Better Than The Models ?

[Coast]

Tomorrow:

http://www.knmi.nl/waarschuwingen_en_verwachtingen/provincie.php?pr=ZE

http://www.knmi.nl/warnings_and_forecasts/warnings_coastal_waters_and_dutch_continental_shelf.html

[MurcieBoy]

Coast, wow! Thanks for putting these charts up...

[Coast]

If you want to keep an eye on any specific thunderstorms tomorrow, this European one may give the greatest coverage:

http://www.euclid.org/realtime.html

I'll take a look at any ESTOFEX warnings in the morning with regard to thunderstorm activity for the day.

[Timmytour]

OK so its 1st of Feb and Murcieboys prediction is 100% verifying, 130 mile an hour winds and major destruction is on the cards

The only problem is its on the other side of the world in Australia

But joking apart if MB had come on here in November and predicted a destructive storm with 130 mile an hour winds would hit at this exact

time somewhere in the world that would be some prediction to come off.

Maybe he is on to something but just needs to tweet the data

The Nort Sea Storm (Australia)

:lol:

Well it's the beginning of February and the weather is certainly in the news....but for what it's doing in Australia AND the United States!

I did hear that severe gales are expected in Northern Scotland tonight.....

Still keeping an eyes on things but, right or wrong, (and I have to concede the latter seems much more likely) this thread has kept my interest in the weather in a month which has otherwise been after the Lord Mayor's show!

[LomondSnowstorm]

:lol:

Well it's the beginning of February and the weather is certainly in the news....but for what it's doing in Australia AND the United States!

I did hear that severe gales are expected in Northern Scotland tonight.....

Right you are, pretty blowy here too with some sleet and snow to low levels tomorrow morning and again on Friday, with winds climbing towards storm force in some parts. Not exactly a North Sea storm but noticeably stormy up in haggisland

[Roger J Smith]

Day 1 validation (24h to map time)

_____________________________________

Very little change in the model output, except that both GFS and ECM have slightly developed the low around 57N 18W at map time (now just 24h away from the prog maps at 12z).

Therefore, no basic change in commentary, and only slight movement in the correlation stats (below).

Before listing those, would also note 500 mb wind at Shetland is SW today. Will be posting a full report on the other details of validation starting tomorrow. This will actually favour MB more than the correlation given that there should be a peak of strong winds late tomorrow. I am forecasting max gusts in Ireland near 65 knots and northern Scotland near 70-75 knots.

statistic ...... DAY 10 .. DAY 09 . DAY 08 .. DAY 07 . DAY 06 .. DAY 05 . DAY 04 .. DAY 03 . DAY 02 . DAY 01

GFS

day to day ..... ----- ...... 0.87 ...... 0.77 ...... 0.84 ...... 0.90 ...... 0.92 ...... 0.97 ...... 0.96 ...... 0.99 ..... .99

GFS

vs MB .......... 0.15 ....... 0.05 ...... 0.06 ......-0.22 ...... 0.08 ...... 0.19 ...... 0.12 ......-0.07 ......-0.12 .....-0.08

GFS

vs RJS ......... 0.53 ....... 0.36 ...... 0.60 ...... 0.56 ...... 0.52 ...... 0.40 ...... 0.44 ...... 0.39 ...... 0.44 ..... 0.41

ECM

day to day ..... ----- ......-0.19 ...... 0.59 ...... 0.64 ...... 0.85 ...... 0.88 ...... 0.92 ...... 0.98 ...... 0.97 ..... 0.98

ECM

vs MB ..........-0.79 .......-0.15 ......-0.20 ...... 0.48 ...... 0.32 ...... 0.12 ......-0.15 ......-0.09 ......-0.00 .....-0.09

ECM

vs RJS ......... 0.07 ....... 0.71 ...... 0.31 ...... 0.17 ...... 0.36 ...... 0.37 ...... 0.43 ...... 0.38 ...... 0.38 ..... 0.44

ECM

vs GFS .........-0.37 ....... 0.65 ...... 0.55 ...... 0.81 ...... 0.94 ...... 0.96 ...... 0.92 ...... 0.99 ...... 0.97 ..... 0.99

See you again tomorrow when we'll have some interesting comparisons of major model 10-day forecast performance vs reality and the two experimental forecasts thrown into the mix.

