January 2021 C.E.T. and EWP forecast contests

[BornFromTheVoid]

Latest projections and probabilities

 

At the moment, we have 91.3% chance of finishing between 1.5C and 4.5C

The chances of finishing:
Close to average (3.9 to 4.9C) is  to 22.6% (2 days ago was 22.2%)
Above average (>4.9C) is  to 0.8% (2 days ago was 2.8%)
Below average (<3.9C) is  to 76.6% (2 days ago was 75.0%)

No projections finish below 0C.

[Froze were the Days]

9 minutes ago, BornFromTheVoid said:

Latest projections and probabilities

 

At the moment, we have 91.3% chance of finishing between 1.5C and 4.5C

The chances of finishing:
Close to average (3.9 to 4.9C) is  to 22.6% (2 days ago was 22.2%)
Above average (>4.9C) is  to 0.8% (2 days ago was 2.8%)
Below average (<3.9C) is  to 76.6% (2 days ago was 75.0%)

No projections finish below 0C.

On that basis my assumption that we struggle now to see a winter month below -1c anomaly of the 61-90 CET series looks like continuing...I went for 4.2c but might be a bit high...time will tell.

[damianslaw]

A rise expected over the coming days, so we could be around the 3 degree mark in a week's time, thereafter, hard to know what will happen in the last third, but the suggestion is for a colder trend, but may be a taller order than was looking just 3 or 4 days ago of recording a notably colder than average month say something below 1 degrees, but still within realms of a finish in the 1 to 2 degree range which is preety cold by any standards. Notwithstanding I would say a below average month is now almost a cert unless we see a major flip in the longer term trends.. (I may be eating my words I hope not!).

[Roger J Smith]

This is a rather simple calculation for the CET, but if we reach 3 C by Tuesday 19th then with 12 days to go, if the cold outbreak takes up 70% of the time remaining, and modified Atlantic air masses similar to what we've seen recently take up the other 30%, then that would be like 8 days near an average of zero C and four near 6 C, so, 19x3 is 57 degree days, add another 24 (4 x 6), that's 81 degree days and a mean CET of 81/31 or 2.65, let's say 2.5 after corrections. If you think the cold will be deeper and/or longer lasting, then a rule of thumb might be to reduce the 2.5 by 0.1 per 2 extra days with extra degrees of cooling applied to the number of days you think that might apply, so if you think for example, 9 days that average -1 instead of my suggestion, the calculation changes to 57 + 18 - 9 = 66 / 31 = 2. 15 which probably ends up around 1.9 or 2.0. 

From the current model run this looks like a warmed up version of Jan 1784 unfolding,  SST values back then were no doubt 1-2 C lower than today, which probably accounts for all phases of the month running 1-2 deg colder and ending up -0.6 (after a three-day mild spell mid-month boosted the running average to 1.0). 

I think today, even if we got the exact same synoptics as a Jan 1740 or 1795, we would come in just -1.5 or so, those air masses would be carrying extra surface warmth from crossing warmer oceans and also the jet stream nowadays probably cannot drop quite as far south as it may have done when the whole atmosphere was slightly cooler. It all makes Dec 2010 seem that much more remarkable, somewhere else on this forum I showed that running 31-day anomalies were coldest of any year (since 1772) in 2010 from about the 23rd Nov-23rd Dec stat to 30th Nov-30th Dec, the only day that 1890 led was the 31st. Before 2010 moved into the lead, 1878 held a running 31-day lead for a time ending when 2010 picked it up. 

Also remarkable would be the fact that Jan 1963 was within 0.4 of 1795 and rival 1814 at some point around the 26th-27th, dropping back only a bit to finish a degree or so behind them. 

