Jump to content

*** PLEASE REGISTER AND JOIN OUR DISCUSSION!!! ***

THE STAFF WANT YOU TO JOIN US AT PHILLYWX!!!

Register as a member today, and become part of the Delaware Valley weather community!

Our pro and seasoned amateur meteorologists, and weather enthusiasts from around the PA and NJ area together form a great group discussion, and we're asking folks that read our site today to register as members and post along with us!

Don't be intimidated if you're not an expert, ask questions if you're curious or want to build your knowledge!

Whether it's adding to our local profiles by reporting observations (and maybe becoming a SkyWarn Spotter!), or contributing more on the model interpretation side, we'd like you to join us in a constructive and insightful dialogue around all things Philly Weather!


Sign in to follow this  
Rainshadow

2019-20 Cold Season Model QPF/Snowfall Fcstg Performance. November 18th Event. Break Up The GFS, Another Win. SWETs Last Place.

Recommended Posts

Just to let you know, Tom & I plan on starting this with the first synoptic event on/about November 1st.  I am going to finalize last year's stats between now and then.

There is a new Euro, a new GFS and the same old NAM for the upcoming season.  We are going to go with the standard 10:1 snow/liquid equivalent ratio on model snowfall forecasts.  This would make it an apple to apple (except for the SREF which is Kucheraish) model  comparison as best we can.  PHL climo is 11:1. 

We thank the 13,000 views on last year's thread.  That was pretty good for a fairly dull winter.  Please don't hesitate to comment.

  • Like 5

Share this post


Link to post
Share on other sites

Because the Halloween/November 1st event has predicted upright convective elements, the start of the cold season qpf will begin with the next forecast/observed event that has .1" or more predicted/observed within 84hrs of starting time in November. 

I have finished the summary tables from last year and will post them once we really begin.

 

Share this post


Link to post
Share on other sites
On 10/27/2019 at 5:16 PM, Rainshadow said:

Because the Halloween/November 1st event has predicted upright convective elements, the start of the cold season qpf will begin with the next forecast/observed event that has .1" or more predicted/observed within 84hrs of starting time in November. 

I have finished the summary tables from last year and will post them once we really begin.

 

Lol now that we are ready to go and the convective outcomes are gone, it looks dry.  Should have started it a week earlier.

Anyway here are the average errors for all 50 or so evaluated qpf events from the last cold season vs 2017-18 final stats.  Also every observed/predicted snowfall event in which either near 1" fell or 1" or more was forecast by multiple models or multiple forecasts by one forecast system, yes I am talking to you GEFS, NAM.

 

1.JPG

2017-18:

2.JPG

3.JPG

2017-18 errors                                                                 GFS                          GEFS                 NAM                     SREF                      EC                        EPS

4.JPG

Share this post


Link to post
Share on other sites

euro

11/3 0z: 0

11/3 12z: 0

11/4 0z: .08

11/4 12z: 0

11/5 0z: 0

eps

11/3 0z: .02

11/3 12z: .01

11/4 0z: .08

11/4 12z: .03

11/5 0z: 0

Share this post


Link to post
Share on other sites
15 hours ago, tombo82685 said:

euro

11/3 0z: 0

11/3 12z: 0

11/4 0z: .08

eps

11/3 0z: .02

11/3 12z: .01

11/4 0z: .08

We will go thru the 00z run on the 5th, the start is too soon to include the 12z run tomorrow.

Share this post


Link to post
Share on other sites
On 11/3/2019 at 7:00 PM, tombo82685 said:

euro

11/3 0z: 0

11/3 12z: 0

11/4 0z: .08

11/4 12z: 0

eps

11/3 0z: .02

11/3 12z: .01

11/4 0z: .08

11/4 12z: .03

Don't forget the 00z run and sorry for being a PITA, but we start the Thursday night event with the same 00z/5th run.

Share this post


Link to post
Share on other sites

euro

11/5 0z: .14

11/5 12z: .23

11/6 0z: .24

11/ 6 12z: .23

11/7 0z: .16

11/7 12z: .14

eps

11/5 0z: .23

11/5 12z: .21

11/6 0z: .22

11/6 12z: .24

11/7 0z: .17

11/7 12z: .11

Share this post


Link to post
Share on other sites

Well.  Here we go!