Edited February 2, 2011 by Roger J Smith

[Coast]

http://www.wunderground.com/global/stations/10149.html

http://www.elbe-webcam.de/elbblick

http://www.umwelt.sachsen.de/umwelt/wasser/7806.htm

[Coast]

Regarding the thunderstorm milestone for today in the forecast, here is what ESOTFEX have forecast:

Storm Forecast

Valid: Thu 03 Feb 2011 06:00 to Fri 04 Feb 2011 06:00 UTC

Issued: Wed 02 Feb 2011 20:52

Forecaster: GATZEN

SYNOPSIS

A deep trough is located near Greece and will slowly move eastwards on Thursday. To the north, a strong westerly jet expands across northern Europe.

DISCUSSION

Scotland, North Sea, and south-western Scandinavia

A region of interest will be north-western Europe. A strong westerly flow will affect this area on Thursday. Late in the period, another, very strong mid-level jet streak is forecast to reach more than 60 m/s at 500 hPa over the northern British Isles. A frontal wave will be located to the east of this jet streak and will rapidly move eastward. Current scenario is that the moist low-level air mass will be not overlap with the deep lift near the jet streak axis, where quite cold mid-levels are forecast to lead to CAPE. So only a shallow line of convection may develop along the cold front and only sporadic lightning is forecast if any. Nevertheless, this strongly forced line may produce severe wind gusts and tornadoes given a bulk shear up to 20 m/s in the lowest kilometre. An upgrade to level 1 may be needed when sufficient instability will develop along the cold front.

To the west of the cold front, severe synoptic winds are forecast with the sting jet of the low. Weak instability in the cold air mass will likely assist for some showers that may lead to a mixing of the high winds that will be about 45 m/s at 850 hPa to the ground.

[Coast]

Hornsea wave data:

http://www.channelcoast.org/data_management/real_time_data/charts/?chart=72

Buoy 62164 (The North Sea Station) current data:

http://www.ndbc.noaa.gov/station_page.php?station=62164&unit=M&tz=STN

[MurcieBoy]

^ My word, I never knew where to look for such detail...

Back to the comparison charts for today at 6z and actuals:

3 Feb

4 Feb

5 Feb

Actuals:

28 Jan

30 Jan

1 Feb

2 Feb

[peterf]

Hornsea wave data:

http://www.channelcoast.org/data_management/real_time_data/charts/?chart=72

Buoy 62164 (The North Sea Station) current data:

http://www.ndbc.noaa.gov/station_page.php?station=62164&unit=M&tz=STN

superb info Coast,thanks very much peter.

[Coast]

My word, I never knew where to look for such detail...

In my simple way, it's the best method I can find of graphically showing comparisons for people to analyse, I hope you feel they are fair and representative too.

Current rain radar:

Webcams around the North Sea:

Happisburgh, Norfolk

http://www.bbc.co.uk/norfolk/coast05/coastcams/

Gabriel Lunds gate, Farsund, Norway

http://www.webcamsinnorway.com/webcams.php?viewcam=677

Ostend, Belgium

http://www.rnsyc.be/ostend/webcam.asp

superb info Coast,thanks very much peter.

As above, I hope it is fair and gives everyone something constructive to use in their summaries.

Edit - God, does all of this stuff make me ever so slightly geeky???? :lol:

[TonyH]

God you know your stuff coast!

Great charts, graphs, cams, etc.

[Coast]

God you know your stuff coast!

Great charts, graphs, cams, etc.

Not really, I've just saved a lot of links in My Favourites in the last few years!!! :lol:

[MurcieBoy]

In my simple way, it's the best method I can find of graphically showing comparisons for people to analyse, I hope you feel they are fair and representative too.

When I made the 12 Nov 2010 prediction, I never knew GFS or ECM forecast pressure charts existed (let alone wind or jetstream charts); indeed, I didn't even know where to go to get detailed historic weather data.

The analysis and support from all on NW (in particular RJS, Coast, OldMetMan) in analysing the Storm Forecast has really helped a great deal in this regard.

I am also looking forward to the end of the forecast period, so one can just step back and see what actually has happened (in overview terms) during 1-5 Feb 2011; and analyse closely the variances with the actual.

Edited February 3, 2011 by MurcieBoy

[Coast]

The analysis and support from all on NW

Does that get us a go in the LP700-4 when it turns up then?

[MurcieBoy]

Does that get us a go in the LP700-4 when it turns up then?

Well, lets just say, I live only about 5 miles south from the A27!

[Coast]

Well, lets just say, I live only about 5 miles south from the A27!

Giallo Evros?

