Pre storm analysis with Derived Products and METOPS sounding

The Grand Forks CWA was recovering from a previous line of showers.  The sun was returning and heating the environment.

Temps were recovering into the mid 80s with DPs in the low to mid 60sby 20z.
Metars showed a wind shift line SW/NW right on the border of our CWA.

Derived Cape at 1902z ran from 450-500 in our northwest counties to 175-200 in the southwest. There was a gradient that lined up with the wind shift line (west of wind shift line dropped  to zero).  Animation showed the gradient moving with the wind shift.

Derived TPW ran in the 1.2-1.3" range on the average.  Animation showed a gradient moving from west to east with the highest values along the wind shift, dropping to 0.6" behind  the wind shift. 

Conclusions of Derived Cape and TPW:

It was fascinating to see the satellite derived CAPE and TPW oriented to the wind shift line as it moved from west to east (whether the numbers are exact or not, the relative orientation was there).  We expect initiation along that line.

We also picked a METOPS-A sounding in the far southwest part of the CWA where instability looked the highest.

The original METOPS-A sounding was from 1647z, and we were now at 20z.  CAPE was zero, not very unstable.
The original is below:

Modifying it to 20z ob T=84F and C=63 we got SB CAPE 1443.  I used the method I learned yesterday of following mixing ratio to modify DP up from surface.  Also modified T a little bit above the surface to look more like reality.
Edited METOPS-A Sounding below:

Conclusion of METOPS-A:

Modifying the sounding was necessary since 2 hours had elapsed from 1706-1902 and heat and moisture had jumped considerably.

Davis Nolan
WKRN Meteorologist

Our environment is not overly unstable.

Pre-Storm Analysis Across FGF Area

During pre-storm analysis over the FGF CWA, I looked at baseline derived products from GOES-16 in a 4 panel as shown below, which included the moisture difference product. 

There is a tongue of slightly higher CAPE values, approximately 500 J/KG, entering the western portion of the CWA. This coincides with a tongue of higher TPW, around 1.2 inches and better LI values (-3 to -5). The better moisture can also be seen in the moisture difference panel on the better right.

I also looked at a 4-panel water vapor procedure, which included the upper, mid, low, and simple water vapor RGB. This was mainly to get a feel for the larger scale moisture and flow pattern. Once some instruction on the RGB was given, I found it useful in assessing where the moisture contributions were coming from.

A look at radar data from both KMVX and KMBX reveals a fine line or surface boundary approaching the FGF CWA from the west. This line is the leading edge of the drier air, which means the focus for convection is generally going to be along and east of this boundary. The CAPE and TPW gradients were generally along and east of this line as well above in the derived products.

I also looked at a modified MetOps sounding in the western part of the CWA. The modified sounding had CAPE values around 700 J/kg.  Before modifying the sounding, there was little to no CAPE. These values were generally consistent with the baseline derived, although slightly higher.

-Ironman

Significant Issues with Dervived Motion Wind

With tropical storm  Cindy in the Gulf of Mexico, it seemed reasonable to use the derived motion wind to estimate the strength of the tropical storm. I went ahead and loaded the visible satellite imagery with the associated derived motion wind vectors:

Unfortunately, the performance of the derived motion wind was poor. There were a lot of spurious wind vectors across southern Texas and Mexico, which had values between 50-80 knots.  These are likely due to the algorithm tracking small areas of cumulus.

In addition, the  wind vectors across the southern part of the tropical storm were in the opposite direction. These winds should be from the west instead of the east.

The overall poor performance of the derived motion wind gives me zero confidence in operational utility of the data.

-Lost Met

GOES-16 TPW Near Cindy

On 21 June, Tropical Storm Cindy approached the Louisiana Gulf Coast. The main threat with this system was heavy rain and flooding. The GOES-16 TPW baseline product provides forecasters with rapidly updating information about how much moisture is available in the atmosphere. Widespread TPW over 2.0" was sampled in the vicinity of Cindy, with some localized areas over 2.5" (Fig 1). These observations agree with nearby radiosonde data (LIX), which measured 2.36" of PW  (Fig 2).

Figure 1: 21 June 2017 GOES-16 TPW (color) and IR (gray).

Figure 2: 21 June 2017 LIX 12z radiosonde.

- Bill Line, NWS

LIX: Tropical System (New Orleans)

Tornado Warning Issued prior to ProbTor identification: Warning environment is within outer rainbands of Tropical Storm Cindy. 

• Warning decision mainly based on velocity couplet and mini-super cell appearance on reflectivity. Velocity couplet is weak and probably on the lower end threshold but ProbTor had not identified it by 19:35. 

Velocity Couplet

• ProbTor identified the cell by 19:38z with a prob of 9%.

ProbTor at 9%

• 1946Z: 0.5 velocity signature weak, but ProbTor increased to 15% probability.    Forecaster  -Tahoe

EWP Week 1 Day 3 Operations

For day 3 of week 1 of the 2017 GOES-R/JPSS Spring Experiment, participants will begin operations in the Grand Forks, ND and New Orleans, LA CWA's. The New Orleans folks will monitor tornado potential associated with Tropical Storm Cindy.

EWP Week 1 Day 2 Summary

Our two groups of forecasters remained in the Dodge City and Chicago CWA's for the duration of Tuesday.

