDM... we said yes, but they will have to ask the regions to get it. UAH CI is already being delivered to southern region, UWCI is also available via LDM to a few regions... both products are also available on the web (see this blog post).
- Is this product still useful in pulse type non-severe storms? "Yes, absolutely... as long as there is no cirrus contaminating."
- "The probabilistic CI will be really useful for aviation purposes since we're not looking at just severe convection."
- "There were some times when using both products at the same time became confusing because the CI would show low probabilities, but the CTC would show some strong signals." ... This is an effect of the products being completely independent of each other.
- "I definitely think running both together provides the most information for us... my biggest observation was that when the CI switched to night time mode, maybe put a threshold on the lower signals to cut out on some of the amount of signals."
- "There were some good cases where the CI would switch from 60% to 80% and then the CTC would flag the storm and we would get some good lead time on the first lightning strike... it was nice to have it all work in tandem."
- "Every storm that went severe yesterday in MO had a -30 C/15 min and every signal that strong lead to a severe storm, so it was a complete success."
- "I did play around with the color table to make it look more like radar and used it a bit, but realized that I didn't really like it... the standard color table worked really well."
Nearcast
- "That was my favorite product... it really seemed to capture the motion of the atmosphere than any of the products we had been looking at."
- "I really liked the 500 mb theta-e product... it really had a clear picture of the gradient of moisture and showed a strong push of cooler/dryer air where storms did not end up forming."
- "It really picked up on an area SE of Denver where the instability was strongest and identified the region where storms formed."
- "It seems like the colors were a little flipped, with the more unstable being the cooler colors... as a forecaster I want to see the warmer colors identify regions I should be paying attention to... red means bad in our field."
- "I'm wondering what this is going to look like in the west when the levels aren't as applicable... are they thinking about adding more levels?" ... Yes
Simulated Satellite
- "The simulated satellite picked up on some of the isolated storms pretty well all week, but the more organized things like cold fronts or the tropical system did not do so well." ... This is more of an issue of the model, not the simulated satellite imagery itself and seems to be a feature of this week only... may be due to the fact that there was some significant capping over much of the US.
- What did you think of using this versus other model fields? "It's really useful, we have to do sky fields in GFE and it would be very helpful to get this in GFE." ... Dan Lindsey volunteered to do this and was given a contact to do so.
- "The low bias in extent of high/anvil clouds was very apparent... is there any effort to fix that, but still keep the ability to keep low cloud info from the model? I think it might be useful to have two separate products, one for the regular imagery and one for low clouds."
- "If you could code the low clouds generated in a IFR, VFR, MIFR, MVFR color table, that would be incredibly useful for aviation." ... This may be more of a suggestion for the low cloud/fog product.
- "It would be nice to have more model runs of this." ... It would be hard to do so because of the computational requirements.
- We mentioned to the forecasters that this product is available via LDM and the forecasters were encouraged to contact Dan Lindsey to get the process started... it is also available on the web (see this blog post).
- "We did notice that the simulated WV was much dryer than the observed." ... This is because the band being simulated is different than the one that is available on current GOES and sees slightly lower in the atmosphere.
- "I stumbled on the band difference and I basically said, 'I don't know what this stuff is' so I backed off." ... This suggests more work needs to be done with forecasters on basic satellite topics like band differences, weighting functions and radiative transfer... a short description was understood very well, so it's not a complexity issue, it's more of a lack of required training.
- "I was using the band difference yesterday to pick out the dry line over much of OK... it was a little difficult to see because of the low clouds over the area."
Sounder Airmass
- "I looked at it twice... it was interesting to see the initiation occurred along the sharp moisture gradients that were associated with the strong shortwave troughs... as well as being very useful for picking out the jet maximum... I really love this product."
PGLM
- Unfortunately another miss this week... no significant weather within range of the LMAs.
Overall / Training
- "When stuff begins firing... I switch away from the models."
- "Any extra lead time you can give on training materials is beneficial... It can be difficult to try and weave that in to your operational shift schedule... It did take more than one day to get through all the material... I had to go through some of the material twice before I got a real grasp on it."
- "The WES was very useful to just play around with the data... the job sheets were helpful, but I just liked playing around with the data."
- "If we came in the first day and tried to cover all this material I would have been totally overwhelmed."
- "Can you condense the presentations? Get right to the point and be concise... you can provide additional written background material that we can go over later."
- "Keeping stuff off a required NOAA domain would help so I could access it at home."
- "I found that I was spending a lot of time the first two days just trying to find the products... anything you can do to have procedures ready would be very helpful." ... This is difficult since we localize over anywhere in the US, so procedures built for one localization will not work for another.
