December 2022 C.E.T. and EWP contests _ start of a new contest year

[LetItSnow!]

23 minutes ago, cheeky_monkey said:

December came in here with a mean of -17.5c which is -8.6c below normal..did not get above 0c at all during the month.

How did 2022 come out overall? IIRC you guys had had some quite warm weather during the autumn. I know in your part of the world you've actually managed some cooler than average years in the past 5 or more years.

[cheeky_monkey]

7 minutes ago, LetItSnow! said:

How did 2022 come out overall? IIRC you guys had had some quite warm weather during the autumn. I know in your part of the world you've actually managed some cooler than average years in the past 5 or more years.

first six months of the year every month was just below normal but nothing out of ordinary..Jul - Oct were very warm and dry .. Nov and Dec were quite a bit below normal 

[Roger J Smith]

I am not seeing the usual table update on the 5th, so I can't finalize the EWP scoring yet. Check back later or tomorrow. Normally this value pops up before noon in the table, if not earlier, so this is unusually late for them. I added up the daily values that are listed now (the tracker got to 101.6) and those also add up to 101.6 mm. 

[Guest]

On 29/11/2022 at 21:50, DCee said:

When the comp is closed can you take a average of all the estimates (CET and rainfall) and post that please? 

It'll be interesting to see if the average of all the estimates are near the measured values....

As I mentioned earlier in the thread the average estimate of all our guesses is spot on! 

The accuracy of prediction when using a knowledgeable dataset/forum is impressive!

[Guest]

On 30/12/2022 at 22:57, snowray said:

I noticed that the consensus of the 91 forecasts was in fact 3.5c, quite incredible.

Indeed, as I predicted it might be. See above post.

I often wonder why we dont develop a model which simply takes all the day to day runs and does a rolling average on them all to produce a single model run. Updated at the end of each day....

[Roger J Smith]

Our CET consensus over the past three years has an average error of 0.8 C which is a slight improvement over random (the average difference of outcome from recent normals). This past month was closer than most, but we have been as much as 2 deg out on some other occasions. 

Still waiting for the EWP confirmation. Our consensus there was not so good (63 mm vs 101.6 outcome). As a group we have shown less skill with EWP than CET when compared to the performance of normal values, which actually won the December contest. (DCee is currently high human). 

[Roger J Smith]

Still waiting on the EWP update, might be a few days now with the weekend, or could happen tomorrow. It's not just this monthly (and therefore annual) update that is late, also the tracker has not been updated since the 4th (to the 2nd, it is always two days late rather than the one day lag of the CET tracker). 

[Roger J Smith]

We finally have a table value for December EWP which is 102.0 mm (up 0.4 from the tracker).

This will mean some minor scoring adjustments which I will make back in the thread tables, it won't affect the top three or much else, but having a value that rounds off (like most forecasts) is going to create ties, so under EWP contest rules, nobody will lose points from being tied but some may gain. Differentials will be 0.13 points for most of those changes. 

102.0 mm is just slightly below the 1991-2020 normal value of 103.6 mm which was actually the closest "forecast" but among Net-weather forecasters,  DCee and Robbie Garrett were 1-2 with their 98 mm forecasts. Weather26 was third at 96 mm. Summer8906 was fourth at 109 mm and Bobd29 was fifth at 93.5 mm. Our consensus of 63 mm was not anywhere near as good as the temperature forecast consensus of 3.5 C. I suppose with the dry start being indicated, few believed that we would get as much rain as we saw from 17th to end of month. 

[Roger J Smith]

December was 89th wettest out of 257 Decembers, but even so, drier than all but three from 2009 to 2021 (those being 2010, 2014 and 2016). The year 2022 was 84th driest of 257 years with 857.2 mm, similar to 1989 (857.1 mm). Since 1975, there have been twelve drier years out of 47 in that span, namely 24th driest 1975 (758.7 mm), 2003, 1996, 2011, 1976, 1991, 2010, 2005, 1995, 1990, 1997 and 1989. 

Given that seasons do not quite overlap years, the contrast was stark between a dry spring-summer (42nd driest spring, 7th driest summer) and a wet autumn (26th wettest). The winter of 2021-22 was fairly close to average despite the dry January (it ranked 104th wettest) but this winter (2022-23) could rank a lot wetter, if February and the rest of January keep up the current pace, if we finish with 125 mm in January then it would only take about 100 mm in February to reach the same wet ranking of 26th as autumn managed. It would then take a very dry February to match the ranking of the previous winter (18 mm on top of the 125 estimate for January). To make this the wettest winter on record, 140 mm in January (a higher estimate) would then require 213 mm in February to match 2013-14 but 150 mm in February would be good for 6th wettest and 120 mm would be sufficient for about 15th wettest. So a continuation of trends since mid-December would leave this winter quite high in the rankings, remarkable when it is considered that the first half of December was very close to being driest on record.