Jump to content

Roger J Smith

Members
  • Content Count

    4,825
  • Joined

  • Last visited

  • Days Won

    13

Roger J Smith last won the day on October 13 2018

Roger J Smith had the most liked content!

Community Reputation

6,055

11 Followers

Profile Information

  • Gender
    Male
  • Location
    Rossland BC Canada
  • Interests
    forecast research on a global scale, thus an interest in UK weather as part of a larger research program that concentrates mainly on severe weather in North America.

Recent Profile Visitors

32,815 profile views
  1. I am going to attempt to rephrase an earlier (now removed) question in politically neutral terms that can surely cause no offense to any reader. What percentage of the warming since 1980 or any other defined recent starting point of the AGW signal is attributable to natural variability? If we agreed that the CET has risen about 1 C deg since a pre-AGW mid-20th century average value, then what would it be if there was no human activity on the planet? And what is the reasoning behind that conclusion? (Just for the record, I would state my opinion that the warming is somewhere between 50% and 80% natural in origin, so the human signal is 20 to 50 per cent, and without human activity the mean CET would be 9.8 instead of 10.2.) There is a related question that forms part of my own research focus. Does the AGW signal warm up all air masses proportionately? For example, are mild winter air masses warmed up by 1 degree as well as arctic air masses, or is it more that arctic air masses are warmed up 2 degrees and also we see them less often that causes the warming? (I have not finished the research ongoing to provide an opinion on this myself but could note that my study of Toronto temperatures over 180 years indicates that air mass temperatures are not changing very much, it is air mass frequency that has changed, and the biggest step function of that change was actually as far back as 1897 to 1915. This is something for which I can provide evidence when I publish the "Toronto 180" study that is in proof-reading stages now -- Toronto began recording in March 1840). Another question for you: What part of the AGW process creates simultaneous surges of warming in distant locations, namely, Toronto and central England? Why are record high temperatures often set in clusters around the same months in these two distant locations? What leads you to think that this is not evidence of natural variability having a powerful influence?
  2. It's just part of a general principle with temperature anomalies that they can vary from location to location by 1 to 2 deg. I've seen a lot of monthly reports of temperature anomalies and they often spread out over about that sort of range, so +1 here, -1 there, somewhat at random. It can be partly because Pit has already adjusted his data to the 24-hour calendar day while the CET preliminary data are being reported for intervals that will later be adjusted (which is why they often drop rather than rise since colder evenings get more involved in the overall picture). Also one would need to know what normal period the Pit uses as a base, if he's not using 1981-2010 but his station average over time, then we're not comparing anything since the CET might have a different anomaly value if compared to its own data for the same period of time. So maybe Pit can tell us when he reports a temperature anomaly, what's the period against which this month is being compared? I could then quickly work out what the CET anomaly for that same period would be. Also we need to know if his report is for a running mean 1st to yesterday's date or end of month -- I know his reports of precip anomalies are against end of month averages. For example with his rainfall report, at the last instance he says 19 mm, 21.9% of average, but that is for 11/31 of the month so in reality if we assume Dec precip falls at a constant rate, the 19 mm is (21.9 x 31/11) per cent of average which is 61.7% of expected amount to the 11th. If temp is similar, that creates a potential difference in itself since his end of month temperature average is probably 0.5 to 1.0 lower than his station 1st to 11th average. However, that would imply he should have a higher rather than a lower temperature anomaly (since CET compared to end of month is running closer to 1.6 above that normal).
  3. EWP update -- 18 mm to 9th, looks like about 6 more on the 10th (much heavier in north Wales and Cumbria again, but grid average probably 6-8). Then the ten-day GFS is a little drier now than in recent days, as the focus of heavier rainfalls seems to be further west into Ireland. Would say 20-30 there, and maybe another 30-50 towards the end of the 00z run after the ten days are finished. Totals could be 75 to 105 mm from those estimates. Month also not quite over at that point, so 80-120 looking good. Despite having 119 forecasts on file, nobody is entirely out of the hunt yet. Although I predicted 77.8 mm, based on what I am now seeing my pick is 111.1 of Stationary Front to 120 (Relativistic, IIRC) or in that range of forecasts. CET still looks like mid 5 to low 6 is a good bet.
  4. Except that the reporting of anomalies on a daily basis here is against running averages for this part of the month, so it's always equally likely to be above or below that statistic (not quite true in a warming climate it will be maybe 55% likely to be above and 45% below) ... what Rel was saying was, when you have only a few days, the average can be further up or down from normal than it's likely to be when you average 31 days at the end. The coldest running averages of December CET occur before the halfway point even though the second half is a bit colder on average. (check my post on 1st of Dec with normal and extreme values of CET, the lowest running CET was on the 8th of Dec 1879 at -3.65 C, the values then converge on the eventual coldest entire month, 1890 at -0.8). Also, the contribution of each day diminishes as you move forward. Today (the 11th) will be 1/11 of the average after today, tomorrow will be 1/12, the 31st will be 1/31, so if all three days are 3 deg below normal, today will have almost three times the impact on the running mean as the 31st would have.