(this is a link to the thread about cold 31-day intervals -- what if we had no calendar and all 31-day averages were compared? the list of coldest "months" would be similar to what we find familiar, but there are some intruders that people don't know were there, such as mid Dec 1860 to mid Jan 1861, the two months each had CET values in the 1-2 range but a 31-day interval averaged -1.4 and was the t-14th coldest with calendar month Jan 1940 which could not be improved with any other days involved. In fact, it's rather interesting how many of the top cold spells just about maxed out in their calendar months, but another example of a mid-month to mid-month player was in 1917 (15 Jan to 14 Feb, --1.2 C). 

The same is true of summer warmth of course, there are two 20+ "months" in the CET records, in 1976 and 1995, if you toss the calendar, but July 2006 had no way of improving on its status as warmest calendar month. 

____________________________________________

link to thread about cold hidden "months" 

https://www.netweather.tv/forum/topic/92395-beyond-the-calendar-uncovering-the-hidden-cold-months/?tab=comments#comment-4081977

 

Edited January 15, 2021 by Roger J Smith

[Roger J Smith]

Note also for future reference, the coldest 31-day interval for this winter may then be 28 Dec to 27 Jan if it turns milder around the 28th, as all four days at end of Dec 2020 were 1.0, 1.0, 1.0, -1.3 (1.7 deg days total), so if that same four days of January were to average 5 C let's say, then the January mean, whatever it turns out to  be, can be lowered by 0.6 (20 deg days replaced by 1.7, 18.3 fewer, will reduce the average by 18.3/31). 

When you think about it, the twelve calendar months are meaningless divisions of time, the reason we have 12 is that most ancient calendars were lunar, but as societies evolved they realized a need for an annual reset at similar climatic intervals, so they added extra calendar months, since the lunar month is 29.53 days on average, 12 of them last for 354.36 days. It would have been just as easy to say we'll have 13 months of 28 days, except that every February (or whatever) would be 29 days and every leap year would be 30 -- that would have led to the exact same problem being sorted out by Pope Gregory in the 16th century (10 extra days elapsed since Roman times due to the faster motion of the earth around the Sun, in 365.242 days rather than 365.25 as envisaged in the model of adding an extra day every fourth year). Other than that, the twelve calendar months have a vague symmetry with the twelve zodiac periods, all of which are approximate rather than exact, and also they drift against the calendar. I'm not into astrology personally but I realized in my studies of precession that the zodiac signs are now a month later (in terms of when the Sun is moving through constellations) than astrology says they should be. That's because astrology developed in pre-Roman times when the Sun was in those constellations in the periods ascribed to them. Nowadays, the Sun reaches each constellation a good five weeks later (the 26,000 year precession cycle will take this further and further into the lateness realm until at some point the zodiac signs will be opposite in the calendar from where they are supposed to be). If you're having trouble visualizing that, I'll use Taurus which is supposed to run from about 22 April to 21 May. Taurus is the constellation over the right shoulder of Orion (as we see Orion, it would be his left shoulder since he's looking at us). The Sun reaches the summer solstice over top of Orion which is why it features prominently in our mid-winter sky. ... So the Sun nowadays moves through Taurus from about the 24th of May to the 23rd of June before it reaches Gemini.  It's a full month and a couple of days later than an astrologer practising in pre-Roman times would have determined (you can only "see" it at total eclipse, and even then not easily because not many stars become visible at total eclipse as I found out in Oregon in 2017, I could only see three bright planets and one bright star at totality). Maybe at a longer totality (my encounter was only 2 min) more stars appear, but while it's quite dark at totality, it isn't pitch black like midnight, more like how it appears about a half hour after sunset on a clear summer's evening. 

[The PIT]

Sunny Sheffield still at 1.8C -2.7C  below average, Rainfall 56.1mm 69.3% of the monthly average.

[Roger J Smith]

EWP update ... 37 mm to 13th, about 5 mm additional on 14th for 42 mm to date. 

GFS 00z guidance gives 30 mm on average through 25th 00z, maps for 26th-31st suggest potential for another 20-30 mm rain then.

