CI With Radar

Couldn't tell if there was something weird with the CI probs...it looked like the storms had already developed.  Unless it picked up that more towering CU was developing right next to the new storms.  Which in this case it seems possible as the storms were building off each other and morphing into a line.

-Penny Gardens

*Later note. I tried using this to overlap outflow boundaries with the CI or CI Severe to see if any CU developed along the boundary.  Didn't see anything, but that may be because of the cirrus.
And is there a probability tool for a gust front? I'm always reminded of the Indiana State Fair situation.  I kept seeing gust fronts ahead of our storms today...however the peak wind seemed to be only about 30 mph.  

CI Probability: Some questions (and another potential use)

We've evaluated CI Probability products this week, with some mixed results, but that has me thinking (which, as my wife will attest, is usually not a good thing)...

Have CI developers noted any correlations between the % shown and how quickly cells with reflectivity values > 35 dBZ develop?

One of my co-evaluators noted a small area of 90% CI Prob yesterday and within a few minutes, a strong cell developed.

I have observed several areas where there were consecutive CI Prob % in the 40-70 range within which convective cells developed, but that process took a little longer, closer to 20-35 mins.

It may seem intuitive that a 90% CI Prob area would generate convective cells quicker than an area where the CI Prob potential is in the 40-70% range, but is this really the case? Maybe this is one use for the CI Prob product that could/should be explored further.

And, how long does an area need to be highlighted with a decent (50%) CI Prob before a forecaster should be confident enough in the development of a cell with reflectivity values > 35 dBZ? Our evaluation group hasn't been able to answer that question during this week of testing.

- Thomas Bell

Moderately High Severe CI Prob for consecutive time periods...did intense convection initiate in the area highlighted?

Across E-cntrl NE, the Severe CI Prob had an area of > 60% for several consecutive time periods over what looked like developing convection. Stepping ahead in time a bit, severe convection did not appear to develop within the convective clouds that the Severe CI Prob product identified as having a pretty decent potential of becoming severe. However, intense storms did develop at the far southern extent of the "general area" that the Severe CI Prob product identified as having a > 60% severe CI potential. The jury is still out regarding out regarding how useful the Severe CI Prob product is.

2002Z:

2007Z:

2012Z:

2107Z:

- Thomas Bell

Analyzing CI Via Multiple Datasets

[20:30 UTC] Convective initiation has begun across central and eastern portions of the Billings CWA. Decided to take a look at some low-level WV imagery in conjunction with the CI product to monitor development underway in extreme northern portions of WY (in an area of poor radar coverage) as of 20:00 UTC. These supplemental datasets are very valuable in areas where radar coverage is sparse and/or non-existent.

Fig. 1: Marked increase in low level WV being shown on Ch. 10 that is associated with agitated cu field blossoming in northern WY.

Fig. 2: The day cloud convection RGB showing the cu field becoming more agitated with a gradual transition from liquid to ice phase as the development increases

Fig. 3: CI shows this cluster as a 40% of convective initiation (which is by far the highest of any "area" within the region of developing cu field). Initiation appears to be just starting following the capture of the above image.

Rosie Red

Severe CI

I found a 95% value on the severe CI! This jumped out in the sea of blue pixels, since it was color coded red.  This developing convection is under a watch area, and will be moving our way in a few hours.  Looking at the visible you can see this area is prime for our LI & CAPE values to become ripe for severe storms as the storms are moving into the clear sky area.

-Penny Gardens

Analyzing The Situation For Billings

I loaded the GLM data, visible & CI overlapping in addition to the severe CI.  The severe CI probabilities are 1% so I'll revisit that product in an hour or so.  The CI clusters have a few areas 40-50%.

This area in SW Montana does correlate with the HRRR model showing convection developing, then rotating around an area low low pressure.  There is a little bit of shear, lapse rates are marginal.  The SPC has a slight risk for hail & wind.  The tornado threat is just southwest of our WFO.  

I'm digging into the RGBs which are slow to load.  But I used the airmass RGB to see the location of the low.

-Penny Gardens

Severe CI Prob vs CI Prob in central OH

Just took a look at an area where some convection developed across central OH. The Severe CI Prob product indicated a low potential for severe TS development, which appears to have been the case. The CI Prob product indicated > 50% chance of CI for 3-4 consecutive time periods. Convection did develop in the area, which, interestingly looks like it was in an area of outflow boundary convergence (at least, it looked like that to me on the vis sat imagery I used in conjunction with the CI products).

2002Z (note the somewhat continuous, but still splotchy, areas in green, especially S of the broken line of convection):

2007Z:

2012Z:

2037Z (A few cells with reflectivity values > 50 dBZ have developed within the "area" that the CI Prob highlighted as > 50% for more than 2 consecutive time periods):

 - Thomas Bell

Developing Storm Using CI

I noticed at 20:17z the cluster of CU showed up as green for development.  On radar the returns were green and low but showed some development.  When the CI went away in the next frame, we had yellow returns on radar...then it grew to a t-storm by 21:38z.

