NUCAPS Gridded Elevated Lapse Rates...

Hello from the Hazardous Weather Testbed in Norman, OK!

So, we were taking a closer look at the NUCAPS (NOAA Unique Combined Atmospheric Processing System) data, both in the sounding and in the gridded format for severe weather applications.  Today, we are specifically focusing on the potential for deep/severe convection in northern Illinois and western Kansas...and surrounding areas.  It's important to note here that NUCAPS soundings are satellite derived (Soumi NPP, Metop-A, Metop-B and JPSS (future)) and independent of model data.  The NUCAPS soundings have been utilized in the HWT now for three years.  The NUCAPS gridded data meanwhile are a relative new-comer to the HWT, having been implemented briefly at the end of last year's HWT...but are more of a formal product being evaluated this year.  The efforts to get the gridded NUCAPS data into AWIPS were led by the SPoRT group at NASA MSFC and have played a role in weather forecasting in Alaska, especially for the Anchorage CWSU.  Now, we are taking a closer look at the data for potential severe weather applications and trying to think of proper data displays and visualizations.  One thought that came to mind was looking at lapse rates of the NUCAPS vs various model output.  This image shows 700-500 mb lapse rates for the NUCAPS gridded data along with GOES-16 0.64 um imagery (upper left), HRRR (upper right), GFS20 (lower left) and NAM 40 (lower right).

Notice that the 700-500 mb lapse rates from the NUCAPS data are very similar in pattern to the various model data shown above. While specific values at various locations differ by relatively small percentages, patterns and gradients are overall fairly similar.  However, the NUCAPS data suggests perhaps slightly steeper lapse rates (C/km) over portions of the Great Lakes than the suite of model data.  In the environment in that area today, these levels correspond to temperature values roughly in the 0C to -20C temperature range...per NUCAPS soundings.  So, we would expect to see perhaps slightly more robust convective development in this region today, than model data alone might imply.

Next, we're going to take a closer look at some of the NUCAPS soundings in context of both the gridded data and other analysis data sets.

-Kris W and Nadia S

KDDC: SVR Use of Prob Severe In Warning Decision

Issued a severe thunderstorm warning as 50 dBZ was reaching 35,000 feet and there was a significant storm top divergence signature aloft. A three body scatter spike was also seen. Prob Severe values for hail increased to over 90% and then have been consistently around 98% since warning issuance.

See blog post called Prob Severe compared to Classic Radar Interrogation Techniques

An LSR for hail of hen egg size (2.00 inches) came in 2 minutes after the warning was issued.

At 3:21 pm, a report of ping pong ball sized hail (1.50 inches) was reported.

I was having an issue with the text workstation, which delayed the issuance by a few minutes.


Issued a new severe thunderstorm warning on the same storm as Prob Severe has consistently been running over 90% for hail and has even shown an increase in Prob Wind which is now 80%. 60 dBZ was also reaching 35,000 feet which is well above the -20C (~22,000ft).

Several reports have also come in since the time of new issuance of warning including hail half-dollar to the size of golf balls, which the MESH was indicating as well.

The storm briefly weakened before similiar signals were indicated in all tilts with 60 dBZ reaching ober 30,000 feet and a consistently high Prob Severe value at 98%.

-Ironman

GLM data vs Earth Networks in Prob Severe

A strengthening storm over the NW part of DDC area had big increase in CG lightning activity, but there was very little shown in the GLM data. See image below.

GLM data continued to be much lower and insignificant whereas the CG data was showing about 50 flashes per minute. This storm had about 60 dBZ over 20,000 feet, so higher cloud flashes would be expected, but its not being shown.

GLM vs. NLDN

GLM showed important utility by providing lead time for eventual cloud-to-ground lightning.

At 1850 UTC, GLM indicated the presence of flashes (and associated events):

 This quantity of events and flashes increased by 1920 UTC:

By this time (20 minutes later), the first cloud-to-ground strikes were detected by NLDN:

-Lost Met

Chicago: Meso-Analysis

Meso Analysis: Chicago (LOT):

Main points up front:

• GOES 16 derived stability indicies help diagnosis the anemic convective development in LOT's CWA. 
• GOES 16 derived stability helped reveal areas where convection is most likley to become active later (western CWA).
• Metop-A Nucaps passes added confidence to the analysis seen in GOES-16.

Initial line of convection and outflow boundary struggling to sustain deep convection across the northern CWA and little new convection firing along the outflow boundary as seen by the radar above. 

The GOES-16 Derived Products:

GOES-16 derived stability fields show helpful information of weak CAPE values of only a few hundred (J/kg) and positive LI's across much of the LOT CWA.

 

An analysis about 45 minutes layer and using a compressed color scale to increase contrast shows the most favorable activity would be along the western CWA which showed higher CAPE (300-400 j/kg), negative LI's, and few tenth's higher on PWs. This matches well with surface observations which show higher surface temps and dewpoints in the mid-50s to lower 60s in Davenports area. 

Metop-A NUCAPS & GOES-16 Airmas RGB:

Overlaid Airmas RGB and Metop-A pass to compare. No strong signal evident of moisture or large dry layers intruding into the areas. Did sample and modify Metop-A dervide soundings in the Davenport sounding to the west and more stable environment to the east. Overall, modified soundings show a similar pattern of slightly more unstable to the west where convection is initially firing. 

Forecaster: Tahoe

Met-Op Sounding

A met-op pass occurred around 17z over the DDC region. The surface temperature of the sounding was consistent with actual observations in the area. I did have to modify the surface dew point as the sounding had 62F and the surface obs were closer to 68F. The modified sounding is below, which showed about 4,500 J/Kg of SBCAPE.

This value is more consistent with data in SPC mesoanalyis, whereas the original sounding showed only about 2,000 J/KG of SBCAPE.

-Ironman

VIS/IR Sandwich Procedure

The "Sandwich"  combo is very useful for situational awareness and developing convection. When the first developing cell intensified, colder cloud tops were clearly shown in the IR, with the cu field with less vertical depth clearly identifiable with weaker cells.

-Ironman

4-Panel Baseline Derived Products

A useful 4-panel for mesoscale analysis using Baseline Derived Products is to have TPW, CAPE, and LI with the Moisture Difference product. This gave me a quick look at the changing environment with increasing moisture and instability. The moisture difference product adds value due to its higher resolution. Once the product was explained in more details, it makes sense why areas with thicker cloud appear darker. This could be confusing at first glance since drier areas are darker.

-Ironman

IR/Visible Sandwich

A procedure that combined both visible and IR GOES-16 data provided excellent situational awareness for initiating convection. The "sandwich" display has visible with semi-transparent IR overlay.

In the example below, the appearance of cooling cloud tops gave a heads up that deep, moist convection was initiating:

-Lost Met

EWP Week 1 Day 2 Operations

On day two of the 2017 experiment, our two groups are operating in the DOdge City, KS and Chicago, IL CWA's.