All told, it adds up to 92-102 mm totals, not that different from my 88.8 mm scoring file estimate ... so I won't update that yet. By the way, if you already looked at the excel file (posted a couple of days ago) I've replaced it with a more current version, changes are mainly those second-order scoring adjustments I added to the procedures, so everyone dropped a fraction (those between 10th and 70th slightly more than the people at either end). 

The lowest outcome I could imagine verifying now would be 65 mm and the highest 130 mm, chances look like this (to me) ...

65-80 mm __ 25% chance

80-95 mm __ 25% chance

95-110 mm _ 25% chance

110-??? mm _ 25% chance

_

LG did not submit his usual 200 mm forecast, not sure if he meant to and forgot, but in any case unless the outcome is >110 mm that would have scored zero as second highest error with a late penalty. He was in fourth place after Dec but with or without that forecast would probably end up middle of the annual standings after January. Based on 88.8 mm, snowray had taken the lead in the annual as well as being top score for Jan (89 mm forecast). Of course that can only go down and not up, so somebody else may be our leader by end of the month. The current rest of top 12 from my estimate of 88.8 mm (annual or two-month total) would be JeffC (85), feb1991blizzard (80), Federico (80), Reef (97), Midlands Ice Age (90), Blast from the Past (75), Bobd29 (78), summer18 (70), shillitocettwo (105), Godber1 (75) and Don (65). Their forecast values are shown to indicate in which way the standings might shift for different outcomes. December leader SteveB sits 16th (57 mm) but would be into the top ten if the outcome is below 80 mm, as would Born from the Void (t18th) with 62 mm. Virtualsphere (103 mm) in 13th can move up well into the top ten if the outcome is above 100 mm. Our consensus was 68 mm so at the moment normals (93, 94.2) are doing better it would seem.  The two normals would be in the top five of our contest at 88.8 mm and probably at any other value between 85 and the upper limit as there aren't a lot of higher forecasts and some of them are from first-timers with no December score to add. 

[Roger J Smith]

Long way to go, but if the GFS was essentially correct to end of month, CET would finish around 2.0. This is a rough matrix of all forecasts in the region of 1.0 to 2.8 CET that have an EWP above 50 mm. The centre of this matrix is where I currently estimate "best combined forecast" to lie ... snowray can hit a double at 2.0 since he was into the CET contest before Kentish Man. At 1.9 though, snowray can't do the double first because the two forecasts at 1.8 were both in before. Of those two, Earthshine (60 mm) can have a double first being first in at both 1.8 and 60 mm. None of the 2.1 forecasters can hit a double first since Typhoon John (no EWP) was first in on the contest. (I suppose really I should only compare CET of people with an EWP forecast in this best combined ranking, hasn't been a factor much in the past, but I just take the CET order of merit as is). Anyone with a solo CET punt can obviously make it to double first by hitting their CET number on the button. 

some EWP forecasts are almost identical, there's an & symbol when that happens, and in all cases, the EWP value is about where the first letter of any name appears (first example being 62 mm for wx-history and 63 mm for davehsug both at 2.8 CET. (not shown, consensus 2.8 68 mm).

Note added Jan 18 _ Expanded the table upward to 3.5, moved the target to 2.6 130 mm.

At the end of any CET row, are any other forecasts not in matrix range or lacking an EWP component.

 

CET/ __ EWP>> 60 __ 65 __ 70 __ 75 __ 80 __ 85 __ 90 __ 95 __100 __105 __110 __115 __120 __125 ___150 

3.5 ...........SteveB star&J10.............................JeffC............................................................................................... 

3.4 ........................................................................................................................................................................... (no fcsts)

3.3 ........................................................................................................................................................................... DLAW no EWP

3.2 ............................................................................................................................................................................ (no fcsts)

3.1 ........GC...................................................................................................shil......................................................