- "If you can be more upfront and tell us 'You should be looking at this product now' would help us not miss things we should be seeing... don't be afraid to interrupt us." ... Maybe have the PIs running a separate workstation and have their displays up on the Situational Awareness Display would help.
SATCAST / UW Cloud-top Cooling
- Forecasters asked if they could get the CI products via LDM... we said yes, but they will have to ask the regions to get it. UAH CI is already being delivered to southern region, UWCI is also available via LDM to a few regions... both products are also available on the web (see this blog post).
- Is this product still useful in pulse type non-severe storms? "Yes, absolutely... as long as there is no cirrus contaminating."
- "The probabilistic CI will be really useful for aviation purposes since we're not looking at just severe convection."
- "There were some times when using both products at the same time became confusing because the CI would show low probabilities, but the CTC would show some strong signals." ... This is an effect of the products being completely independent of each other.
- "I definitely think running both together provides the most information for us... my biggest observation was that when the CI switched to night time mode, maybe put a threshold on the lower signals to cut out on some of the amount of signals."
- "There were some good cases where the CI would switch from 60% to 80% and then the CTC would flag the storm and we would get some good lead time on the first lightning strike... it was nice to have it all work in tandem."
- "Every storm that went severe yesterday in MO had a -30 C/15 min and every signal that strong lead to a severe storm, so it was a complete success."
- "I did play around with the color table to make it look more like radar and used it a bit, but realized that I didn't really like it... the standard color table worked really well."
Nearcast
- "That was my favorite product... it really seemed to capture the motion of the atmosphere than any of the products we had been looking at."
- "I really liked the 500 mb theta-e product... it really had a clear picture of the gradient of moisture and showed a strong push of cooler/dryer air where storms did not end up forming."
- "It really picked up on an area SE of Denver where the instability was strongest and identified the region where storms formed."
- "It seems like the colors were a little flipped, with the more unstable being the cooler colors... as a forecaster I want to see the warmer colors identify regions I should be paying attention to... red means bad in our field."
- "I'm wondering what this is going to look like in the west when the levels aren't as applicable... are they thinking about adding more levels?" ... Yes
Simulated Satellite
- "The simulated satellite picked up on some of the isolated storms pretty well all week, but the more organized things like cold fronts or the tropical system did not do so well." ... This is more of an issue of the model, not the simulated satellite imagery itself and seems to be a feature of this week only... may be due to the fact that there was some significant capping over much of the US.
- What did you think of using this versus other model fields? "It's really useful, we have to do sky fields in GFE and it would be very helpful to get this in GFE." ... Dan Lindsey volunteered to do this and was given a contact to do so.
- "The low bias in extent of high/anvil clouds was very apparent... is there any effort to fix that, but still keep the ability to keep low cloud info from the model? I think it might be useful to have two separate products, one for the regular imagery and one for low clouds."
- "If you could code the low clouds generated in a IFR, VFR, MIFR, MVFR color table, that would be incredibly useful for aviation." ... This may be more of a suggestion for the low cloud/fog product.
- "It would be nice to have more model runs of this." ... It would be hard to do so because of the computational requirements.
- We mentioned to the forecasters that this product is available via LDM and the forecasters were encouraged to contact Dan Lindsey to get the process started... it is also available on the web (see this blog post).
- "We did notice that the simulated WV was much dryer than the observed." ... This is because the band being simulated is different than the one that is available on current GOES and sees slightly lower in the atmosphere.
- "I stumbled on the band difference and I basically said, 'I don't know what this stuff is' so I backed off." ... This suggests more work needs to be done with forecasters on basic satellite topics like band differences, weighting functions and radiative transfer... a short description was understood very well, so it's not a complexity issue, it's more of a lack of required training.
- "I was using the band difference yesterday to pick out the dry line over much of OK... it was a little difficult to see because of the low clouds over the area."
Sounder Airmass
- "I looked at it twice... it was interesting to see the initiation occurred along the sharp moisture gradients that were associated with the strong shortwave troughs... as well as being very useful for picking out the jet maximum... I really love this product."
PGLM
- Unfortunately another miss this week... no significant weather within range of the LMAs.
Overall / Training
- "When stuff begins firing... I switch away from the models."
- "Any extra lead time you can give on training materials is beneficial... It can be difficult to try and weave that in to your operational shift schedule... It did take more than one day to get through all the material... I had to go through some of the material twice before I got a real grasp on it."
- "The WES was very useful to just play around with the data... the job sheets were helpful, but I just liked playing around with the data."
- "If we came in the first day and tried to cover all this material I would have been totally overwhelmed."