  5. Recent METARs (21z) for Reyjkavik airport shows 330 (NNW) 31G46 knots and -2 C with snow and blowing snow. vis 900m _ at 22z this stayed similar with vis 600m. Recent METAR (21z) for Akuyeri airport shows 330 (NNW) 28 knots (nearby location is 41G58kn) in mod snow vis 1.4 km. _ at 22z this also stayed similar with vis 1.1 km. Assume it's even worse in parts of the northwest. Will post some updates if conditions are worse than these. Previous hours were similar at both locations but the nearby location to Akuyeri (which is Vadlaheidi) had gusts to 69 knots a few hours ago.
  6. EWP estimated to be around 18 mm now (was 14 mm after 7th, then 8th and first part of today look to have added 4 mm over the grid) ... 10-day GFS showing healthy amounts that average at least 50 mm over the grid, perhaps 55 mm for a total of 73 mm. The six days after that don't look overly wet on the charts but it's a pattern that looks a bit dubious so anyway, we will likely be somewhere between 70 and 90 mm by Christmas. That probably doesn't entirely eliminate any of the wetter forecasts from consideration given how stormy the Atlantic has been so far this month, it could lash the UK at any point towards the end. As for CET, I don't see a lot of movement from current levels, some colder looking maps at times but the flow remains broadly speaking westerly and that makes me suspicious that the thicknesses either won't verify or won't perform as they might in a different sourced air mass. And a few days in the mix looking quite mild. So it will probably stay in the general vicinity of the mid-5 to 6.0 range although I could imagine it getting as low as 5.0 or as high as 6.5 given the set-up.
  7. EWP remained zero for the first four days of December, looks to have added perhaps 4-5 mm on the 5th with 15-25 mm amounts in Wales and Cumbria, small returns elsewhere. The ten-day GFS has all areas at least 30-40 mm with a few (Wales and Cumbria again) well over 50, will estimate 40 mm for the grid there. The 11-16 day maps on GFS look fairly wet too and could add another 30 mm. That would bring the totals to 75 mm by the 22nd. As for CET projections, some days are quite cold but the pattern looks increasingly mild after mid-month with several days that look capable of touching 10 C. Overall I would estimate that we might be near 5 C at the end of the current run but would give 4 to 6 as the reasonable chances zone. A lot of forecasts are in play depending on what might lie over the event horizon. I had mentioned in my LRF the possibility of battleground scenarios and strong gradients with 25-26 Dec an indicated energy peak. If we stay in the pattern shown towards end of today's GFS run, look out for major storminess around the holidays. At least most people will already be where they want to be having travelled a few days earlier, so not the worst time for disruption. In this pattern, disruption would most likely be wind-related south, flood-related central and snow-related north.
  8. Final EWP contest results -- Contest Year 2018-19 This is the final "top 25" of our contest, with points totals and average errors with rankings shown for those also (the correlation while high is not absolute between points and errors, depends to some extent on which months you had larger errors). The ranking for average error was against those who entered at least 7/12 contests, some who entered fewer had smaller average errors for the months they played. Rank __ Forecaster _______ Points _____ Avg error __ Rank _ 01 ___ Born from the Void __ 91.4 _____ 15.8 ______ 1 _ 02 ___ Reef _____________ 83.8 _____ 18.8 ______ 2 _ 03 ___ Don ______________77.4 _____ 22.3 ______ 4 _ 04 ___ J10 ______________ 77.2 _____ 24.3 ______ 8 _ 05 ___ Dr(S)No ___________76.6 _____ 22.8 ______ 6 _ 06 ___ EdStone __________ 72.2 _____ 24.1 ______ 7 _ 07 ___ weather-history _____ 71.7 _____ 26.9 _____ 15 _ 08 ___ Blast from the Past __ 71.0 _____ 29.0 _____ 22 _ 09 ___ Godber.1 __________ 68.9 _____ 20.1 ______3** _ 10 ___ Stationary Front _____68.3 _____ 25.3 _____10 _ 11 ___ Feb1991Blizzard ____ 68.0 _____ 25.0 _____ 9 _ 12 ___ Stargazer __________67.9 _____ 27.2 _____17 _ 13 ___ Virtualsphere _______ 67.3 _____ 22.6 _____ 5** _ 14 ___ seaside60 __________66.5 _____ 26,3 _____13 _ 15 ___ daniel* ____________ 65.1 _____ 25.8 _____ 12 _ 16 ___ Jeff C _____________ 64.9 _____ 28.5 _____ 20 _ 17 ___ Midlands Ice Age ____ 64.7 _____ 27.6 _____ 19 _ 18 ___ Mulzy _____________ 64.5 _____ 28.6 _____ 21 _ 19 ___ Kirkcaldy Weather ___ 64.3 _____ 27.1 _____ 16 _ 20 ___ syed2878 __________ 62.4 _____ 31.7 _____ 27 _ 21 ___ jonboy _____________61,4 _____ 29.6 _____ 23** _ 22 ___ Steve B ____________59.9 _____ 31.5 _____ 26 _ 23 ___ CheesepuffScott _____58.9 _____ 31.7 _____ 28 _ 24 ___ brmbrmcar _________ 57.7 _____ 31.0 _____ 24 _ 25 ___ Norrance ___________57.4 _____ 31.3 _____ 25 _ 26 ___ DAVID SNOW _______56.1 _____ 25,5 _____11** -------------------------------------------------------------------------------------------------------------------------------- ** Of the top twenty-six, all but Godber1, virtualsphere, jonboy and DAVID SNOW played 12 months; those four played 11 times. ... ... this is why they generally ranked higher in the average error than points gained. (note Jonboy had a forecast in August that was too late to count, with that added his total would be close to 70 for about 8th place) (14th best average error went to Let it Snow who played only 8/12 so finishing in 32nd for total