3.0 ..................................non........stew................................................................................................................... B87 30mm

2.9 ...........................Don..MrM&sum18&Mul&Tim.Norr............................................................................ DWW,SumSun no EWP

2.8  .......................wx-h&dave ..............................................................................................................................

2.7 ...........................................Feb78 ....................................................................................................................

2.6 ................................................................................................................................................................X..... (MWB no EWP)

2.5 ...............................The PIT ............................................................................................................................ 

2.4 ........................................2010cold .. feb91 .................................................................................................... (Q89 no EWP)

2.3 ....sundog ..........................................................................................................................................................

2.2 .............................. sea ..... GrPlum ..................................................................................................................

2.1 ...........................BFTV ........boof&Dunc&BFTP................................................................................................ (TyJohn no EWP)

2.0 .............................................................................snowray ................................................................................. (Kent no EWP)

1.9 ............................................................................................................................................................................. (no fcsts)

1.8 ...........DT....Earth................................................................................................................................................

1.7 ............................................................................KirkWx......................................................................................

1.6 .............................................................................................................................................................................. (no fcsts)

1.5 .........................................................Godber1.......................................................................................................

1.4 ......................................RJSmith ...........................................................................................................................

1.3 .....................................NeilN ...moor .................................................................................................................. (SMU no EWP)

1.2 .dan.............................LIS .........................................MIA ..................................................................................john.. 

1.1 ..............................jonboy .............................................................................................................................................

1.0 .................................................................................................................................................................................... (no fcsts)

0.5-0.9 ...........Paul&WWS ... EG .................SFront .....................cymro....................................................................... (CSW 25 mm)

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

It might be interesting to draw a map of the British Isles (and France) and show where each forecast in the whole set actually verified. 

I may do that because this is quite an interesting month for variations. 

 

Edited January 18, 2021 by Roger J Smith

[Summer Sun]

2.3c to the 14th

1.2c below the 61 to 90 average
2.1c below the 81 to 10 average

___________________________

Current high this month 2.3c on the 14th
Current low this month 0.5c on the 1st & 2nd

[Man With Beard]

7 hours ago, Summer Sun said:

2.3c to the 14th

1.2c below the 61 to 90 average
2.1c below the 81 to 10 average

___________________________

Current high this month 2.3c on the 14th
Current low this month 0.5c on the 1st & 2nd

Little chance of ending above 4C from here

[feb1991blizzard]

1 hour ago, Man Without Beard said:

Little chance of ending above 4C from here

Little chance of it ending up at 2.4c though either, i am gutted i didn't stick with my 3.6c or whatever it was that i got up early, my 2.4 sure to bust on the low side now.

Birmingham - BBC Weather

WWW.BBC.CO.UK

---

14-day weather forecast for Birmingham.

Always find this is a good guide, average to slightly above for rest of the month mainly.

[Frigid]

39 minutes ago, feb1991blizzard said:

Little chance of it ending up at 2.4c though either, i am gutted i didn't stick with my 3.6c or whatever it was that i got up early, my 2.4 sure to bust on the low side now.

Birmingham - BBC Weather

WWW.BBC.CO.UK

---

14-day weather forecast for Birmingham.

Always find this is a good guide, average to slightly above for rest of the month mainly.

I doubt we will go above the 61-90 average for January. 2.5 to 3.0 will be the most likely outcome granted we get the cold last week including a downwards correction. However it does look very uncertain for the next 14 days so who knows we could end up close to average or considerably below. 

[Roger J Smith]

EWP is going on another rampage according to the GFS. 

42 mm in the barrel by 14th, added perhaps 4-5 mm on 15th (20-25 mm rainfalls spread over southwest England and south Wales, much less further east). From the 00z GFS ten-day projection (which would include all of today's rain after midnight) at least 50 mm more will fall on average by the 25th and then from charts 25th to 31st another 30 mm would not be surprising. 