12Z analysis is out from the Met office. Again I've overlaid the forecast to see what matches up:

Apart from the position of a low between Sweden and Finland I can't find any pressures, fronts or isobars that really correspond today. The general storm predictions in the forecast also seem to have missed, but let's give it until the end of the day to see what happens with the thunderstorms prediction.

http://www.skystef.be/obs-lightning.htm

Worth keeping an eye on this thread, just to see what a real Atlantic storm can do in Scotland:

[Roger J Smith]

Comparisons of Forecasts and Actual map for 12z 3 Feb 2011

__________________________________________________________________

The locked-in nature of the GFS and ECM since about day five has reduced the suspense in this operation considerably, but with the actual map now in the data vault, we can go back and see how the two main global models fared in the ten day period to validation, and how the two experimental forecasts from further out (22 to about 40? days) compared.

I would refer the reader back to my previous post for the validation of the experimental forecasts with various prog charts in that period.

First of all, a brief map description for the actual. In general, it represents fast WSW flow with a semi-permanent low of about 990 mbs in central Sweden for several days, a fast-moving developing low northwest of Iceland with a map-time pressure near 970 mbs, and a parent low northeast of Iceland moving slowly northeast. A ridge of high pressure that was centered near the Azores ridged into central Europe. The flow between 50 and 55 deg north was mostly from a SW to W direction from the western edge of the map grid to the eastern edge.

The correlation of the actual with the MB experimental map was -0.14. This represents a slight "anti-correlation" but most of the negative sign can be attributed to the overall upper flow being near right angles to the actual pattern. The RJS experimental (which was issued later at day 22) has correlation of 0.43. If one postulated a random nature to the RJS forecast program, this correlation would not be exceeded more than five times out of a hundred. My subjective opinion is that this correlation is about normal for the program in its current stage of development, and points to some large-scale pattern correlation ability, although I would think that if the basis for the pattern had been fast flow and not block break-down, the correlation might have been a lot higher, as timing of features was good.

Both maps hit on their identified primary low pressure, albeit with large intensity errors and track direction errors. The MB experimental map correctly placed a low in central Sweden but had it 40 mbs too deep and moving at about 300 deg whereas map analysis showed a slow drift from 240 deg. The RJS described a 1002 mb low west of Donegal Bay tracking ESE, in reality this was an explosive 970 mb low tracking ENE to NE, but position at map time was within 3 deg of lat and long (wave position on map used was identical).

Now, looking at the performance of the two major global models from day 10 to day 1, the GFS had a good fix on the outcome from day 10 and probably from about day 12-13. The correlations rose from 0.58 at day 10 to 0.87 at day 8, both higher than average for this statistic, and then somewhat randomly fixed onto the eventual 0.98 correlation at day one. The ECM was nowhere near at day 10 with a correlation of 0.08, but outperformed the GFS on days 9, 5, 4 and 3. The day 4 ECM was actually the highest correlating map in the series at 0.99 (tied with day 1 ECM).

For the record, here are the comparisons of these ten-day statistics (all positive correlations in .00):

MODEL .. d10 ... d9 .. d8 ... d7 ... d6 .. d5 ... d4 ... d3 .. d2 ... d1

GFS ..... 58 ... 67 ... 87 ... 87 ... 85 ... 88 ... 92 ... 96 ... 98 ... 98

ECM ..... 08 ... 73 ... 82 ... 72 ... 79 ... 91 ... 99 ... 98 ... 97 ... 99

-----------------------------------------------------------

Now a few added notes on the forecast validation. Today, there have been strong winds already in northwest Ireland and westerrn Scotland. A gust to 72 knots was reported from Belmullet around 14z. Public reports speak of high waves and a developing minor storm surge in Galway Bay with the evening high tide (new moon was earlier today at 0232z). Western Scotland has reported wind gusts to 77 mph at South Uist and will probably peak around 90 mph later. This wind peak is likely to continue to track northeast and hit parts of Norway later but progs suggest that it will dissipate somewhat. A second wind peak is expected with a wave tracking through similar regions later Friday. Most of this activity is directed more towards Norway than the North Sea although it will get moderately windy at times in the places warned of severe winds in the MB forecast.

Today's 12z Shetland 500 mb reading was WSW and the wind speed at Valentia was about 80 knots at 500 mb (a moderately high value, not considered extreme). Further west a 250 mb peak wind of 161 knots was reported over the east-central Atlantic. Once again, this is moderately high for time of year but not exceptional. I will check the 00z map for expected slightly higher values of these statistics for the overall validation. Today's low position check awards max points (4) to both experimental forecasts after RS(0,0) and MB(0,1) the first two days. Can already say the likely points for tomorrow are 1,1 then 0,0 for day 5.