points). ---------------------------------------------------------------------------------------------------------------------------------- The four seasonal winners were: WINTER _ Blast from the Past SPRING _ DiagonalRedLine SUMMER _ J10 AUTUMN _ coldest winter ------------------------------------------------------------------------------------------------------------------------------------- The most points gained from 2017-18 contest year to 2018-19 for any forecaster who entered 11/12 or 12/12 in both years was 38.7 for EdStone who went from 44th place in our first contest year (11/12 forecasts) to 6th this year. Our contest winner Born from the Void picked up 30.2 points and moved up from 15th to 1st. (some part-timers from year one also moved up a lot of points and ranks but I only looked at regular entrants in both years) The excel file now contains a comparison of year one and year two total points, in columns DO, DP and DQ. Most people who were regulars both times seem to have finished remarkably close to where they finished in year one. The most points earned in total over two years goes to Don who was second and then third in the two contest years. Close behind was J10 who has finished third and fourth in the first two contest years. ============================================================== One more stat, from the rankings table ... the 12-month forecaster who had the best "worst ranked forecast" was once again Born from the Void whose ranks never dropped lower than 21st place. His average rank was 12th, the next highest among regular entrants was 15th for Reef. (Coldest winter who only entered the last four contests finished with three high-ranked efforts, 5th, 6th and 7th in the autumn). Congrats to all of our forecasters and good luck going forward. FULL SCORING DETAILS IN THE ATTACHED FILE EWP20182019NOV.xlsx
  9. The November EWP has been posted now at 138.8 mm. That won't change much in the excel scoring file, but I need to check some numbers in the "ultimate scoring" section as they also updated (just September, October is unchanged so far) as is the custom at Hadley EWP. (added later _ Nov scores changed by 0.2 up for Lettucing Gutted and I Remember Atlantic 252, down for Mulzy and timmytour, as the slightly revised EWP changed their ranks and scores, after BFTP only five others had forecast above the outcome and the other three did not change ranks or scoring as a result). Can also confirm that coldest winter has won the autumn segment of the contest with Don, SteveB, DR(S)NO and virtualsphere in the next four positions. Eventual contest 1-2 finishers Born From the Void and Reef were sixth and seventh in autumn. Stationary Front, jonboy and syed2878 round out the top ten for autumn. So I will post some summaries of annual scoring. This will likely confirm also the table I posted a few days ago of "best combined forecast" results since all ranks are going to remain the same after today's minor adjustments. The closest forecast remains the 138.0 mm of Feb1991Blizzard now followed a bit more closely by second place Blast From The Past (140 mm). Back in a while with the final version of scoring for the contest year.
  10. About that second half of the months (beyond the last gasp of the guidance for even the latest forecasters to make a punt) ... In the contest year 2018-19, the mean daily CET for the various days of the month, compared to 1772-2019 average values (from my research file). I have averaged each five days to simplify the presentation. Interval ___ Mean daily CET ___ Average of all data 01-05 _____ 9.8 _____________ 9.4 06-10 _____10.0 _____________9.4 11-15 _____ 9.6 ______________9.4 16-20 _____10.4 _____________ 9.3 21-25 _____11.9 _____________ 9.4 22-26 _____12.1 _____________ 9.4 (max dep) 26-31 _____ 10.7 _____________9.5 ===================================== As you can see, the departures from normal (even this long-term normal cooler than 1981-2010 by 0.8 C) were small until after the 15th. They rose to a peak of 2.7 above that normal (would be close to 2 C above recent normal) by 22nd to 26th of the twelve months. That's why I added the one extra line to the table, to show the maximum departure. So as luck would have it, most months (January probably an exception) warmed significantly relative to normal values after the guidance period. The graphs converge towards the end because February drops out and then on the last day, April, June, September and November drop out. That suggests that the effect was a bit more pronounced in February and the 30-day months. The coldest relative to long-term normal was the 4th which ran 9.0 vs the rather static long-term 9.4 (when you average so many months the numbers stay fairly flat, always between 9.32 and 9.44, except that they rise slightly on 29th and 30th as February is the only missing month from those normals, which are 9.49 and 9.52. You'll recall that the summer months all had heat wave conditions after the 25th for at least a day or two. People who noticed this second half warmer trend probably did a bit better than most in the CET contest. The attached graph shows the daily averages, which peak at 12.7 on the 25th and 12.2 on the 26th (22nd-26th is the warmest five-day average). GRAPH: MEAN DAILY C.E.T. Contest Year 2018-19 by DAYS OF MONTH (blue) vs 1772-2019 AVERAGES (for existing dates)* * removes data that would be 29-31 Feb, 31 Apr, 31 Jun, 31 Sep 31 Nov. The removal of that data explains the bump upwards of the average (red) on 29th-30th as only February is excluded from that data. __ Note: CET values are 10x actual, e.g. 120 = 12.0 C