That 50 mm ten-day is somewhat conservative as large percentages of the grid are well above that, so let's say 50 to 70 ... the totals derived from the above range from 126 to 147 mm. (yikes)

I guess I will be redoing my scoring file unless later model runs start to emphasize more dry cold and less intrusive rain/snow events. That might return the target to below 100 mm. 

We had one 300 mm forecast which will finish last anyway unless this month gets above 163 mm, then it would gradually climb a few spots, by 180 mm it is moving into the top third of scoring. Otherwise, these are the forecasts higher than normal (93 mm) ... Polar Gael 151.3, Captain Shortwave 150, moffat 130, SLEETY and DAVID SNOW 115, Relativistic 110, shillitocettwo 105, virtualsphere 103, DR(S)NO 100, Cymro, Reef 97.

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

Meanwhile, the CET from latest guidance looks to be settling in around 3 C, could see it reaching that by Tuesday 19th, then back down to around 2.4 by end of the week, up and down to end of month, more up than down ... but that could change if we get back into longer cold outbreak scenarios. 

The missing 5.9 C from December's ranks is well known, but here's a perhaps even more unlikely outcome, in 363 years, only one January hit 3.1 C (that was 1832). Usually the values near the median of all observations in this large data set have plenty of ties, and the mean for January (all CET data) is 3.3, the median 3.5, but only one year managed 3.1. So if this year joins it, that will be interesting. Of course it's not really 363 years that could have hit 3.1, as almost all the 1659 to 1710 means are multiples of 0.5, but to show how odd the frequency is, this is the number of Januaries that ended up at 2.8 to 3.5 ... 8, 6, 5, 9, 1, 6, 6, 9, 12. In the cold direction, the first instance of a lone wolf is 0.2 (1895) and the first missing value for January in the cold direction is -0.5. In the mild direction, the first lone wolf was 6.1 (1989) and the first missing mild outcome is 7.2 C. Looking at February and December, there were 7 Febs and 6 Decs that ended up 3.1, and even March had 3 while November had only one (not surprisingly as that was 6th coldest 1851). 

 

Edited January 16, 2021 by Roger J Smith

[The PIT]

Sunny Sheffield at 1.7C -2.8C below average. Rainfall 61.5mm 75.9% of the monthly average.

[Summer Sun]

2.3c to the 15th

1.2c below the 61 to 90 average
2.1c below the 81 to 10 average

___________________________

Current high this month 2.3c on the 14th & 15th
Current low this month 0.5c on the 1st & 2nd

[BornFromTheVoid]

Latest projections and probabilities

 

At the moment, we have 91.5% chance of finishing between 1.5C and 4.5C.

The chances of finishing:
Close to average (3.9 to 4.9C) is  to 32.3% (2 days ago was 22.6%)
Above average (>4.9C) is  to 0.4% (2 days ago was 0.8%)
Below average (<3.9C) is  to 67.3% (2 days ago was 76.6%)

No projection finishes below 0C, and only 1 goes below 1C.

[The PIT]

Sunny Sheffield still at 1.7C -2.8C below normal. Rainfall 62.9mm 77.7% of the monthly average

[damianslaw]

Half way juncture of the month, a cold first half that may now be cancelled out somewhat by a mild second half.. not seeing any signs for anything substantively below average in the CET zone in the outlook, indeed looking fairly mild with limited frost due to wind and rain. Further north, colder weather looks like returning second part of coming week, but may not make much inroads to much of the CET zone.

Still 2 weeks to go and we may see a flip to colder weather more widespread by last week, but probably unlikely. 

With this, expect a continued rise in CET over coming days, will most likely be in the 3s fairly soon. A final finish in the 3s looks a good bet, where I'm not sure.. A near average month now most probable rather than notably below but still think good chance of coming in below the 61-90 average.

This month had very high potential to be our first notably cold January since 2010, alas not meant to be. Still it's refreshing to be experiencing a January where the odds of below average are higher than above, not really been able to say that since 2010. In 2013 the cold arrive a bit too late.