Lightning started showing up just after the CI fields disappeared. 

Severe CI

At first glance I used severe CI to see where the best chance for a severe storm may develop.  But further investigation showed that I misinterpreted what it actually meant.
When I scrolled over the value, it was blue and read 1%.  I thought blue could be 10-20% chance, but in this case it was very low.   That 1% chance is too low to plot, in my opinion.  The color table had no change from 0% to 30% so maybe get rid of everything before 30%?  Trying to filter out the noise and the unnecessary data.

-Penny Gardens

Observations Of Products

ProbSevere-
The color table is hard to follow...probability for hail went from 71% to 61%.  It wasn't obvious that it went down just by glancing.  I looked at the color table at the top and rolled the mouse over it to see what happened.  

CI-
The colors jumped to yellow quickly, but would go away with the next scan.  Not sure what caused the spike.  A lot of noise and clutter with the lower values included. 

Mid-level clouds are picked up on the Water Vapor AGB.  We had to reference our "cheat sheet" to figure out that's what we were seeing.  It looked odd and jumped right off the imagery in the dry area.  Compared to the CI it had jumped to increased chance for convection to occur there.

*Side note on the Water Vapor...nice imagery! Easy to use to see where the dryline was located as there were barely any METAR stations around that area. The water vapor was obvious with the color table.  I could immediately tell where the dry air was located compared to the moisture.  The storms were trying to develop in the drier air mass...

 And just discovered the severe CI product.  This is a nice way to hone in on where ingredients come together for severe storm development.  The areas matches where there are mid-level clouds rapidly developing.  This area also  had a rapid increase in CAPE.




-Penny Gardens

ABQ - First look at CI Probs

2145Z: Latest HRRR has backed off on convection this evening over southern NM, with the ongoing clusters north of ABQ likely remaining the main event. However, the last couple HRRR runs have suggested a lone supercell tries to develop across southern NM so decided to analyze the CI probability & Severe CI probability across this area.

The first issue was deciding which GOES satellite to use, east or west. Obviously I would prefer to use the higher spatial & temporal resolution with GOES 16, but the ABQ WFO is getting near the edge of the GOES 16 disk. Data is available in AWIPS for the astern half of the CWA (east of I-40) but I don't know how much I can trust the data this far from the nadir.

Looping through about an hour of data I see a lot of sporadic jumps in probabilities for both the CI & Severe CI products. On the Severe CI product gif below, values are generally low around 10% but the product will often jump up to 50% or even 100% & back to 10% from frame to frame. While the values overall remain low & no increasing trend is evident, the sporadic increases in data lead to lower confidence & trust in the data. Similar but less extreme jumps in values were evident in the baseline CI product. I did notice a slight trend in increasing CI probabilities but we're only talking from 10% to 30-40% so I'm no longer expecting much in the way of development across southern NM this afternoon. I really like viewing the CI data overlaid on GOES-16 Channel 2 visible imagery to better identify the cu clusters that may be the next to initiate.

I also looked at the CI & Severe CI probs from GOES West. It's tough going back to the coarse resolution & 15-minute imagery of GOES-15 when you've had a taste of the new & improved GOES-16. In our current GOES-16 world I guess I'm just too impatient to wait 15 minutes for data & it was hard to zero in on favored areas of cu development with the coarse resolution "blobs" . The values for CI & severe CI generally matched up well with the values from GOES-16 so that bodes well for the tuning of the GOES-16 product.

Initial thoughts about the CI Probability Product

The CI Probability product is a bit of a "noisy" display with a lot of seemingly "random" areas of CI Probability percentages that percolate up and down, disappear, then reappear, etc., but...

If one filters through the noise and pays attention to areas where 2-3 (or more) consecutive frames display a CI probability of ~ 50-70% (or higher), it seems that convection with reflectivity values of at least 35 dBZ is more likely to form in those areas.

Maybe it would help to get rid of values below 50% on the color/percentage bar? That would filter a lot of the noise and make it a more useful product.

Also, perhaps using a different time-step (is 5 min too long or too short? I don't know...) or tweaking the algorithm a bit might help distinguish those areas more likely to experience convective initiation.

- Thomas Bell

Severe CI Evaluation - ABQ

[19:00 UTC) As a forecaster for WFO ABQ, certainly the attention of the day has already been drawn to the ongoing convection in the western and southern portions of the CWA. However, in review of the various GOES-East imager suite of products (GOES-West temporal scales were a bit coarser and thus not ideal for rapidly-evolving situations like this case), decided to take a look at the Severe Convective Initiation (CI) Probability data. While the domain of data doesn't quite stretch far enough west to take a look at ongoing strong to severe convection in the western and southern parts of the area, did notice that the Severe CI probability was a bit jumpy with some fair weather cu in my east (with thin cirrus overhead, too).