- Bill Line, NWS

• GOES-16 ABI Imagery and products
• Imagery
• Sandwich product is a top 5 image to look at in operations
• Compared to RGB’s, sandwich was better. Looked at daytime convection, couldn't see much of the gradient in the cloud top temperatures. In sandwich, could see it well while preserving the gradient.
• Cheat sheets for various imagery bands and combinations will be very helpful
• Derived products (CAPE, LI, winds, dsi, tpw)
• Instability very low compared to SPC mesoanalysis and most model data
• In kansas, GOES cape was half as much as spc mesoanlysis and GFS
• Convection developed within the highest values of instability and TPW
• Channel differences
• 
  
• RGB composites
• I am having trouble wrapping my head around the Airmass RGB
• GLM
• In Illinois, I used glm to try to get lead time to cg, but didn’t have lead time
• In kansas, GLM data provided lead time to first NLDN cg by around 20 min
• My favorite display was the 5-min product updating every minute.
• Most forecasters looked at the gridded events and flash combo.
• I like looking at the flash because it is what I’m familiar with
• For tv, would mostly show flashes, but maybe last frame would show the extent (gridded event) to get across that lightning can extend far from the storm core, and could potentially hit the surface far away from thunderstorm
• ProbSevere
• In illinois, the environment was marginal. Probsevere values stayed low confirming the low end threat. One report was from strong outflow with very low dbz, so probsevere had no probs on the outflow and very low probs with associated storm. This happens a lot out west, so would be nice to be able to track situations like this (outflow winds).
• Not much velocity data in the algorithm. Storm with high std, I saw this before probsevere increased. Would be nice to somehow incorporate more vel data.
• When I had a lot of storms and warnings going, probsevere was very useful. Instead of viewing all tilts with all storms, prob severe was great for SA, directing me to storms that needed attention. For all storms that were producing hail, probhail was over 95%. Gave me confidence that the same types of structures were being seen with new storms as with old storms.
• NUCAPS
• Afternoon pass confirmed low instability in our area. Storms did not go severe
• Using plan view display, looked at dew point depressions just above the surface, helped to confirm dry atmosphere.
• Loaded op and exp modified ahead of line. Experimental made well a mixed layer, added instability to around 500 j/kg. This even seemed a little high, considering convection was very weak.
• Modified was much better. Non modified had a lot of CIN, unbreakable, but we had explosive development., modified was much better. Went from 300 to 2400. Southern plains will get nice well mixed layer, but east coast you won't really get this.
• Some flag to show where NUCAPS is significantly different from a model would be helpful. Perhaps above a certain layer
• Cross section was useful to see how deep dry layer was
• Would be beneficial to have plan view CAPE of modified soundings

Some thoughts and observations on NUCAPS gridded data...

It's been interesting to observe and think about ways the gridded NUCAPS data can be used in an operational setting dealing with convective weather.  The recent blog post highlighted lapse rates and their similarity to analyses provided by the Storm Prediction Center.  For this one, I was going to take a quick look at Convective Inhibition (CIN) data as provided by the NUCAPS gridded data, in conjunction with satellite observations of the region.  This first image shows the NUCAPS derived CIN data over the southern and central High Plains at ~19Z overlaying GOES-16 0.64 um imagery.  

The color curve in the above image is set so that generally brither colors (light blues to greens) indicate higher values of CIN and thus regions where convection is being inhibited.  Notice that low clouds can be seen in areas of western Kansas where CIN values are lower.  The same can be said for portions of northern Texas.  Meanwhile, CIN values are peaked generally in the TX/OK panhandles and parts of southern Kansas where there is little cloud development in this scene.  Taking in the big picture here, this would indicate any development might be favored to continue in western Kansas where one cell can already be observed to develop. A post earlier showed that lapse rates were not that steep in this area during the early afternoon, however, steeper lapse rates just to the west were about to advect into western Kansas.

Forecasters are going to be inclined to check on the data individually and take it on an individual basis sometimes.  Comparing the NUCAPS gridded data with the soundings can present some problems sometime.  Take for example, the gridded data in the far NE corner of the TX panhandle.  The observation at this point shows a value of ~279 J/Kg (of CIN) between the 850 and 500 mb levels, which is where the inversion is located in this area.  Now, looking at the actual sounding for this same location shows different values.

Although not shown, the CIN values for a ML parcel in this scene according to the sounding data were about 600 J/Kg.  Granted, this is just one example shown, but was representative of the broader issues.  So, there perhaps needs to be a better way to obtain the gridded data so that there are no potential discrepancies between the NUCAPS Soundings data, and the gridded data.

-Kris W

Gridded NUCAPS Verification

Gridded NUCAPS verified well with the SPC Mesoanalysis when comparing mid-level lapse rates:

Steep mid-level lapse rates (>7 C/km) are located near the Kansas/Colorado border on both analysis. 

Westerly flow aloft was progged to advect the steeper lapse rates eastward into deeper moisture across western Kansas. This led to explosive development a couple hours later.

-Lost Met

GLM Parallax error?

I wonder if I observed the "parallax" error in GLM?

At first I thought that the GLM lightning was lining up with the gust front instead of the core behind it.

However, when looking at ENTLN lightning, it lines up with the core.

Davis Nolan
WKRN Meteorologist