- "Can you condense the presentations? Get right to the point and be concise... you can provide additional written background material that we can go over later."
- "Keeping stuff off a required NOAA domain would help so I could access it at home."
- "I found that I was spending a lot of time the first two days just trying to find the products... anything you can do to have procedures ready would be very helpful." ... This is difficult since we localize over anywhere in the US, so procedures built for one localization will not work for another.
- "If you can be more upfront and tell us 'You should be looking at this product now' would help us not miss things we should be seeing... don't be afraid to interrupt us." ... Maybe have the PIs running a separate workstation and have their displays up on the Situational Awareness Display would help.
Web links for products
There have been a lot of requests by the forecasters for the websites for the products being demonstrated... below is a comprehensive list for future reference.
SATCAST
http://nsstc.uah.edu/~jwalker/UAH_SATCASTv2_CI_VIS.png (most recent)
http://nsstc.uah.edu/~jwalker/UAH_SATCASTv2_CI_VIS_m1.png (minus 1)
http://nsstc.uah.edu/~jwalker/UAH_SATCASTv2_CI_VIS_m2.png (minus 2)
http://nsstc.uah.edu/~jwalker/UAH_SATCASTv2_CI_VIS_m3.png (minus 3)
http://nsstc.uah.edu/~jwalker/UAH_SATCASTv2_CI_VIS_m4.png (minus 4)
UWCI
http://cimss.ssec.wisc.edu/snaap/convinit/quicklooks/
Nearcast
http://cimss.ssec.wisc.edu/model/nrc/
Simulated Satellite Imagery / Band Difference
http://cimss.ssec.wisc.edu/goes_r/proving-ground/nssl_abi/nssl_abi_rt_js.html (imagery)
http://rammb.cira.colostate.edu/ramsdis/online/goes-r_proving_ground.asp#Synthetic_GOES-R_Imagery_from_Real-Time_NSSL_4_km_WRF-ARW (imagery/band differences)
Sounder Airmass
http://rammb.cira.colostate.edu/ramsdis/online/goes-r_proving_ground.asp#GOES_Sounder-Based_Products
SATCAST
http://nsstc.uah.edu/~jwalker/UAH_SATCASTv2_CI_VIS.png (most recent)
http://nsstc.uah.edu/~jwalker/UAH_SATCASTv2_CI_VIS_m1.png (minus 1)
http://nsstc.uah.edu/~jwalker/UAH_SATCASTv2_CI_VIS_m2.png (minus 2)
http://nsstc.uah.edu/~jwalker/UAH_SATCASTv2_CI_VIS_m3.png (minus 3)
http://nsstc.uah.edu/~jwalker/UAH_SATCASTv2_CI_VIS_m4.png (minus 4)
UWCI
http://cimss.ssec.wisc.edu/snaap/convinit/quicklooks/
Nearcast
http://cimss.ssec.wisc.edu/model/nrc/
Simulated Satellite Imagery / Band Difference
http://cimss.ssec.wisc.edu/goes_r/proving-ground/nssl_abi/nssl_abi_rt_js.html (imagery)
http://rammb.cira.colostate.edu/ramsdis/online/goes-r_proving_ground.asp#Synthetic_GOES-R_Imagery_from_Real-Time_NSSL_4_km_WRF-ARW (imagery/band differences)
Sounder Airmass
http://rammb.cira.colostate.edu/ramsdis/online/goes-r_proving_ground.asp#GOES_Sounder-Based_Products
Thursday, May 24, 2012
UW-CTC prompts experimental warning with great success in NW MO
The UW-CTC product, again, showed persistent signals of strong cooling from 2140 UTC to 2210 UTC for a storm developing within NW MO. The following images show the sequence of cooling rates for this cell:
 |
| 2140 UTC |
 |
| 2145 UTC |
 |
| 2155 UTC |
 |
| 2202 UTC (Not AWIPS screen capture due to AWIPS-2 technicality) |
 |
| 2210 UTC |
 |
| Official EAX warning text |
 |
| SPC Storm Reports (see NW MO) |
| This development was noticed by a forecaster paying attention to the EAX WFO here at the HWT. Based on the environment and the observed CTC rate of < -50K/15 min, the forecaster issued an experimental warning for this cell. The actual radar-based EAX severe thunderstorm warning was not placed until 2254 UTC, meaning with the added experimental information, the HWT forecaster had a issuance lead time of 44 minutes over the official warning (see above for official warning text from EAX)! Additionally, this particular cell had severe hail and wind reports with the hail reaching 1" criteria at 2305 UTC (see above reports). This means that the UW-CTC detection had an hour and 25 minute lead time over the observed hail report. |
UAHCI along cold front in Kansas
Over the past several hours, several storms have fired along the cold front in Kansas and subsequently moved off towards the northeast. This area continues to be defined by a thin line of convergence at 2255 UTC (see Figure 1). The UAHCI product continues to indicate the storms will continue to fire along this portion of the front with a stronger strength of signal on the northern end. Subsequent visible imagery shows that the northern end of this line segment did indeed generate more convection as can be seen at 2345 UTC with the mature cumulonimbus in Figure 2. The southern end of the line continues to show strong strengths of signal, with future development likely in the unstable environment.