Right off the bat some signs worth noting because they are repeats of the last two seasons already.

If the GEFS is drier than the OP or other models, watch out.  Other than that, it should quit while it is ahead.

The NAM gets schooled any time the precip source is mainly from the Atlantic Ocean (low or not associated).

The lately qpf excitable Euro is lucky we were not verifying for ACY because it might have come in last.

The SREF has too many expressoed members.  This was not just an ARW issue (although they are normally the culprits).

2019-20qpf.xlsx

 

 

Share this post


Link to post
Share on other sites
13 hours ago, tombo82685 said:

euro

11/5 0z: .14

11/5 12z: .23

eps

11/5 0z: .23

11/5 12z: .21

Lol, I am going to have to include a snow column because of the GEFS.  I am assuming the EPSs are 0 for this one.

Share this post


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

Lol, I am going to have to include a snow column because of the GEFS.  I am assuming the EPSs are 0 for this one.

0.00

  • Like 1

Share this post


Link to post
Share on other sites
6 hours ago, tombo82685 said:

0.00

Lol.  Looks like we can go thru 12z on the 7th for this one.  BTW gefs mean snowfall is up to 0.2" for Friday.  ❄️

Share this post


Link to post
Share on other sites
On 11/5/2019 at 7:10 PM, tombo82685 said:

euro

11/5 0z: .14

11/5 12z: .23

11/6 0z: .24

11/ 6 12z: .23

11/7 0z: .16

eps

11/5 0z: .23

11/5 12z: .21

11/6 0z: .22

11/6 12z: .24

11/7 0z: .17

Yup 12z will do it, it started to rain.

Share this post


Link to post
Share on other sites

First measurable event in the book and all models on all runs were too wet.  A positive tilt modeling bias.  Anyway it was down to the least wet by average and that was the OP ECMWF.   The NAM which started the wettest, never gained ground.  Sad because this was not a coastal event.  The GEFS had members that had PHL changing to snow before ending on three successive runs and it gets the last place finish on the snow side.  Only one SREF run had measurable, the rest of the modeling was correctly zero.

2019-20qpf.xlsx

Share this post


Link to post
Share on other sites

euro

11/9 12z: .16  snow .4

11/10 0z: .35 snow  .8

11/10 12z: .35 snow .7

11/11 0z: .26 snow .4

11/11 12z: .24 snow .2

11/12 0z: .19 snow .1

eps

11/9 12z: .23 snow  .7

11/10 0z: .36 snow .7

11/10 12z: .34 snow .6

11/11 0z: .31  snow .4

11/11 12z:  .26 snow .4

11/12 0z: .17 snow .2

Share this post


Link to post
Share on other sites
23 hours ago, tombo82685 said:

euro

11/9 12z: .16  snow .4

11/10 0z: .35 snow  .8

11/10 12z: .35 snow .7

eps

11/9 12z: .23 snow  .7

11/10 0z: .36 snow .7

11/10 12z: .34 snow .6

Will make the 00z run on the 12th the last one for this one, precip starting too soon on the morning of the 12th.  Sorry we have to keep more on top of this for this season. 

Share this post


Link to post
Share on other sites

Well the EPS is off to a rather inauspicious qpf debut this season.  It now has come in last place in two of three events.  I know the EC was "upgraded" in June, not sure if the EPS was too.  All the models were too wet again in a pos tilt event, not as bad last time, if that means anything.  The GFS came in first place being the least worst (wettest).  On the snow side, the we've been NAMMed run was too much for the correct 0 on its last run to overcome, it came in last.  The OP EC came in first, the only model within 84 hours of this event to not forecast an inch or more on any run.  Granted this is a straight 10:1, so once we were within 84hrs the models essentially (other than the we've been NAMMed runs) were coating PHL adjusting for a Nov S/LE ratio.