  11. In the coming year, I am hoping to work on a project to create a third approach to climate change which is neither the orthodox (human dominated) scenario nor any kind of denial. Whether it would still qualify as skeptical depends on how you interpret the approach. It is not meant to downplay the critical aspects of the situation. In fact it might well underscore that the situation is not only critical, but irreversible (by us) and therefore in need of a different strategy. First of all, I don't deny the physics of climate change as being in part a response to rising greenhouse gas levels, and human activity is obviously the main source of those. It has to be remembered though, warming of any cause will increase the greenhouse gas levels in the atmosphere to some extent as formerly frozen ground thaws and vegetation spreads. So there's one area of imprecise understanding, how much of the rise in greenhouse gases is a cause, how much is an effect, and could the climate warm naturally without the external addition of greenhouse gases. Past history suggests that it could. I have always been skeptical of some aspects of the orthodox climate science that is driving public policy debates. I have said on past occasions that I thought the warming observed from about 1990 to 2015 was at least half caused by natural variability of the same kind that produced natural warmings on several other recent occasions, for example, from the Maunder to the mild 1710-39 period, from the Dalton to the somewhat milder mid-19th century, and from the later solar downturn cooling (aided by Krakatoa dust) around the 1880s, to rather sharp warming signals in the late 1890s. The period from 1921 to 1960 also appears to have been impacted by a considerable natural warming signal. Because there was a cooling trend in many climate zones around the 1960s to early 1980s, a resumed natural warming cycle after 1987 looked to researchers like it was very likely to be of human origin because around then, the increases in greenhouse gases well underway by 1960 had accumulated and (as they theorized) overwhelmed the natural balance of our atmosphere to the point where a possibly catastrophic warming was then underway. This view has not changed much except for its broadening out to include just about every variation in weather and climate as possible "symptoms" of human-driven climate change. Some in the atmospheric sciences (and by no means just a few eccentric outsiders which is how I would surely come across in this day and age) are rather quietly grumbling away about an overkill in that broadening, to the point where every weather event that probably would have happened in any case is being blamed on you and me driving our cars and heating our homes. It makes no sense to me to blame floods and windstorms and droughts on human climate change when those elements were present in climate zones long before the industrial age or even human settlement in some cases. What was causing them back then? Buffalo farts? In recent years, climate change has even been blamed for colder episodes like the polar vortex phenomenon. Any casual reading of 19th century climate records will show that in a natural cool phase, the arctic vortex often forms and moves south. So what are the orthodox scientists really saying here? That climate change may produce the outcome we were looking for anyway? It makes no sense to people, which is why the politics of climate change has proven to be a tougher sell than its proponents expected. The average person does not want to be shelled for extra taxes for no plausible reason. Somebody has to identify the 800-lb gorilla in the room -- carbon taxes do not change the weather. They don't do much to change behaviour either. They just impoverish people, forcing them to pay more for essentials perhaps at the cost of medical care or food. When the zealots cry out, "what will we tell our grandchildren?" perhaps tell them we were trying to give them enough money to survive? The orthodox theory is rather like the old floating iceberg theory of erratic boulders, it sounds plausible but a better theory is creeping up on it waiting to kill it off when enough people realize it is wrong.That better theory is natural variability in control of climate. We are just minor players who (for some) think we are in charge. We are not creating weather patterns. We are not really even modifying them. The natural changes are modifying them. It's getting warmer because storm tracks are shifting north (at about the same rate as the magnetic pole). When storms moved further north in the cold climate of the 19th century, temperatures approached modern levels. A few records were set then too. The longer residence time in cold air kept the ground chilled which had to be overcome by those air masses moving north. This is why old records often fall by 1 or 2 degrees now. They don't stand much of a chance having been recorded in an era when the ground temperatures were several degrees lower than today. Orthodox climate change theory is wrong. It is obvious that natural variability has not gone away. Most of the natural variability since 1700 has been in the warming direction. We are in an inter-glacial period quite a bit removed from any widespread glaciation outside of the zones we see today. Sure, small ice caps on Baffin Island have shrunk down, mountain glaciers are retreating, and there has been some marginal melt from Greenland. But that process could very well be expected to be ongoing towards some hypothetical mid-inter-glacial minimum that our science frankly cannot reliably estimate. What if any logic is there in saying that the climate of 1921-50 was somehow an ideal climate that we should fight to preserve? Why not the Maunder climate or the warmer Neolithic warrm spell (warmer than nowadays from the biological records left behind)? More to the point, how are we supposed to prevent the climate from warming naturally? I suppose there would be possible ways to do this, but carbon taxes would not be very useful in that regard. Our addition to the natural warming signal is slight, in my opinion, perhaps one quarter or less. And if we could