 | ||||
| Figure 1 |
 |
| Figure 2 |
Sea breeze along W-coast of Florida
Forecaster comments from EWP blog...
The synthetic IR/WV forecast did a good job in placing strongest
activity over Florida along the W coast, where eastward moving sea
breeze probably caused strongest convergence in easterly wind regime.
Pulsating storms featured reflectivity peaks in excess of 50 dBz.
Harrison County MO CTC Lead Time
Forecaster comments from EWP blog...
The following CTC image led to the issuance of a warning at 2028Z. Severe hail signatures were noted with storm at 2112Z and severe hail was reported at 2130Z.
The following CTC image led to the issuance of a warning at 2028Z. Severe hail signatures were noted with storm at 2112Z and severe hail was reported at 2130Z.
RSO mimics GOES-R temporal resolution for UAHCI
Away from the main activity occurring in the Upper Midwest (previous blog posts), thunderstorms are also developing in association with the upper low. The UAHCI algorithm has captured this development over Nebraska. Severe weather is not anticipated with this activity. GOES is currently in Rapid Scan Operations, so this will closely mimic the temporal resolution of GOES-R. The first image is from 2032 UTC with SOS values over 70.
This second image is from 2040 UTC, with a strength of signal over 80.
The last image is from 2045 with a strength of signal continued over 80. The storms went onto initiate over central Nebraska! The temporal resolution of GOES-R will surely help convective initiation algorithms.
This second image is from 2040 UTC, with a strength of signal over 80.
The last image is from 2045 with a strength of signal continued over 80. The storms went onto initiate over central Nebraska! The temporal resolution of GOES-R will surely help convective initiation algorithms.
Continued Development throughout IA/MO Captured by UW-CTC
Continued convective development has been occurring throughout IA/MO over the past hour. UW-CTC has been consistently detecting strong cooling within the cloud tops of the developing storms. The following sequence of images (valid 1955 UTC through 2015 UTC) shows the consistent nature of the detections for this case by the UW-CTC algorithm:
A fellow colleague has stated that these detections were not only ahead of the warnings associated with the storms, but also ahead of the severe watches that were issued. The two following images show the watch areas issued for the region. The first watch, which is a Tornado Watch for MN/WI area was first issue at 1925 UTC. UW-CTC had multiple detections in the MN area prior to the watch being issued in the MN/WI domain.
The second watch area, a Severe T-Storm Watch for S.W. WI, IA, and western IL, was issued at 2035 UTC. Again, the UW-CTC product had multiple strong detections within this domain prior to the watch being issued.
Perhaps the UW-CTC could aid in situational awareness for watch issuance, too?
 |
| Could not get AWIPS 2002 UTC image due to technical issues with AWIPS-2 |
A fellow colleague has stated that these detections were not only ahead of the warnings associated with the storms, but also ahead of the severe watches that were issued. The two following images show the watch areas issued for the region. The first watch, which is a Tornado Watch for MN/WI area was first issue at 1925 UTC. UW-CTC had multiple detections in the MN area prior to the watch being issued in the MN/WI domain.
Impressive Cld top cooling rate, Iowa
Forecaster comments from EWP blog...
Severe convection currently evolves along the cold front over C Iowa with Cld top cooling product showing impressive cooling rates below -65 K/15 min e.g. over Warren and Polk counties.
Severe convection currently evolves along the cold front over C Iowa with Cld top cooling product showing impressive cooling rates below -65 K/15 min e.g. over Warren and Polk counties.
SimSat has problems with tropical low
Forecaster comments from EWP blog...
A weak tropical low moved from Cuba to the NE during the daytime hours and was forecast to affect far SE Florida. Comparing SimSat and real IR/WV data, significant differences can be seen especially regarding the cirrus canopy. The main reason for the differences may be the much stronger development of this depression than expected 21 h ago. NHC also increased probs to 40 %, indicating a consolidating system. This evolution may have resulted in a better defined depression’s center with stronger convective development. Therefore those products may have to be used carefully for tropical lows, which reveal stronger intensification rates compared to what models forecast.
