Jump to content
  • Member Statistics

    17,609
    Total Members
    7,904
    Most Online
    NH8550
    Newest Member
    NH8550
    Joined

WxChallenge 2013-14


Recommended Posts

  • Replies 1.9k
  • Created
  • Last Reply

A 21+ knot would've helped me compared to the national consensus, but am surprised it only verified at 20 knots as well...for tomorrow I'm debating the high, and I doubt we decouple tonight, but we could quickly see things decouple tomorrow evening with a really quick drop in temperatures lending to a likely 6z low tomorrow night. Do I want to drop the temperatures to below 20 by 6z for my forecast is the question.

Link to comment
Share on other sites

38/18/14/0

 

The GFS and NAM both seem to be hinting at an inversion tomorrow with a much chillier start...so even with mid-level temps warming several degrees held highs in the upper 30's. With the heart of the high pressure moving overhead tomorrow evening, what should be clear skies the first half of the night and dew points around 10 I expect a free fall in temperatures in the evening. If anything happens to prevent a decouple initially for whatever reason things may not get below 20 before 6z which is why I didn't quite go all out on the low. For the winds, with a NW flow and still a bit of a pressure gradient early went higher than most guidance...we'll see however if 14 knots busts too high, which it may if an inversion is in place. Really wanted to go aggressive where I could and move up more than I did today.

Link to comment
Share on other sites

39/21/12/0

 

It got to 38 today... why not 39 tomorrow on the rebound? Quick de-couple and drop fast tomorrow night. NE winds are bad in Concord!

 

Quick de-couple and drop fast is not 21F. Dew points are near 10F, and almost always when they decouple, they fall to the dewpoint before 06z.

 

I have no idea why I'm the lowest in the range. There's support for 10F.

 

I'm 42/12/15/0 tomorrow.

Link to comment
Share on other sites

Quick de-couple and drop fast is not 21F. Dew points are near 10F, and almost always when they decouple, they fall to the dewpoint before 06z.

 

I have no idea why I'm the lowest in the range. There's support for 10F.

 

I'm 42/12/15/0 tomorrow.

Verification wise, there is no support for 12. Anything from 15F to 18F yes there is support but for 12 that is way out there. I do detailed MOS verifications and 15 is the lowest I can see. I even do EURO verifications and it has 26F at 6z. The most it was off on a low was 14 and also 13, but those happened once. The more frequent errors were 11 to 8 and the median, mean, and mode for nights where they decouple and cool was 8, so my statistics say that your 12 is one hell of a ballsy forecast low. 

Link to comment
Share on other sites

Verification wise, there is no support for 12. Anything from 15F to 18F yes there is support but for 12 that is way out there. I do detailed MOS verifications and 15 is the lowest I can see. I even do EURO verifications and it has 26F at 6z. The most it was off on a low was 14 and also 13, but those happened once. The more frequent errors were 11 to 8 and the median, mean, and mode for nights where they decouple and cool was 8, so my statistics say that your 12 is one hell of a ballsy forecast low. 

 

I've looked at decoupled nights in past years and have found numerous times where the 06z temperature was 5-8F below MOS at the same time.

 

And just physically speaking, on calm clear nights in November over the past 5 years, they almost always reach their dew points before 06z (and often fall another 1-3 degrees below that). I can see their dew points being higher than modeled (say 14F, like the Euro suggests) to start the night, but that's why I think 12F is a good forecast.

 

On top of that, the 12z Euro doesn't have them decoupled for whatever reason. The 00z Euro did, and got them down to something like 20-22F. I'm only asking for eight to ten degrees on top of that, which per your own analysis, is perfectly reasonable.

 

 

And what about day 3? MOS forecasts were 26F (GFS) and 28F (NAM) and they found 21F. MOS dew points were 19F and 21F. Going with the MOS dew point for tomorrow gives 10F.

Link to comment
Share on other sites

I've looked at decoupled nights in past years and have found numerous times where the 06z temperature was 5-8F below MOS at the same time.

 

And just physically speaking, on calm clear nights in November over the past 5 years, they almost always reach their dew points before 06z (and often fall another 1-3 degrees below that). I can see their dew points being higher than modeled (say 14F, like the Euro suggests) to start the night, but that's why I think 12F is a good forecast.

 

On top of that, the 12z Euro doesn't have them decoupled for whatever reason. The 00z Euro did, and got them down to something like 16F. I'm only asking for four degrees on top of that, which per your own analysis, is perfectly reasonable.

 

 

And what about day 3? MOS forecasts were 26F (GFS) and 28F (NAM) and they found 21F. MOS dew points were 19F and 21F. Going with the MOS dew point for tomorrow gives 10F.