[Summer Sun]

2.4c to the 16th

1.1c below the 61 to 90 average
2.0c below the 81 to 10 average

___________________________

Current high this month 2.4c on the 16th
Current low this month 0.5c on the 1st & 2nd

[Don]

2 hours ago, damianslaw said:

Half way juncture of the month, a cold first half that may now be cancelled out somewhat by a mild second half.. not seeing any signs for anything substantively below average in the CET zone in the outlook, indeed looking fairly mild with limited frost due to wind and rain. Further north, colder weather looks like returning second part of coming week, but may not make much inroads to much of the CET zone.

Still 2 weeks to go and we may see a flip to colder weather more widespread by last week, but probably unlikely. 

With this, expect a continued rise in CET over coming days, will most likely be in the 3s fairly soon. A final finish in the 3s looks a good bet, where I'm not sure.. A near average month now most probable rather than notably below but still think good chance of coming in below the 61-90 average.

This month had very high potential to be our first notably cold January since 2010, alas not meant to be. Still it's refreshing to be experiencing a January where the odds of below average are higher than above, not really been able to say that since 2010. In 2013 the cold arrive a bit too late.

I think my 2.9C guess is probably the lower limit of where we finish this month and something in the 3's most likely ATM as you say.

Edited January 17, 2021 by Don

[Frigid]

3 minutes ago, Don said:

I think my 2.9C guess is probably the lower limit of where we finish this month and something in the 3's most likely ATM as you say.

Looking at the outlook, it does look milder than average. Quite disappointing tbh since this month could've been the one with the -ve AO/NAO, appreciable amounts of northern blocking and the SSW. I guessed 2.9c as well, will take a lot to get near that figure by the looks of it. Just hoping for something better in Feb

[Don]

34 minutes ago, Frigid said:

Looking at the outlook, it does look milder than average. Quite disappointing tbh since this month could've been the one with the -ve AO/NAO, appreciable amounts of northern blocking and the SSW. I guessed 2.9c as well, will take a lot to get near that figure by the looks of it. Just hoping for something better in Feb

It doesn't look very mild, though, so I can't see the second half of the month cancelling out the first.  Yes, it has been very frustrating and I was almost convinced the country would see widespread snow before the end of the month but this not looking likely now.  As you say, hopefully February will make this winter for us!

[Roger J Smith]

Always hard to project when there are going to be days with a mild start and a cold finish, the provisionals will come in a lot warmer than the eventuals when you get days like that (Wednesday will be an example, can see a few others possibly down the road). I think there's a tendency for the final values to be adjusted more to calendar day where the provisionals pick up a daily minimum that will later be revised down. (morning low around 7, afternoon high 9, evening near 1, if you take average of 7 and 9 that's 8 but an average of 9 and 1 only 5 -- right there a 0.1 C difference in the monthly average at end of month). Do that three times and you're out by 0.3 ... will say this, the estimated CET will almost certainly get hammered down by at least 0.3 and maybe more this month, whatever values are involved. Best guess would be 3.0 down to 2.5 at this point. 

[damianslaw]

Yes I'm not expecting a particularly mild second half to January, but it will feature much milder weather at times compared to much of the first half, at least in the southern portion of the CET zone. Tomorrow widespread maxima in this part of the zone in double figures. There could though equally be punctuations of much colder weather which may have the effect of cancelling out the milder conditions, but on balance expecting the milder days to cancel out the colder ones at least in the southern part of the CET zone. Further north and in Scotland a cold January now almost certainly on the cards, much of Scotland this coming week will be below average, maxima not likely to get much above 4 degrees for many, and further sub-zero nights especially from mid-week onwards. End of the month maps will make for interesting viewing, expect a classic cold to warm north to south pattern.

[DAVID SNOW]

Worth remembering that 2 out of 3 CET stations are in the south of England, this is where the cold/mild boundary  will dictate the final CET value.

Very hard to predict ATM.