A weak tropical low moved from Cuba to the NE during the daytime hours and was forecast to affect far SE Florida. Comparing SimSat and real IR/WV data, significant differences can be seen especially regarding the cirrus canopy. The main reason for the differences may be the much stronger development of this depression than expected 21 h ago. NHC also increased probs to 40 %, indicating a consolidating system. This evolution may have resulted in a better defined depression’s center with stronger convective development. Therefore those products may have to be used carefully for tropical lows, which reveal stronger intensification rates compared to what models forecast.
Iowa CTC Lead Time
Forecaster comments from EWP blog...
The cloud top cooling product in central Iowa indicated nearly 40 C/15 km of cloud top cooling for a storm in Boone and Story counties. This product preceded intense radar reflectivity in the area by nearly an hour.
The cloud top cooling product in central Iowa indicated nearly 40 C/15 km of cloud top cooling for a storm in Boone and Story counties. This product preceded intense radar reflectivity in the area by nearly an hour.
CI north of Kansas City
CI has been occurring along a cold front stretching from the surface low over the upper Midwest. The UAHCI has been producing Strength of Signals values over 60 for many of the southern storms that have begun developing in Iowa and Missouri. This example of the UAHCI at 1915 UTC occurs near in Northeast Kansas, yet with the fast flow, the cell did not initiate until just south of the Iowa border in Missouri. The UAHCI algorithm produced approximately 20 minute lead time on this particular cell considering satellite scan time and algorithm run time. The radar imagery in the second image is from 1951 UTC.
MLB-CI/CTC Example
Forecaster comments from EWP blog...
Here is an example of how UAH Convective Initiation (CI) and CIMSS Cloud Top Cooling (CTC) can be used to anticipate convective development. Here in the first image at 1940Z, let’s look at the CI product in the top left panel:
Although this did not ultimately become a severe thunderstorm, we did
observe a few lightning strikes shortly after this time. This
progression from strong CI signal to strong CTC can be a good indicator
of a developing thunderstorm.
Here is an example of how UAH Convective Initiation (CI) and CIMSS Cloud Top Cooling (CTC) can be used to anticipate convective development. Here in the first image at 1940Z, let’s look at the CI product in the top left panel:
CI (top left panel) at 1940Z.
Focus on the Lake/Marion/Sumter County line. There is a maximum in
CI there of 80, indicating likely convective development. If we look at
the CTC in the same area on the next image (top right panel), we see a
developing max CTC over -20C/15min, which is a strong indicator of a
developing severe thunderstorm.
1945Z CTC (top right panel).
CI along sea breeze over NE Florida
Forecaster comments from EWP blog...
Westward moving sea breeze front helped to spark sporadic showers/thunderstorms during the afternoon hours. We captured the interaction of an already ongoing weak shower, as front moved in from the east.
The southern parts of Nassau and Charlton county and Baker county were monitored. At 1932Z an increase in cloud top cooling was noticed and 8 min later, CI product also had probabilities of 70 % plus for this area. At the same time, a cooling rate of roughly -15K/15 min was detected. Rapid development occurred thereafter with 50 dBz and more seen in the JAX radar side at 1947 Z, which gave a good lead time for focusing on the potential hot-spot over NE Florida. The storm was a pulsating one with rapid weakening thereafter.
Westward moving sea breeze front helped to spark sporadic showers/thunderstorms during the afternoon hours. We captured the interaction of an already ongoing weak shower, as front moved in from the east.
1932 Z; UAH-CI and Cloud top cooling
1940 Z; UAH-CI and Cloud top cooling
The southern parts of Nassau and Charlton county and Baker county were monitored. At 1932Z an increase in cloud top cooling was noticed and 8 min later, CI product also had probabilities of 70 % plus for this area. At the same time, a cooling rate of roughly -15K/15 min was detected. Rapid development occurred thereafter with 50 dBz and more seen in the JAX radar side at 1947 Z, which gave a good lead time for focusing on the potential hot-spot over NE Florida. The storm was a pulsating one with rapid weakening thereafter.
UAH-CI Signal Precedes UW-CTC Signal…Both Precede Convective Initiation
Forecaster comments from EWP blog...