tone down our contribution, that other 75 to 90 per cent would roll on without interruption. Now there is one silver lining in this approach, which is that a natural cooling trend could set in. This has happened perhaps three times since 1739 (on a large scale) and perhaps three or four times between the postulated quasi-20th century MWP and the onset of the Maunder. The time scale appears to be roughly once a century. We see a longish downturn in solar activity at least rivalling the Dalton minimum. Perhaps we will get lucky and see a long cooling trend. Our warming signal will mask that and cause a flatter trend to emerge. Some say that has already begun to happen. One of the strongest signs of natural variation being in control is how different zones warm and cool at similar times. An external driver must be suspected when different climate zones have their peak temperatures in similar time periods. All of them are getting a bit of extra heat from somewhere. But why would that be human caused? Would an ongoing, never-varying human output of heat not have equal influence all the time? There are problems with the orthodox theory that should cause it to fall soon. And the paranoia of its proponents is a tip-off that it is a weak theory and they know it is weak. Join me. I'm going to try to put an end to this idiocy once and for all. Our futures may depend on it. An ongoing natural warming would imply irreversible sea level changes and other outcomes that will need to be mitigated because they cannot be prevented. The political implications of this are self-evident. Our progressive political parties are on the wrong path trying to prevent something that nobody can prevent if it's going to happen. And those who take a denialist, skeptical approach ("nothing bad will happen") could be equally wrong. I wouldn't want to bet against natural variation causing further warming of climates all the way through to some sort of tipping point when sea levels will start to rise appreciably. Are we ready for this? No, our response is to blame humanity for something it is not causing, and to place trust in a hail-Mary political strategy that can't possibly work (even within its own paradigms, it can't work, not without causing such severe economic distress that people will naturally rebel against it). These are big stakes and it's time for real science to happen. Climate science is a social science that got past real scientists somehow and fooled them into thinking it was legitimate. It is bogus. I think two-thirds of the people on weather forums hold this same opinion. The people who have seen the most actual past weather data are naturally skeptical. But I don't think a lot of them will instantly agree with my third option. It has become comfortable to dissent by holding to steady-state minor fluctuation paradigms. There again, anyone familiar with climate history will know that our climate doesn't do gentle changes, it lurches rather violently from one state to another (the Dryas periods for example, at the end of the last glacial era). I hope the moderators don't remove this post. These ideas need to be discussed and debated. We have gone a bit too long now without a plausible alternative to the orthodox theory.
  12. The EWP tracker blew off the rain in Cornwall on 30th and remained stuck on 137 mm. I won't bother to update my scoring file (set to 138) until the final number is posted, as it won't be useful work to do twice. The standings are not likely to change on any value between 136 and 140 mm basically. Will publish the full details on contest year scoring on the 5th when that final number is made known. Meanwhile, on to the business of tracking our three robotic contestants (they aren't half bad actually) ... REPORT ON CONSENSUS and NORMALS SCORING for NOVEMBER 2019 and the contest year wrap-up Dec 2018 _____________________________ Jan 2019 _______________________ Feb 2019 __________ ____FORECAST _ error __ rank __ points ____ FCST __ error__ rank _ points _____ FCST __ error __ rank __ points Consensus_ 5.0 __ -1.9 _ 32 to 32 _ 52.3 _____3.5 ___ -0.5 __16 to 18 _ 75.7 to 78.6 _ 4.0 __ -2.7 _ 33 to 37 _ 48.3 to 54.1 1989-2018*_4.9 __ -2.0 _ 33 to 33 _ 50.8 _____4.7 ___ +0.7 __21 to 23 _ 68.6 to 71.4 _ 4.9 __ -1.8 _ 14 to 14 _ 81.3 1981-2010__4.6 _ -2.3 _42 to 44 _ 33.8 to 36.9 _4.4 __ +0.4 __14 to 15 _ 80.0 to 81.4 _ 4.4 __ -2.3 _ 17 to 19 _ 74.2 to 77.0 March 2019 _____________________________ April 2019 ______________________ May 2019 ____FORECAST _ error __ rank __ points _____FCST_error _ rank __ points _______ FCST _ error _ rank __ points consensus _ 6.9 _ --0.9 _ 29 to 33 _ 48.4 to 54.8 _8.4 _ --0.7 _ 24 to 28 _ 55.9 to 62.3 _ 12.3 _ +1.2 _ 28 to 31 _ 50.0-55.0 1989-2018 _ 6.8 __ --1.0 _ 34 to 35 _ 46.2 to 47.8 _8.8 _ --0.3 _ 9 to 13 _ 80.3 to 86.9 __ 12.0 _ +0.9 _ 17 to 20 _ 68.3 to 73.3 1981-2010 _ 6.6 __ --1.2 _ 39 to 39 _ 39.7 ______8.5 _ --0.6 _ 20 to 23 _ 63.9 to 68.6 __ 11.7 _ +0.6 _ 11 to 11 _ 83.3 June 2019 __________________________________ July 2019 ______________________ August 2019 ____FORECAST __ error __ rank __ points _______ FCST _ error _ rank _ points ________ FCST _ error _ rank _ points consensus _ 15.0 _ +0.8 _ 27 to 31 _ 48.1 to 55.0 __ 17.0 _ -0.5 _ 19-24 __ 62.9 to 71.0 ____17.0 _ --0.1 __ 3 to 13 __ 81.0 to 96.8 1989-2018 _ 14.6 _ +0.4 _ 11 to 15 _ 75.8 to 82.7 __ 16.9 _ -0.6 _ 25-30 __ 53.2 to 61.3 ____16.5 _ --0.6 __38 to 43 __ 33.6 to 41.5 1981-2010 _ 14.5 _ +0.3 __ 6 to 10 _ 84.4 to 91.3 __ 16.7 _ -0.8 _ 34-39 __ 38.7 to 46.8 ____16.4 _ --0.7 __44 to 44 __ 32.0 to 32.0 ____________________________________________________________________________________________ September 2019 _____________________________ October 2019 _____________________ November 2019 ____FORECAST __ error __ rank __ points _______ FCST _ error _ rank ___ points _______ FCST _ error _ rank ___ points