In general this larger validation will likely prove inconclusive -- I will make one final post on the 6th with a roundup on this -- but the main point emerging from the MB validation is a successful large-scale hit on pattern intensity, especially if we factor in a subjective error that was widely spotted by forum readers from the maps provided. My earlier point that strong winds in the region usually come from a source to the WSW has been illustrated here as well. With that one relative success, I would encourage MB to keep working on his method and despite the rather stark differences in outcome for specific locations, not to concentrate on that alone but to take some encouragement from the more general successful indication.

Edited February 3, 2011 by Roger J Smith

[AtlanticFlamethrower]

Thanks RJS and, Coast. RJS - is there any way you can simplify your analysis - or provide a simply paragraph summary? I can do correlations but when I read the thread in the evening my brain is often half-switched off and I can't give what you're saying the attention it warrants. Interesting that you think MB should keep at his method. Is there going to be another test date, I wonder?

Oh - MB's forecast might not have gone as precisely hoped but it's windy outside... for the first time for quite a while. It's not a Great Storm but it's a little bit spooky!

[MurcieBoy]

Thanks again for the analysis RJS and for the overall "large scale" point. I am going to take a good few days after the end of the forecast period and produce a very candid appraisal video (which I hope will bring out some interesting points for discussion/reflection).

AF, yes I have a couple of extreme events forecasted for spring/summer this year (both in the USA). I shall formally publish once I have had time to digest the Feb Storm forecast and see what refinements I need to make to the US forecasts. I really do need to get back to some more research (as I have hardly done much "stepping back" thinking since September 2010 - the "variances" in the current storm forecast have given me a few things to think about!).

[peterf]

Thanks again for the analysis RJS and for the overall "large scale" point. I am going to take a good few days after the end of the forecast period and produce a very candid appraisal video (which I hope will bring out some interesting points for discussion/reflection).

AF, yes I have a couple of extreme events forecasted for spring/summer this year (both in the USA). I shall formally publish once I have had time to digest the Feb Storm forecast and see what refinements I need to make to the US forecasts. I really do need to get back to some more research (as I have hardly done much "stepping back" thinking since September 2010 - the "variances" in the current storm forecast have given me a few things to think about!).

cheers Nick,very interesting indeed.

[TEITS]

Well in my opinion the forecast was a failure. Doesn't matter if its windy outside with a deep LP system to the N of Scotland. The fact is we do not have +100 NW,ly winds and synoptically the forecast was incorrect.

I certainly hope nobody is suggesting this forecast was correct. A LP system between Iceland/Scandi bringing gales in the winter is hardly notable and remember MB said this storm would be inferior to 1953.

Im not mocking MB but here is a prediction. On the 15th Dec 2011 we shall see HP over the UK. On the 15th January 2012 we shall see a deep LP system to the N of Scotland bringing windy weather. Now in my opinion my forecast has a 50% chance of being correct!!

What put me off this forecast is how sensational it was and reminded me of the type of forecasts Piers Corbyn & the Daily Express make.

Edited February 4, 2011 by THE EYE IN THE SKY

[Coast]

Just for the record, here is yesterdays thunderstorm activity. It's hot-linked so after today (04/02/11) it will revert back to whatever the previous 24 hours are. If you are reading on the 4th Feb, you will notice the thunderstorms were confined to Scotland and came from a Westerly direction. Last but one forecast day today and I'll put the 12Z MetO fax comparison up as soon as it's released after lunch.

I think my overview will be on Monday.

[Weather-history]

What put me off this forecast is how sensational it was and reminded me of the type of forecasts Piers Corbyn & the Daily Express make.

Same here and the detail as well. No one can predict such detail that far out. For this forecast to have been a success then the detail had to be correct, getting the general pattern right was not good enough because a general pattern was not predicted, specific weather patterns at specific dates were predicted. I think its nigh on impossible to do this at so far range to say I expect 80mph winds on the 17th March 2011 across Scotland, Northern Ireland and northern England because there are so many variables in between now and that date that it only requires a slight tweek of one of these and the low pressure is weaker, delayed, came earlier or didn't come at all even if you are using scientific analysis.

Edited February 4, 2011 by Mr_Data