The more I look at these products, the more I’m convinced they work very well in tandem with each other. Given an environment that supports convective growth, a strong UAH-CI signal (upper image, top left panel) typically precedes a strong UW-CTC signal (middle image, top right panel). In this case, the strong UAH-CI signal (1615 UTC) preceded the UW-CTC signal (1645 UTC) by about 30 minutes. By 1715-1730 UTC, 30-45 minutes later, deep-moist convection did indeed form as evidenced by the numerous CG strikes (bottom image, bottom left panel). We’ve seen this all week…UAH-CI gives us a heads up that general convection is initially building, then a short time later UW-CTC typically picks up on the stronger convection building. Given a strong cap isn’t present and instability is sufficient, deep-moist convection typically develops 30-60 minutes thereafter.
The more I look at these products, the more I’m convinced they work very well in tandem with each other. Given an environment that supports convective growth, a strong UAH-CI signal (upper image, top left panel) typically precedes a strong UW-CTC signal (middle image, top right panel). In this case, the strong UAH-CI signal (1615 UTC) preceded the UW-CTC signal (1645 UTC) by about 30 minutes. By 1715-1730 UTC, 30-45 minutes later, deep-moist convection did indeed form as evidenced by the numerous CG strikes (bottom image, bottom left panel). We’ve seen this all week…UAH-CI gives us a heads up that general convection is initially building, then a short time later UW-CTC typically picks up on the stronger convection building. Given a strong cap isn’t present and instability is sufficient, deep-moist convection typically develops 30-60 minutes thereafter.
1615 UTC. Notice the 70% UAH-CI signal in the top left panel.
1645 UTC...30 minutes later. Notice strong cloud-top-cooling signal in the UW-CTC product (top right panel).
1715-1730 UTC...30-45 minutes later. Deep-moist convection did indeed
develop in the vicinity of aforementioned CI/CTC signals, as evidenced
by numerous CG lightning strikes (bottom left panel).
UW-CTC Working as Expected in MN/IA
The UW-CTC product has been capturing the initial and ongoing convective development taking place first within the MPX CWA domain, and later further south in north central to central IA. The beauty of these events is that this showcases how the UW-CTC product still does pretty well, even though the area has a lot of apparent cloud contamination within the scenes. The below images are shown in chronological order with the original strong convection beginning in the MSPL area at 1515 UTC (Fig. 1).
Later on at 1725 UTC, the UW-CTC captured good development in the MSPL area that later is an ongoing robust thunderstorm (Fig. 2).
Some further strong convective development along the frontal boundary was noted in northern and central IA. Figs. 3 (1932 UTC) and 4 (1955 UTC) capture this development unzipping further south along the boundary and still within areas of significant amounts of cloud debris.
 |
| Fig. 1 |
 |
| Fig. 2 |
 | |
| Fig. 3 |
 |
| Fig. 4 |
CI mainly inland across FL peninsula
Forecaster comments from EWP blog...
Latest CI image shows moderately high probabilities for CI along ahead of the westward moving sea-breeze boundary. Best chance for thunderstorms appears to be from the interior counties towards the west coast. Some initiation is possible across Northern Brevard County…though the prospects for lightning are looking slim at this point.
Latest CI image shows moderately high probabilities for CI along ahead of the westward moving sea-breeze boundary. Best chance for thunderstorms appears to be from the interior counties towards the west coast. Some initiation is possible across Northern Brevard County…though the prospects for lightning are looking slim at this point.
Overview of sea breeze fronts in Florida
Forecaster comments from EWP blog...
We’re currently monitoring the rapidly inland moving see breeze front over NE Florida with enhanced TCU present along the convergence zone, whereas the west coast see breeze has trouble to move inland within dominant easterly wind regime. Both convergence zones enter an area over N-C Florida where slightly lower dewpoints (lower to mid 60s) are present. With ongoing diabatic heating, cap continues to weaken or has already vanished with isolated initiation possible over N-C Florida in the following hours but ongoing large-scale weak subsidence seems to delay initiation a bit. HRRR quite aggressive with initiation mainly along the western sea breeze front with RUC also showing sporadic storms over N-C Florida. Overall set-up is not impressive and favors only the development of a few thunderstorms. Moderate CAPE and weak shear environment may result in a few strong and slow moving storms.
We’re currently monitoring the rapidly inland moving see breeze front over NE Florida with enhanced TCU present along the convergence zone, whereas the west coast see breeze has trouble to move inland within dominant easterly wind regime. Both convergence zones enter an area over N-C Florida where slightly lower dewpoints (lower to mid 60s) are present. With ongoing diabatic heating, cap continues to weaken or has already vanished with isolated initiation possible over N-C Florida in the following hours but ongoing large-scale weak subsidence seems to delay initiation a bit. HRRR quite aggressive with initiation mainly along the western sea breeze front with RUC also showing sporadic storms over N-C Florida. Overall set-up is not impressive and favors only the development of a few thunderstorms. Moderate CAPE and weak shear environment may result in a few strong and slow moving storms.
