Week 4 Day 2 Wrap Up

Wrap from operations yesterday is posted below. We spent the day operating in the Sterling, VA and Midland, TX CWAs.

ProbSevere

Popped up really fast yesterday where it would go from single digits to high percentages in 10 minutes.

It was hard to have much lead time with it yesterday since storms ramped up quickly

One storm had some rotation but ProbTor remained low which helped some with decision making

Seemed like it helped show the increasing trends with the storms in Texas

GLM showed good increases too which helped verify the ProbSevere trends

ProbSevere would jump whenever storms were clustered but would go back down when they would
distinctly distinguish the storms

ProbWind didn’t do very well picking up on the bowing line segments

The linear storms modes are still really hard to use it too usefully in operations with the storm clustering

Again, using different tracking mechanisms might be better even if there was some increased latency.
It would at least be good to look at.

GLM

Primarily used Flash Extent Density and the Total Energy and issued a warning pretty much just off of
the lightning going up in an environment conducive for explosive growth

Lightning dropped off drastically as the storm reached its peak intensity on radar and ProbSevere
parameters

Primarily used the 5 minute with 1 minute update

Using the 2 minute was a bit more sporadic

Looked at the event density and there didn’t seem to be anymore added value over the what the
flash extent density already gives us

One did look at the event density of a developing thunderstorms and the event density seemed to help
show the ramp up in the storm prior to FED

Paired the lightning with some of the satellite data and could see the updraft grow on the imagery as
the lightning increased

It was ok to come in cold without any bias from any prior training

It would be helpful to have a baseline scale of what values are high or low to assess storm strength

There hasn’t been as much correlation between lightning trends in west Texas and Eastern New Mexico
and storm severity

NUCAPS

A lot of the area was cloudy yesterday and there were lots of yellow dots

Compared some of the soundings to the mesoanalysis and the uncorrected ones were quite a bit off in
the boundary layer

Would probably rather just have the corrected soundings since they seem to be much closer to reality in
the boundary layer

Liked having the continuous field of data on the gridded data and would rather have the continuous field
and ignore the bad data rather than black out a lot of the area

Gridded data seemed to compare pretty well to mesoanalysis

Was able to pick out some of the gradients in the data

Need to just default to an image and not have the gridded option

Lapse rates seemed to be the most useful for using the gridded data with

The PWAT values seemed to compare pretty nicely with the LAP data and the mesoanalysis

The continuous field gridded data helps to visualize the gradients more than having to try to piece it
together

Didn’t really look at the operational soundings

The timing was pretty good for yesterday’s storms

All-Sky LAP & CI

The all-sky CAPE started out pretty low in the area but increased throughout the day to 1000-1500 and
Cu started to develop and eventually storms fired

Used the CI overlap with the visible and it was still very jumpy and would go up and down and would kind
of decrease the confidence

On the Severe CI it might be good to just not show anything until the percentage gets to 30% or so

It did seem to show a trend in west Texas where convection eventually initiated

The Severe CI never really went higher than 10% at all

Storms didn’t really become severe yesterday but there were a few severe areas

Over long time periods where CI would increase some were areas where storms initiated, so it did seem
to hone in on some of those storms

The All-Sky PWAT values compared very well with the Mesoanalysis values and showed the boundaries
very well while the CAPE values were overall a bit low but the gradients matched up very well spatially.

ABI

Used a few of the RGBs to help track initiation and storm development

The Simple Water Vapor RGB helped show the dryline placement yesterday in west Texas where
observations were sparse

Used a 4 panel of RGBs to look at different features throughout the region

The Day Cloud Convection RGB was pretty neat to use yesterday where the color contrast helped pick
out the areas of lower clouds streaming into the storms in the DC area

The channel differences were a little hard to use and the color tables could use some adjustments,
especially the split window