consensus ________ 0.0 _ 01 to 01_ 100.0 ________ 10.9 _ +0.9 _ 29 to 34 _ 45.0 to 53.3 __ 6.9 ___ +0.7 __ 27 to 30 _ 55.1 to 59.8 1989-2018 ________-0.1 _ 02 to 11 _ 84.6 to 98.5 ___11.0 _ +1.0 _ 35 to 36 _ 41.7 to 43.3 __ 7.4 ___ +1.2 __ 51 to 52 _ 21.1 to 22.7 1981-2010 ________-0.3 _ 20 to 25 _ 63.0 to 70.7 ___10.7 _ +0.7 _ 24 to 26 _ 58.3 to 61.7 __ 7.1 ___ +0.9 __ 38 to 41 _ 38.1 to 42.8 =========================================================================================== __________ Average (12 months) __________ ___________ abs err __ rank __ points consensus ___0.93 __ 23 to 26 _ 59.8 to 66.0 1989-2018*___ 0.88 __ 23 to 27 _ 60.3 to 62.5 1981-2010 ___ 0.93 __ 27 to 29 _ 56.8 to 60.7 *note played Dec 2018 as 1988-2017. ________________________________________________________________________________________ Analysis of these scores ... Our consensus caught up to the two normals in November and all three scores are basically in the same ball-park. They are roughly equivalent to the range of 11th to 15th in the CET scoring table. The range of ranks and points reflect the fact that more than one forecaster usually has the same error values as these three robotic forecasters. But in terms of the average error of their forecasts, 0.93 was the average of our 22nd place forecaster (average error is not quite in the same order as total points though, the spread is rather wobbly down the table and 0.88 is the average of our 12th place forecaster. In fact one or two of the top ten had higher average errors than these three robots, while our 31st place finisher was barely above the highest of these errors. So it depends in part on which months you beat these robots and which months you fall short. The best robotic result would be for a forecast that was 0.3 above the recent 30-year normal. That would have averaged an error of 0.87 deg. If you went to 0.6 above the recent normal you would start increasing your average error (there it would be back to 0.92). Anywhere from 0.2 to 0.4 above the recent normal would be about the same as the ideal 0.3. Our consensus did better than either robotic normal in several months but except for August and November, the differentials were small to insignificant. We did statistically worse as a group in February, May and June. The first of those months proved milder than we expected and May/June proved to be cooler than recent normals while we mostly went warmer than them. In January we went colder than the recent normals by about twice the amount that was justified by the outcome, so the consensus and normal scores there were similar although opposite in sign. The bias can be assessed by the actual average as opposed to the absolute average calculated above. The consensus bias was -0.31, meaning that on average, our forecasts were 0.3 C lower than outcomes. The bias for 1981-2010 was -0.55 meaning that this contest year has averaged 0.55 above the 1981-2010 normals. The bias for 1989-2018 was -0.18 deg indicating that the contest year was only 0.18 C warmer than the 1989-2018 averages. This will obviously demonstrate that forecasts about 0.2 to 0.3 above those "normals" will eliminate bias altogether. I can say on a preliminary basis that consensus did better than the recent normals in the EWP. We seem to have a bit more skill there, and the current ranking of consensus is 6th on points, while the two normals are tied for 20th, still better than at least half the regular field, but rather close to random in terms of skill shown. Since recent normals are not all trending in the same direction recently, there is no "sweet spot" relative to recent normals and you're better off to hug consensus which does show considerable skill, rarely finishing much lower than halfway down the pack. Our overall bias in EWP forecasts seems to be to go a bit lower than the actual amounts. But a few players (as with CET) have large biases on the upside, whether they truly believe in them or not, and this may disguise a larger tendency to go too low since we have a handful going way too high on a regular basis. It turns out I am a bit of an offender in that regard (for EWP, not CET), although I am actually trying to hit the numbers.
  13. I have updated the first post of the thread with data from Nov 2019 (6.2). This rather cold value now means that Nov 2020 will have to finish 6.6 to 9.5 to stay on 7.4, as we are now at about 7.36. Anything plausible colder than 6.6 will make 7.3 the new value. A really warm November could take us to 7.5. SteveB, here's the full breakdown of 30-year averages since 1901-30 for your interest. If you want to see earlier values, look in my historical CET thread in the historical weather section. Some of my values may round 0.1 different from Met Office values because I think they derive them from an average of 30-year mean max and mean min. Also you get into third decimal issues about rounding from .x5 situations. But this gives the general picture of what's been happening in the recent past. Interval __________JAN __ FEB __ MAR _ APR _ MAY _ JUN __ JUL __ AUG __ SEP __ OCT _ NOV _ DEC __ YEAR 1901-1930________4.2 ___4.2 ___5.5 ___8.0 __ 11.3 __ 13.8 __ 15.8 __ 15.2 __ 13.1 ___9.8 ___5.8 ___4.5 ___ 9.2 1911-1940________4.2 ___4.3 ___5.7 ___8.2 __ 11.5 __ 14.2 __ 15.9 __ 15.6 __ 13.3 ___9.6 ___6.2 ___4.6 ___ 9.4 1921-1950________3.9 ___4.2 ___5.9 ___8.0 __ 11.3 __ 14.4 __ 16.3 __ 15.9 __ 13.6 __10.0 ___6.4 ___4.5 ___ 9.5 1931-1960________3.5 ___3.9 ___5.9 ___8.1 __ 11.5 __ 14.6 __ 16.2 __ 16.0 __ 13.7 __10.1 ___6.8 ___4.7 ___ 9.6 1941-1970________3.3 ___3.6 ___5.7 ___8.1 __ 11.4 __ 14.5 __ 16.0 __ 15.7 __ 13.7 __10.6 ___6.6 ___4.4 ___ 9.5 1951-1980________3.6 ___3.7 ___5.6 ___8.0 __ 11.3 __ 14.3 __ 15.9 __ 15.6 __ 13.6 __10.5 ___6.6 ___4.7 ___ 9.4 1961-1990________3.9 ___3.7 ___5.7 ___7.9 __ 11.1 __ 14.1 __ 16.1 __ 15.8 __ 13.6 __10.6 ___6.5 ___4.7 ___ 9.5 1971-2000________4.2 ___4.2 ___6.3 ___8.1 __ 11.3 __ 14.1 __ 16.5 __ 16.2 __ 13.7 __10.4 ___6.9 ___5.1 ___ 9.7 1981-2010________4.4 ___4.4 ___6.6 ___8.5 __ 11.7 __ 14.5 __ 16.7 __ 16.4 __ 14.0 __10.7 ___7.1 ___4.6 __ 10.0 ============================================================================================= Trends in the above data reveal that Jan and Feb were milder in the first half of the 20th century than around 1941-80, and we were only edging past those values by 1981-2010. Spring, summer and autumn showed somewhat the opposite trend until the recent warming began.