Neither CI nor Cld top cooling products showing robust signs for
initiation … mainly over C Florida. Both products however show more
vivid activity offshore over the Atlantic, where mid-levels remain a bit
cooler.
1800Z VIS and station data.
1745Z; UAH-CI (upper left) and CIMSS Cld Top Cooling (upper right).
Southeast US still providing areas of CI
The pesky upper low over the Southeast US continues to move offshore. With strong surface heating and "cool" mid-level temperatures with this low, airmass thunderstorms continue to develop over the Southeast US. UAHCI has been indicating many areas with Strength of Signals over 60 from the Southeast up into the Mid-Atlantic States. Many of these storms will not become severe due to the surrounding synoptic environment, yet will produce heavy rainfall and frequent lightning with the stronger storms. This example of UAHCI tracks a slow developing storm in Washington County, GA. These two images are from 1632 UTC and 1645 UTC with Strength of Signals (SOS) of 52 and 57 respectively, suggesting only slow growth at the time.
At 1702 UTC, the SOS ramped up to 73, suggesting more rapid growth and increasing confidence that this cell will actually CI.
At 1715 UTC, the SOS increases further to 83. The UAHCI algorithm stopped tracking the object after this time period as it was deemed mature within the cloud mask, which is evident within the visible imagery at 1725 UTC.
At 1702 UTC, the SOS ramped up to 73, suggesting more rapid growth and increasing confidence that this cell will actually CI.
At 1715 UTC, the SOS increases further to 83. The UAHCI algorithm stopped tracking the object after this time period as it was deemed mature within the cloud mask, which is evident within the visible imagery at 1725 UTC.
Outlook 24 May 2012: Week 3, Day 4
From the EWP blog...
Operating today during the first mod risk of the experiment, however, that is not where we are operating today… at least not to start.
In order to try and get lightning data into the experiment this week, we’ve initialized in the Sterling, VA and Melbourne, FL domains.
For the FL domain, we’re hopeful that storms will get going as the sea breeze front sets up across the state. Already, the UAH-CI product is flagging cumulus for development (consistently 50-60%) along the line from Sanderson to -Orlando to just northwest of Okeechobee.
For the DC domain, the CI alg is also flagging development generally a bit lower on the probability though. A region of SBCAPE of >1500-2000 J/kg extends from eastern NC to over much of the LWX CWA, so it appears possible that we could at least see some convective development across some of the DC LMA though likely only marginally severe.
Our observers and some PIs are keeping an eye over central WI and the associated moderate risk region for development of surface based storms (as to now all development across Minnesota appears to be elevated). Currently instability is limited across much of the region, but we do expect additional destabilization with daytime heating and enhanced lift as the cold front moves in.
Operating today during the first mod risk of the experiment, however, that is not where we are operating today… at least not to start.
In order to try and get lightning data into the experiment this week, we’ve initialized in the Sterling, VA and Melbourne, FL domains.
For the FL domain, we’re hopeful that storms will get going as the sea breeze front sets up across the state. Already, the UAH-CI product is flagging cumulus for development (consistently 50-60%) along the line from Sanderson to -Orlando to just northwest of Okeechobee.
UAH-CI & Visible Satelllite for central FL @ 1830 UTC on 24 May 2012.
For the DC domain, the CI alg is also flagging development generally a bit lower on the probability though. A region of SBCAPE of >1500-2000 J/kg extends from eastern NC to over much of the LWX CWA, so it appears possible that we could at least see some convective development across some of the DC LMA though likely only marginally severe.
Our observers and some PIs are keeping an eye over central WI and the associated moderate risk region for development of surface based storms (as to now all development across Minnesota appears to be elevated). Currently instability is limited across much of the region, but we do expect additional destabilization with daytime heating and enhanced lift as the cold front moves in.
EWP daily debrief 5/24
Yesterday we had a good day in operations for the CI products especially. Below is some feedback we gathered from the forecasters during our daily debrief...
UAH SATCAST (CI) / UW Cloud-top Cooling (CTC)
- "They were very helpful... it was environment you knew the cloud would cool quickly, but you didn't know which clouds would go up... I think it's benefit is really in that marginal environment... discriminating between that severe/non-severe borderline is where we really need help."
- "Really all the satellite based products would be useful for us in terms of aviation forecasting... if we could get an hour lead time on CI, it would be really helpful to us... especially in a place like NY where just one thunderstorm will mess up everything with regards to air traffic... just to see the CI and the CTC, you can update your TAF and contact the proper folks and give them a more confident message than if you're just using radar... with just radar we can't give them the lead time they need."