2019-20qpf.xlsx

Share this post


Link to post
Share on other sites
On 11/10/2019 at 9:40 AM, tombo82685 said:

euro

11/9 12z: .16  snow .4

11/10 0z: .35 snow  .8

11/10 12z: .35 snow .7

11/11 0z: .26 snow .4

11/11 12z: .24 snow .2

11/12 0z: .19 snow .1

eps

11/9 12z: .23 snow  .7

11/10 0z: .36 snow .7

11/10 12z: .34 snow .6

11/11 0z: .31  snow .4

11/11 12z:  .26 snow .4

11/12 0z: .17 snow .2

This next coastal event, we'll have to start jic with the 12z run today, 15th.

Share this post


Link to post
Share on other sites

euro

11/15 12z: .14

11/16 0z: .15

11/16 12z: .09

11/17 0z: .49

11/17 12Z: .17

11/18 0z: .05

11/18 12z: .05

eps

11/15 12z: ..12

11/16 0z: .22

11/16 12z:  .36

11/17 0z: .33

11/17 12Z: .27

11/18 0z: .11

11/18 12z: .1

  • Like 1

Share this post


Link to post
Share on other sites
1 hour ago, tombo82685 said:

euro

11/15 12z: .14

11/16 0z: .15

eps

11/15 12z: ..12

11/16 0z: .22

The NAM & EPS are tripping over each other to get last place.  Looks like the NAM (oh it handles early season coastals qpf terribly) with its boffo 1.07" may have already won last place.  Both are making the GFS look like a qpf genius (so far).

Share this post


Link to post
Share on other sites
On 11/16/2019 at 8:40 AM, tombo82685 said:

euro

11/15 12z: .14

11/16 0z: .15

11/16 12z: .09

11/17 0z: .49

11/17 12Z: .17

eps

11/15 12z: ..12

11/16 0z: .22

11/16 12z:  .36

11/17 0z: .33

11/17 12Z: .27

We can go thru the 12z run today (18th) and call it a wrap.

Share this post


Link to post
Share on other sites
On 11/16/2019 at 8:40 AM, tombo82685 said:

euro

11/15 12z: .14

11/16 0z: .15

11/16 12z: .09

11/17 0z: .49

11/17 12Z: .17

11/18 0z: .05

11/18 12z: .05

eps

11/15 12z: ..12

11/16 0z: .22

11/16 12z:  .36

11/17 0z: .33

11/17 12Z: .27

11/18 0z: .11

11/18 12z: .1

It looks like we may have to split systems for the weekend.  The CFP pcpn starting with the 12z run today (Wednesday) thru 00z Saturday and then the wave on the weekend with a start run TBD.

Share this post


Link to post
Share on other sites

Two in a row for the GFS !    Pcpn at Philly itself had more to do with the negative tilt trof than the ocean low. Granted the latter tracked at the eastern end of the modeling solutions (GFS/Euro much better than GGEM/UKMET) which other than drizzle made it a non factor at PHL.  In essence all the operational models had one "bad" run.  It so happened the GFS bad run was drier than the Euro's & NAM's bad run.  The NAM (12km evaluated, 3km NAM was beyond awful) would have come in first place alas for its first evaluated run (>1") reinforcing its reputation of not being  good at the end of its forecast cycle.  The SREF came in last place. Can't blame this one on the ARWs, it was the NMBs that were juiced, probably relatives of the 3km NAM. 

It looks like the weekend system may be the wettest event (not difficult to do) evaluated so far, we'll see how the GFS does with that.

BTW the GFS has already matched the number of first place finishes it had (2) for the entire 2018-19 cold season qpf evaluation period.

 

2019-20qpf.xlsx

  • Like 1

Share this post


Link to post
Share on other sites
25 minutes ago, tombo82685 said:

euro

11/19 12z: .16

eps

11/19 12z: .09

 

I saw the GFS was dry at 18z, thought maybe we could get away with it, but the Euro crossed the threshold anyway.

Share this post


Link to post
Share on other sites
15 hours ago, tombo82685 said:

euro

11/19 12z: .16

eps

11/19 12z: .09

 

Even if they are all zeroes from here on (For Friday), the die has been cast.

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...
Sign in to follow this  

  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...