  14. 1981-2010 CET normals and daily records (1772 to 2018) ... since we have quite a few new entrants to our CET forecast contest, this table will be the basis for some daily reports you'll be seeing from Summer Sun, who will post the running CET values (first to previous day's date as reported here http://hadobs.metoffice.com/hadcet/cet_info_mean.html and while they still use 1961-90 data for their anomalies, we also have 1981-2010 available in this table. Column 2 shows you the daily averages while column 3 shows the mean running CET to that date. Columns 4 and 5 show the all-time record values for daily mean CET (the average of max and min, obviously the max and min go a bit outside these values). And the final two columns show the highest and lowest values that any month since 1772 has managed to set for running CET. Meanwhile, I will update the EWP tracker information based on latest available data on Hadley EWP site and 24h rainfall maps on meteociel to bring them right up to the latest possible time (as they are reported about 1-2 days after the fact on the Hadley website). Here's a link to that: http://hadobs.metoffice.com/hadukp/charts/charts.html One other thing, in case you hadn't worked this out, J10 is your host for CET forecasts and scores them at end of month. I am your host for EWP. So direct any questions you might have about scoring or reporting issues to the person in charge in each case. Thanks. ____________________ Daily record values _______ Extremes of running CET DATE ___ CET __ CET cum _____ MAX _______ MIN _______ MAX _______ MIN 01 _____ 5.8 ___ 5.8 ______ 12.1 (1775) ___ --3.0 (1947) ____ 12.1 (1775) __ --3.0 (1947) 02 _____ 5.8 ___ 5.8 ______ 12.2 (1985) ___ --4.7 (1796) ____ 11.1 (1775) __ --3.4 (1796) 03 _____ 6.0 ___ 5.9 ______ 12.2 (1985) ___ --4.0 (1879) ____ 10.9 (1985) __ --3.1 (1879) 04 _____ 6.1 ___ 5.9 ______ 12.1 (1888) ___ --3.8 (1879,1925) _10.9 (1985) __ --3.2 (1879) 05 _____ 6.1 ___ 6.0 ______ 12.6 (1898) ___ --3.7 (1844) ____ 10.6 (1985) __ --3.2 (1879) 06 _____ 5.0 ___ 5.8 ______ 12.3 (1898) ___ --5.5 (1844) ____ 10.7 (1898) __ --3.3 (1879) 07 _____ 4.8 ___ 5.7 ______ 12.3 (1856) ___ --5.7 (1879) ____ 10.3 (2015) __ --3.64(1879) 08 _____ 4.4 ___ 5.5 ______ 12.1 (1856) ___ --5.4 (1807) ____ 10.4 (2015) __ --3.65(1879) 09 _____ 4.5 ___ 5.4 ______ 11.8 (1934) ___ --4.0 (1803) ____ 10.1 (2015) __ --3.2 (1879) 10 _____ 4.3 ___ 5.3 ______ 10.9 (1997) ___ --4.5 (1784) ____ 9.9 (1979, 2015)_--2.8 (1879) 11 _____ 4.2 ___ 5.2 ______ 12.6 (1994) ___ --6.7 (1784) ____ 9.7 (1898,2000)_--2.8 (1879) 12 _____ 4.2 ___ 5.1 ______ 12.9 (1994) ___ --8.5 (1981) _____ 9.9 (2000) __ --2.7 (1879) 13 _____ 4.6 ___ 5.1 ______ 11.6 (1918) ___ --6.5 (1981) _____ 9.7 (2000) __ --2.4 (1879) 14 _____ 4.4 ___ 5.0 ______ 11.8 (1912) ___ --5.3 (1878) _____ 9.4 (2000) __ --2.1 (1879) 15 _____ 4.6 ___ 5.0 ______ 11.2 (1985) ___ --5.4 (1788) _____ 9.2 (1898) __ --1.6 (1879) 16 _____ 4.6 ___ 5.0 ______ 11.4 (2015) ___ --6.6 (1859) _____ 9.1 (2015) __ --1.3 (1879) 17 _____ 4.4 ___ 4.9 ______ 12.5 (2015) ___ --7.7 (1859) _____ 9.3 (2015) __ --1.3 (1879) 18 _____ 4.4 ___ 4.9 ______ 12.5 (2015) ___ --7.1 (1859) _____ 9.5 (2015) __ --1.2 (1879) 19 _____ 4.3 ___ 4.9 ______ 13.1 (2015) ___ --6.8 (2010) _____ 9.7 (2015) __ --1.1 (1879) 20 _____ 3.7 ___ 4.8 ______ 11.1 (1833) ___ --7.0 (2010) _____ 9.7 (2015) __--0.9 (1878,79) 21 _____ 4.2 ___ 4.8 ______ 11.4 (1971) ___ --5.9 (2010) _____ 9.7 (2015) __ --1.0 (1878, 2010) 22 _____ 4.5 ___ 4.8 ______ 11.6 (1991) ___ --6.8 (1890) _____ 9.7 (2015) __ --1.0 (1878, 2010) 23 _____ 4.7 ___ 4.8 ______ 11.8 (1991) ___ --6.4 (1870) _____ 9.7 (2015) __ --1.1 (1878, 2010) 24 _____ 4.7 ___ 4.8 ______ 10.7 (1843) ___ --6.9 (1870) _____ 9.7 (2015) __ --1.3 (1878) 25 _____ 4.5 ___ 4.8 ___ 10.4 (1824,1983) _--10.8 (1796) _____ 9.6 (2015) __ --1.5 (1878) 26 _____ 4.3 ___ 4.7 ______ 11.0 (2011) ___ --5.4 (1798) _____ 9.6 (2015) __ --1.4 (1878, 2010) 27 _____ 3.9 ___ 4.7 ______ 10.9 (2015) ___ --8.4 (1798) _____ 9.7 (2015) __ --1.4 (2010) 28 _____ 4.1 ___ 4.7 ______ 11.5 (1987) ___ --8.2 (1798) _____ 9.7 (2015) __ --1.3 (2010) 29 _____ 3.9 ___ 4.7 ______ 12.0 (1987) ___ --4.9 (1874) _____ 9.7 (2015) __ --1.1 (2010) 30 _____ 4.2 ___ 4.6 ______ 11.0 (1834) ___ --6.1 (1870) _____ 9.8 (2015) __ --0.9 (2010) 31 _____ 4.2 ___ 4.6 ______ 11.1 (1859) ___ --7.1 (1783) _____ 9.7 (2015) __ --0.8 (1890) ________________________________________________________________ So from this we see that 2015 had one crack at being warmest Dec (7th to 10th), fell out of first place for a while, then resumed the lead from 16th to end of month. The closest to 2015 that any year came after the 20th was 8.7 (1898 - 21st). As for 2010, it moved into first by the 21st, had a running battle with 1878, shook that off but got nipped at the very end by 1890 which never held the lead at any other stage. Other noteworthy late negative running CET values would include 1788 which was -0.4 on 23rd and -0.3 on 30th-31st, 1796 which was running at -1.0 from 26th to 28th, 1844 which was -0.2 on 25th, 1879 held on at -0.4 to 25th, and more recently 1981 was running at -0.2 from 24th to 27th (finished 0.3).
