Day 1 Wrap up and Feedback

Forecasters wrapped up day 1 of the GOES-R/JPSS experiment operating in Billings, MT and Chicago, IL. There were numerous storms and even some larger hail with storms in the Billings CWA throughout the day. While storms waited until very late to initiate in the Chicago area, forecasters there were able to evaluate various products for initiation and later look at storms very near there CWA, one producing a tornado near Peoria, IL.

Some feedback from the day is posted below:

ABI

Need quick guides for RGBs ASAP to understand the RGB and how to use it

It won’t be used in operations without some sort guide to quickly tell you how it’s useful

Some will like it and some won’t

Individual channel labels should include more than just the micron, but visible, IR, WV, etc…

Stability indices were really helpful for showing areas of concern

They showed gradients well

RAP was higher on CAPE values, but overall gradients were pretty similar

Cloudy regions it’s not helpful

Don’t necessarily want cloudy regions filled in, as it might cause weird gradients filling in with model fields, but might just need to overlay with a model field to see them both and how they compare.

GLM

Trouble understanding events, groups, and flashes

Too many menu choices, what do I look at?

Parallax still an issue when comparing with radar

GLM provides more extent coverage than the ground based networks that just show the flash

Need more training for all of lightning in the weather service. Forecasters aren’t sure what to use and why.

Color table differences between storms. Need different color tables depending on what you are looking for and what type of storms there are.

NUCAPS

Display needs to be improved to keep track of the points that you selected better

Maybe have a pop up window and highlight which point was being selected

Helpful to plot something underneath the points to keep track a little better

Some retrievals still seemed to need to be flagged as not passing QC that were flagged as green from the QC algorithm

NSHARP program shortfalls make some of these viewing options difficult

There should be some type of confidence reading in the sounding, or a reading of telling how much of the sounding might be good and useful

The mid and upper levels can still be useful, even when boundary layer is not good

NUCAPS picked up an inversion well in Billings and it compared well with radar identifying a fire and smoke plume nearby

No models for the point had any kind of inversion that NUCAPS picked up

ProbSevere

Trends in ProbSevere are most useful

Color table is a bit unique

The grouping of clusters seemed to drop ProbTor down lower

Needs to hone in better on the individual cores than grouping the clusters together

Using the full version on a four panel caused the readout to go off the screen

Want to be able to load both full and light on the same panel and toggle back and forth

ProbSevere was pretty successful in the Billings area yesterday locking on to the hail threat

Also seemed to struggle with storms in Montana, especially with storms farther from the radar probably because of ingesting MRMS and lack of radars in the area

-Michael