 

Fig. 1: GOES-East Imager Severe CI Probabilities from 18:17z to 18:32z

Fig. 2: GOES-East Channel 2 Vis from 18:17z to 18:32z 

Certainly this product will provide immense value when there is some temporal continuity, but in this case, its noisiness (jumping from 0% to 77% and back to 0% in consecutive frames) was disregarded for the time being. Will hope to see better temporal continuity for improved interpretation and operational utilization. 

Rosie Red

CI Probs Along WY-SD Border

Another example of the GOES-16 general convective initiation probabilities. Loop below shows CI probs overlaid on VIS imagery. From 1700Z to 1900Z, storms developed N-S along the WY-SD border. Prior to deep convection developing, the CI probs were no higher than about 30 percent. Seems like they should be higher than that. A. Cope

CI/CI-Severe in West Texas Convection

I was able to watch the CI and CI-Severe products in some convection that developed over Glasscock county just west of the San Angelo CWA.

This image was taken at 2147Z on Wednesday, with CI in the upper left, CI-Severe upper right, Goes-16 Visible lower right, and SJT 0.5 degrees reflectivity lower left. While satellite has started to show buildups in the cumulus field in the county, no precipitation or lightning has developed. CI is in the 30-50% range over southeastern parts of the county, while CI-Severe was higher for the central area.

At 2157Z, precipitation developed over the northwestern part of the county, with low CI probability over the eastern part. The CI-Severe jumped up to 80%+ over the northeastern part of the county.

By 2212Z, 4 areas of precipitation had developed. Some lightning was occurring withe the easternmost and northwesternmost cells. Since convection was occurring, the CI and CI-Severe dropped almost completely.

The CI product was useful for anticipating where the showers and thunderstorms would develop, but did not provide a clue as to their intensity. All of the cells were producing lightning by around 2245Z.

Gary Cannalte

Convective Initiation over Arkansas

Looking at another example of the Convective Initiation Product over western Arkansas, from earlier this afternoon. The loop below shows the general CI product overlaid on the "sandwich" VIS/IR imagery, for the period approximately 18Z to 19Z. Looking at the cu or towering cu field over SW AR, the CI values vary considerably from one time to the next. Values are generally around 50 percent or less (green/blue), but sometimes jump up to 60-70 percent (yellow). Thunderstorms did develop over this area shortly after 21Z.  A. Cope

CI regular vs severe

I have noticed on severe occasions where the CI (for regular convective initiation) is much lower than the probability depicted by the Severe CI product.  Intuitively, one would think the probability for convective initiation should ALWAYS be higher than the probability for convective initiation to lead to severe weather.

Fig 1. "Regular" CI product at 2037 UTC. Note the green 50% CI probability toward central AR.

Fig 2. Severe CI product at 2037 UTC. There is a 90% probability for severe convection in the same location as the above image.

wall_cloud

Erratic CI - Especialy CI Severe

Noticed that the CI probabilities were jumpy with some convective development. This was especially noticeable with the CI Severe, which jumped from less than 30% to 90% or higher, before coming right back down. The regular CI also had jumps, but they were much less dramatic.

CI Severe between about 2130Z & 2300Z

CI between about 2130Z & 2300Z

Alex Brown

CI Product Noise

The CI product has been a little noisy this afternoon along the southern edge of the QLCS. I've noticed relatively large areas of enhanced probabilities (40-50%) that drop in and out over the period of a few scans.

The CI product displaying noisy values near MD-PA border.

wall_cloud

CI/CI-Severe Probability

At the tail end of the Pennsylvania squall line near the Pennsylvania/Maryland state line (near Gettysburg, PA), little convection had taken place, but was forecast to develop. The Convective Initiation Probability product did a good job in the 30-60 minute period in forecasting where scattered convection would continue to build southward toward Maryland. The areal extent was underforecast a little bit, but it showed where new development would likely take place. This was along an advancing cold front into an unstable environment where the development was expected, although timing would be uncertain because of some capping. This would have been useful in a forecast office.

Further to the south, additional convection developed over northern Virginia that was not forecast at all by the CI product, while it tried to develop convection from near Richmond, VA to the southern DelMarVa peninsula that never developed. The northern Virginia convection might have been origraphically been induced, as it was ahead of the advancing cold front and just to the east of the Blue Ridge Mountains to the west. It also developed cells to the northeast while the line didin't move eastward at all. The predicted convection from Richmond eastward developed with some mid-level cloud development, but didn't appear to be connected with any near-surface convergence, such as a front or a sea breeze. These areas would have made using the product a little more challenging.

Behind the Pennsylvania squall line, low probability CI's popped up almost continuously where sunshine returned and a field of cumulus clouds developed. Since the probabilities remained low in an areas where convection was unlikely (but not zero),


The CI-Severe product didn't seem to perform too well. The probabilities seemed to be overdone in areas where I wasn't even expecting much convection, and seemed to be underdone in areas where areas along and ahead of the squall line where new cells were forming. The first loop is CI with visible satellite and a regional radar loop; the second loop substitues CI-Severe for CI.

Gary Cannalte