  15. If you had some trouble opening the CET scoring file, I have posted a simplified title link to it at the bottom of this post. This is the combined rankings of CET and EWP for all who posted 11 or 12 forecasts in either venue. The CET rankings are straight from J10's scoring file, the EWP rankings are very slightly subject to alteration when November is confirmed. ... Only those who played both contests are shown in this table, but CET-only (or CET-mainly) forecasters are identified over to the right of the table in a similar position. Some of those had a small number of EWP forecasts and are ranked in that contest but if they played fewer than half their CET entries I don't show those results here. They are tagged with an asterisk though. In the case of nn2013 who played eleven CET and seven EWP, if they had entered the other four and scored just 4/10, EWP rank would have been 30th moving them up a couple of spots, but if I just listed them without EWP rank then they would show up in roughly the same position in this table anyway. The same applies to Leo97t (ten CET, five EWP). In order to qualify for a CET ranking it is necessary to enter at least 10 of the 12 contests. The EWP rankings are not separated by numbers of entries although it is tough to finish much higher than 40th with fewer than ten entries. Ties are broken by highest rank in either contest (e.g., 3rd and 4th below). _ ranks now confirmed for EWP as of 5th December __ BEST COMBINED CET and EWP FORECASTS 2018-19 COMPETITION YEAR __ Rank __ FORECASTER ___ CET rank ___ EWP rank ___ Avg ___ CET-only or CET-mainly in similar ranks _01 ___ Don _____________ 4 __________ 3 __________3.5 ___ Quicksilver (1) _02 ___ EdStone _________ 5 ___________ 6 _________ 5.5 _03 ___ Born from the Void _ 11 __________ 1 __________6.0 _04 ___ Stationary Front ____2 __________10 __________6.0 _05 ___ weather-history ____ 6 __________ 7 __________ 6.5 ___ Summer Blizzard*(8) _06 ___ DR(S)NO ________ 16 __________ 5 _________10.5 ___ _07 ___ Feb1991Blizzard ___10 _________ 11 _________10.5 ___ Damianslaw (9) _08 ___ Reef ____________ 22 __________ 2 _________12.0 ___ _09 ___ Stargazer ________ 12 _________ 12 _________12.0 ___ _10 ___ Mulzy ____________ 7 _________ 18 _________12.5 ___ _11 ___ seaside60_______ T18 _________ 14 _________ 16.0 ___ _12 ___ The PIT __________ 3 _________ T30 ________ 16.5 ___ _13 ___ Midlands Ice Age __ 17 _________ 17_________17.0 ___ _14 ___ J10 _____________36 __________ 4 _________ 20.0 ___ _15 ___ Norrance ________ 15 _________ 25 _________ 20.0 ___ _16 ___ DiagonalRedLine __ 13 _________T30 ________ 21.5 ___ _17 ___ Blast from the Past_ 38 __________ 8_________ 23.0 ___ _18 ___ Kirkcaldy Weather _T27__________19 ________ 23.0 ___ _19 ___ CheesepuffScott ___23 _________ 23_________ 23.0 ___ _20 ___ DAVID SNOW _____21 _________ 26_________ 23.5 ___ _21 ___ Daniel* __________ 33 _________ 15 _________24.0 ___ _22 ___ JeffC ____________32 _________ 16 _________24.0 ___ _23 ___ Godber.1 ________T40 __________9 _________ 24.5 ___ _24 ___ Roger J Smith _____20 _________ 29 _________ 24.5 ___ Jonathan F. (24) _25 ___ virtualsphere ______37 _________ 13 _________ 25.0 ___ Duncan McAlister (25) _26 ___ Timmytour ________14 _________ 36 _________ 25.0 ___ Summer Sun (26) _27 ___ jonboy ___________34 _________ 21 _________ 27.5 ___ Mark Bayley (T27) _28 ___ nn2013 _________ T18 _________ 38 _________ 28.0 __ (note -- entered 11 CET, 7 EWP) _29 ___ Weather26 _______ 30 _________T33 _________ 31.5 ___ Man with Beard (29) _30 ___ brmbrncar _______ T40 _________ 24 _________ 32.0 ___ ^^ sundog (see note below) _31 ___ SteveB ___________46 _________ 22 _________ 34.0 ___ dancerwithwings (31) _32 ___ davehsug _________42 _________ 27 _________ 34.5 ___ _33 ___ syed2878 _________50 _________ 20 _________ 35.0 ___ _34 ___ bobd29 ___________44 _________ 28 _________ 36.0 ___ Walsall Wood Snow (35) _35 ___ I Remember Atl252 _ 43 _________ 35 _________ 39.0 ___ _36 ___ Polar Gael ________ 47 _________T33 _________40.0 ___ _37 ___ Leo97t ___________ 39 _________ 48 _________ 43.5 __ (note -- entered 10 CET, 5 EWP) _38 ___ stewfox ___________51 _________ 37 _________ 44.0 ___ snowray (45) _39 ___ Lettucing Gutted ____52 _________ 60 _________ 56.0 ___ prolongedSnowLover (48) _40a __ Let It Snow! ________40** _______ 32 _________ 36.0** __ _40b___ Emmett Garland ____37** _______ 42 _________ 39,5** __ Kentish Man (49) --------------------------------------------------------------------------------------------------------------- ** Emmett Garland played only eight of each (CET, EWP) and on points would have been 37th in CET if included in the main scoring table. He was 42nd in EWP. On a pro-rated basis these results are about the equivalent of 20th place overall in both contests. ** Let it Snow! played nine CET and eight EWP. On points in CET the finish would have been 40th and in the EWP it was 32nd. Pro-rated basis, about 30th overall. ^^ Sundog who only played CET had enough points in nine outings to finish 26th in the main scoring table. For any other part time players, consult the two scoring files and you can estimate your performance level by pro-rating the total scores to fit (e.g., you played 6/12, double the score you had and see where that lands in the tables). Coldest winter has been scoring at about the same pace as EWP leader BFTV since starting up with the last four entries. They had those same months in CET plus three earlier ones without equivalent EWP. The CET scoring would equate to about middle of the table. If you do this pro-rating, in the CET table make sure you are in the section for total points scoring mid-way through and not the combined scoring as that will give you a false positive. ========================= Here's that link to the CET file in case you had any difficulty opening it. The EWP file will be adjusted tomorrow and reposted in a report on final standings there. It can be seen a few posts back in its current status. Nov19CET.xlsx
×
×
  • Create New...