Thundersnow12

January 10th-12th Winter Storm Potential

Recommended Posts

^

Fantastic read. Not necessary, as we here all know how difficult it is to forecast, but certainly therapeutic to reevaluate how a system evolved and what caused a forecast to not unfold as anticipated. In my laymen's terms....we got screwed Saturday night from a 3-5 " fall by a transient squall line. In the end, we all took it on the shin with this one. Also, certainly more eloquent and disproves ALEK's "rainer" call. ;) Posts like this show why these boards can be extremely relevant and educational. Thanks, again. 

  • Like 2

Share this post


Link to post
Share on other sites
2 hours ago, RCNYILWX said:

Several of the NWS offices, including mine, have done good post mortems on this event along with what went wrong in the forecast.

I'll try to address the various parts that the forecast didn't work out (or worked out okay but differently). I think convection played a noteworthy role in wreaking havoc throughout the forecast. It's likely too simplistic to focus on this factor alone, but I think helps hit the main ideas of what went wrong.

This event was essentially impossible to forecast well from a professional meteorologist perspective because the moving parts were constantly changing even in the shorter ranges when RAOBS typically help build the consensus. We rightly guessed that convection may ultimately modulate what happened.

There were a few "hunch" posts made about how the cold side of the system would be lackluster, and how Friday night would be primarily rain. Those were good calls clearly. My forecast analysis generally didn't work out, but hopefully was still informative for the meteorology beyond this event.

Friday Night Precip:
We had a very convectively active front for January along the inverted trough axis, due to the extreme warmth and moisture under the amplified southeast ridge. In more spring-like setups of this nature, we often look to areas within relatively close proximity to the boundary for the highest rainfall totals due to higher instability and getting the convective rates. It's common for precip to be less than forecast well up on the cold side of the boundary and the precip amount gradient usually ends up sharper than forecast. This is largely what happened, the heavy amounts of 2-3" verified but confined closer to the front in the southeast 1/3 of the LOT CWA.

The sharper cutoff northwest due to moisture robbing led to an earlier end to round 1 of precip and less snow and ice than earlier forecasts. That said, in the NWS, we never forecast the extreme ice amounts some of the models like the NAM kept showing with good reasoning to support why we didn't. That part of the forecast worked out, but then the earlier end and narrower zone of freezing temps still resulted in forecast ice accums in northwest CWA being overdone.

Saturday daytime and Saturday evening:
As of the Wednesday night/early Thursday model runs, the models had trended hard toward the 500 mb wave becoming negatively tilted and supporting a deepening 990s mb surface low lifting north-northeast. This is when I discussed potential similarities to the December 28 2015 event and possibility of significant sleet amounts.

I think a good case can be made that Friday night's convection helped assure a completely separate round 1 precip by: development of multiple competing sfc low centers and delayed deepening of the system as main wave stayed positive tilt for longer. As a result most of the daylight hours Saturday were devoid of steady precip (except for the odd LES under the warm layer into NE IL), precluding any sig ice, sleet and snow accums as cold press from north deepened.

On Saturday PM, there was finally a respectable area of backside/deformation snows with the finally deepening mid-level trough streaming toward the area. Models had hinted at this and it seemed we'd finally get a solid period of snow. However, at the same time, an extensive intense squall line developed over the lower OH valley down to near the Gulf Coast. Despite the support for heavy precip rates that were expected to accompany the Saturday evening snow, it seems likely that the convection had to play at least some role in the deformation precip getting shredded like it did.

What part of the forecast did work out:
The lakeshore flooding forecast: based off reports, photos, and videos of impacts, likely the most significant lakeshore flood episode since the Feb 87 event, surpassing Halloween 2014 due to near record high lake levels. Calumet Harbor lake level gauge came within 1" of the all time record.

If anything, the lakeshore flooding was more significant, with inundation noted a few blocks inland in the South Shore neighborhood. We now have a good baseline for future higher end events in this near record high lake level regime.

Final Thoughts:
Based off the available ingredients for the system, the more significant potential outcomes were reasonable forecasts given the preponderance of the data at hand. One could probably have found ensemble members that mirrored the actual outcome, but those were thought to be outliers at the time. As NWS MKX showed in their what went wrong graphic, the snowfall ended up close to the 10th percentile of the ensemble distribution.

Perhaps we could have done more to collaboratively (with WPC and surrounding WFOs) tighten up the rain forecasts based on conceptual model of a convectively active front, and forecast a smaller footprint off of the charts extreme amounts for January. That would've helped produce better river forecasts.

As for the winter side, once to headline phase, it's tougher to couch things with as much uncertainty, despite the inherent challenges with this forecast. I'm not sure we could've totally avoided a bust in this regard, aside from playing things a bit more conservatively because of the possibility that downstream convection would destructively interfere with the setup.

Have to say this too, that when the typically most reliable model and ensemble, the ECMWF struggles mightily for an event, it does affect the quality of the forecasts. No model did extremely well, some less bad than others.

Despite the failures and frustrations, it's an interesting case to study and learn from for the future to do better next time. Sorry for this being so long, but thank you if you did take the time to read through it.



Sent from my SM-G965U using Tapatalk
 

quoted for a new page so it doesn't get lost, very nice recap

  • Like 1

Share this post


Link to post
Share on other sites

Why do Doctors practice medicine? It allows for error, for when results do not follow expected protocol. Your reflection of the system. Will help others become better forecasters and is a important personal pathway of educational growth. Thanks for sharing this awesome post..  

 

  • Like 1

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...

  • Recently Browsing   0 members

    No registered users viewing this page.