- "When we were deciding when to switch, it seemed like the CI was showing a widespread 30-50% for that whole area, but it didn't really happen." ... Developer explained that sometimes many cumulus clouds close together can become clustered together (especially at night when the visible channel isn't utilized), which will give you relatively lower probabilities over a broader area.
- "That region of Cu was a strongly forced/capped area... that broad view of the CI product gave you an idea that there is some upward motion and told you to pay attention."
- "I really loved the CI products yesterday... seeing the CI and CTC zippering down the line, we actually warning on one CTC signal when we saw that everything else ahead of it was going."
Overall / Training
- "With what you show the forecasters, simpler is better."
UAH SATCAST (CI) / UW Cloud-top Cooling (CTC)
- "They were very helpful... it was environment you knew the cloud would cool quickly, but you didn't know which clouds would go up... I think it's benefit is really in that marginal environment... discriminating between that severe/non-severe borderline is where we really need help."
- "Really all the satellite based products would be useful for us in terms of aviation forecasting... if we could get an hour lead time on CI, it would be really helpful to us... especially in a place like NY where just one thunderstorm will mess up everything with regards to air traffic... just to see the CI and the CTC, you can update your TAF and contact the proper folks and give them a more confident message than if you're just using radar... with just radar we can't give them the lead time they need."
- "When we were deciding when to switch, it seemed like the CI was showing a widespread 30-50% for that whole area, but it didn't really happen." ... Developer explained that sometimes many cumulus clouds close together can become clustered together (especially at night when the visible channel isn't utilized), which will give you relatively lower probabilities over a broader area.
- "That region of Cu was a strongly forced/capped area... that broad view of the CI product gave you an idea that there is some upward motion and told you to pay attention."
- "I really loved the CI products yesterday... seeing the CI and CTC zippering down the line, we actually warning on one CTC signal when we saw that everything else ahead of it was going."
Overall / Training
- "With what you show the forecasters, simpler is better."
Wednesday, May 23, 2012
Nighttime UAHCI Success
As daylight begins to wane along the East Coast, the UAHCI algorithm transitions to a nighttime cloud mask that no longer depends on visible imagery. With the lack of visible imagery, the algorithm appears much more blocky in nature due to the decrease in spatial resolution (1 km visible to 4 km infrared). However, we are able to still track cloud objects that are not stratiform in nature. In this case over Clay County in Southeast South Dakota into Dixon County in Northeast Nebraska, the UAHCI Strength of Signal reached 92 on cells behind the main line of storms. Although not severe, the cell did develop with reflectivity values over 40 dBZ. It is important to note that the UAHCI product is not meant for the predictability of severe storms, but any type convection.
CI product over E-Nebraska
Forecaster comments from EWP blog...
The cold front over NC-Kansas and SE Nebraska was monitored for further development. As expected, storms began to build SW-wards towards the KS boarder and CI products were effective in highlighting the area with highest probabilities for potential development along that boundary.
The event occurred over Clay and York counties (Nebraska). A series of images will be added to show the evolution
Roughly 30 min before surface reflectivity at -10C isothermal level revealed first signs of development, the CI product highlighted both counties with 50-60 % probabilities. The cloud top cooling product showed values of -15 to -25K/15 min at 2315Z (not shown) and the first significant reflectivity signal (56 dbZ ) became visible at 2330 Z. At 00Z onwards a 63 dBz cell evolved out of that NE of York.
The cold front over NC-Kansas and SE Nebraska was monitored for further development. As expected, storms began to build SW-wards towards the KS boarder and CI products were effective in highlighting the area with highest probabilities for potential development along that boundary.
The event occurred over Clay and York counties (Nebraska). A series of images will be added to show the evolution
UAHCI product upper left, CIMSS Cld top cooling upper right, lightning
data lower left and surface reflectivity at -10 C isothermal level at
2232 Z
UAHCI product upper left, CIMSS Cld top cooling upper right, lightning
data lower left and surface reflectivity at -10 C isothermal level at
2245 Z
UAHCI product upper left, CIMMS Cld top cooling upper right, lightning
data lower left and surface reflectivity at -10 C isothermal level at
2302 Z
Roughly 30 min before surface reflectivity at -10C isothermal level revealed first signs of development, the CI product highlighted both counties with 50-60 % probabilities. The cloud top cooling product showed values of -15 to -25K/15 min at 2315Z (not shown) and the first significant reflectivity signal (56 dbZ ) became visible at 2330 Z. At 00Z onwards a 63 dBz cell evolved out of that NE of York.
Subscribe to:
Posts (Atom)