The Euro data I have via WxBell, and AccuWeather had CON at 24F at 6z on its 0z run (winds of 2kts), and 26F at 6z (winds of 4kts) on the 12z run. Both had dews of 15F. With the argument you're making, you're 12F is plausible I think...10F I think is pushing it. If I wasn't in the top 5 and 1 overall, I'd probably have gone lower than the 17F I went. 

 

39/17/16

Link to comment
Share on other sites

The Euro data I have via WxBell, and AccuWeather had CON at 24F at 6z on its 0z run (winds of 2kts), and 26F at 6z (winds of 4kts) on the 12z run. Both had dews of 15F. With the argument you're making, you're 12F is plausible I think...10F I think is pushing it. If I wasn't in the top 5 and 1 overall, I'd probably have gone lower than the 17F I went. 

 

39/17/16

 

Ya I had the wrong Euro low from the 00z run. Still I don't think I'm really out there. If dew points start at 15F, I might be pushing it a degree low or so, but the fact that there's NOBODY below me shocks me. Physically speaking, I just don't see 16F+.

Link to comment
Share on other sites

I've looked at decoupled nights in past years and have found numerous times where the 06z temperature was 5-8F below MOS at the same time.

 

And just physically speaking, on calm clear nights in November over the past 5 years, they almost always reach their dew points before 06z (and often fall another 1-3 degrees below that). I can see their dew points being higher than modeled (say 14F, like the Euro suggests) to start the night, but that's why I think 12F is a good forecast.

 

On top of that, the 12z Euro doesn't have them decoupled for whatever reason. The 00z Euro did, and got them down to something like 16F. I'm only asking for four degrees on top of that, which per your own analysis, is perfectly reasonable.

 

 

And what about day 3? MOS forecasts were 26F (GFS) and 28F (NAM) and they found 21F. MOS dew points were 19F and 21F. Going with the MOS dew point for tomorrow gives 10F.

I wish I knew what the EURO had day 3 at 6z because I only have the lowest it had that day which was 28F so 7 off. Both GFS runs were 5 off so that gets to 15F for a low tomorrow at 6z assuming same error. I still like 15 to 18 better than 10 to 15. 

Link to comment
Share on other sites

Ya I had the wrong Euro low from the 00z run. Still I don't think I'm really out there. If dew points start at 15F, I might be pushing it a degree low or so, but the fact that there's NOBODY below me shocks me. Physically speaking, I just don't see 16F+.

I can understand that. Again, if I was say, at 200th for CON, I would have gone below 15F for sure. I'm honestly surprised national consensus is 19+F. I think my low is on the high side of things for sure, but if its below 15F it'll surprise me a bit.

Link to comment
Share on other sites

42/17/16/0

 

Hoping for a mix to 900 mb (probably won't get it) and maybe that wind can be achieved tonight? Seems doubtful they'll get to anything much above 12-14 with the high moving overhead tomorrow. And like everyone else, can't believe the climate report matched the on the hour observation for wind. That's very rare and of course, suspicious.

Link to comment
Share on other sites

19, 20, 12, 16, 15, 20, 20, 14, 17, 18, 15

 

Those are the winds (in kts) between the 7AM report and the 6PM report (~11 hours). This gives a mean of 16.9 kts and a standard deviation of 2.74 kts. Assuming the winds during that time period were reasonably steady, and the distribution of wind speeds reasonably gaussian (those don't seem to be too bad of assumptions over that time period), for any 2-minute wind period you'd expect that there would be about a 6.75% chance to reach or exceed 21kts. There are 330 2-minute periods during that 11-hour time span. The probability that every single one of those would be 20kts or less is given by 0.9325^330 = 9.65e-11. Based on that analysis, it seems like a statistical certainty that the verification should have been higher than 20kts.

 

By the way, using that same analysis, your 50% probability of exceedence falls somewhere between 24kts and 25kts, closer to 25 kts.

 

The probabilities to reach or exceed a given wind speed based on that mean and standard deviation with 330 (independent) time periods are as follows...
20kts - 100%
21kts - 100%
22kts - 100%
23kts - 98.7%
24kts - 79.5%
25kts - 40.2%
26kts - 13.7%
27kts - 3.7%
28kts - 0.1%
 
Even if you assume the degrees of freedom are only 1/4 of the total number of samples (330), you still get a 100% probability of exceedence for 20kts, and your 50% value is between 23kts and 24kts.
Link to comment
Share on other sites

The low should be 29°F, so in that sense it's definitely wrong. But I'm digging deeper into the wind as well.

 

I agree, that wind seems wrong.  I was feeling great about my 23 kt when I saw 20 kt on the hourly... guess not!

 

Also feeling good about my 17 for the low.  Latest USL has it at 18 at 6Z tonight, which I suspect may be a degree or two too high.  Surprised national consensus was a balmy 19.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.

×
×
  • Create New...