Relativistic

So many people are looking for Spring, meanwhile I'm sitting here hoping that we get as many snow opportunities as we can over the next three months. I must stress however that I won't be actively searching for it like those in the model thread. This is bound to cause some controversy but I think that place is a joke at times. The MJO has shown itself to be waste of time. Charts in FI are useless. People getting hung up over each individual run. I don't think I'll be going back there very often.

Defining the second half of January as 16th-31st, and the second half of February as 15th-28th/29th, the 1987-2016 averages for the second halves of these months are: Jan: 4.5C Feb: 5.1C Some really mild second halves of February are dragging the average up here. The years 1990, 1997, 1998, 2007, 2009, 2012, and 2014 all recorded 7C+ for this period. Only the second halves of January 2002 and January 2008 managed to achieve the same feat during the 1987-2016 period.

I think this is true, but I would only put some of the blame on climate change. January has had mild phases in the past. I've plotted the 30-year January CET means through time. You can see that there have certainly been ups and downs; I think what is most notable however is how mild January was around 1900 (nearly as mild as at present), and how rapidly it dropped off soon after. Based on this graph, I think it's entirely possible that we are on a natural upward trend on top of a gradual warming trend caused by climate change. I wouldn't be surprised to see the average fall off again (as has happened before), but I suspect that should this occur the magnitude of the fall may be somewhat dampened by climate change.

Coldest February ever, -2C.

From that list, 1760 was the only year where both the June and September CETs exceeded 15C, as also happened in 2016. Starting from May, 1760 followed a very similar path to 2016 in CET terms. Unfortunately for cold seekers, this doesn't bode well for the Winter CET: Jan 1761: 5.4C; Feb 1761: 5.8C; March 1761: 6.8C.

Worth mentioning that the 10th-25th January 2013 recorded a CET of -0.1C, but two mild spells bookended the month to give a mean closer to average. Was a very decent spell of weather, I wouldn't turn down another January 2013 if it was offered to me. Edit: Surprisingly, in the last 50 years only five Januaries have managed a 16-day spell as cold or colder than this (1979, 1982, 1985, 1987, 2010).

January: mild, dull, some cold weather thrown in but nothing notable. February: mild again, although turning colder into March, some frost. March: a cold start with frost and some falling snow, otherwise average. April: a chilly month, most notably at the end with falling snow on a number of days (26th was the best), the 30th was a very eventful day with a crazy hailstorm. May: muggy and warm most of the time. June: muggy, warm, and exceptionally cloudy. July: very dry with temperatures regularly in the mid-20s, and that exceptional plume that brought crazy nighttime minima, lack of storms. August: another dry and warm month, with hot days interspersed throughout. September: by this point the ground had huge cracks (easily big enough to get your fist down there) in it because of the warmth and dryness, exceptional heat mid-month. October: dry (bar the 1st) and chilly for large parts. November: quite frosty on the whole and not much mild weather, again dry. December: started very frosty but quickly turned into mild crap, the end of the month saw more frost. Overall, a mild start, cooling off in Spring with some very late dates for falling snow, then leading into a very long and warm Summer (May to September) that featured 30C days in all three of July, August, and September. The remainder of Autumn was overall chilly, with some decent frosty days. December not so great.

The July-August combination in 2013 was 15th warmest on record. A 17C August is notable, as happened this year, at over a standard deviation away from the mean.

Here. It sounds to me that that was strongly implied here. Of course, there could have been a terrible misunderstanding, but I fail to see what else you could mean by this? There is plenty of analysis you could do here, but at 45th warmest on record since 1659, you would expect a Summer as hot or hotter than 2013 once every 8 years. Given that 2003 and 2006 were not long before that, and that 2016 was even hotter at 39th warmest, it still stands that it's pretty unreasonable to say that we are due a hotter one next year. You could argue that it's unfair to compare to years during the LIA, but since 1900 you only expect a 2013-or-hotter Summer once every 7 years. If we're talking really hot Summers, like an overall average of 17C+, then since 1900 you can expect one every 17 years. The last was 2006.

Posted twice for some reason, see the post below.

The way things are going and have gone for us in the past, I expect this month to be very mild. In spite of that, I'll go for 1.8C.

Attached is a graph of the December CET values through the years, with horizontal lines marking the 0C, 1C, and 2C values. Cold Decembers were very frequent during the late 1600s, and also throughout the late 1700s to the 1890s; since then the frequency of sub-2C Decembers has become less frequent. Sub-1C Decembers are interesting in that, with the exception of 1981 and 2010, they all appear to have occurred during clusters. There clusters were 1673-1683 which saw 5 such Decembers; 1784-1796 which saw 3 such Decembers; the duo in 1844 and 1846; and the period 1870-1890 which also saw 5 such Decembers. (Perhaps we haven't waited long enough to see 2010 sit in another cluster...?) To answer your question, since 1659, the occurrence of cold Decembers seems to have been fairly erratic. In the previous 100 Decembers however, 7 have registered as 2C or less (1916 just creeps in here). So that would give one every 14 years.

Sigh, you never change. Other people simply cannot have opinions that differ from your own. Your idea of a hot Summer (where all three Summer months are hot) is absolutely insane. Summer 2006, one of the hottest in recorded history, wouldn't fit this definition; August 2006 was not a hot month by any measure. It's like saying that a Winter can only be said to be cold if December, January and February are all very cold months themselves. And that very rarely happens. Your expectations are terribly unrealistic. We are not overdue a hot Summer if that is your definition. We never are. Such occurrences are extremely rare. And it's delusional to think otherwise. *** EDIT *** Since August 2013 does not fall into the hot category, with a CET anomaly of +1.1C relative to the standard 1961-90 average, then perhaps we should define a hot month as one where the anomaly is +1.5C or greater. In CET history, there have been just three Summers where all three months exceeded this anomaly: 1826, 1976, and 2003. That speaks for itself.

I guess it depends on opinion. July 2013 felt very hot at the time, most days had maxima in the upper 20Cs, with some 30C+ days in there too (1st August saw 34C I think in Oxfordshire). It was also a top 10 July for the CET. Lots of heat records were broken in Summer 2016, admittedly all during short, sharp plumes (of which there were many). From my perspective it may have felt hotter because I spent all of those Summers in the south-east. Not at all unusual to have markedly higher temperatures than the rest of the country during Summer.

Would disagree. Summer 2013 was very hot here in the south-east, 2014 was hot and very thundery, and 2016 was very warm throughout (May to September) with frequent plumes bringing 30C+ plus. Since I took an interest in the weather, Summer 2016 was the first time that 30C was breached in July, August, and September (multiple times during each of those months, might I add). As for 2017, here are my guesses... January: Generally dominated by south-westerlies from an Azores high influence. Very mild minima, maxima around the 10-12C mark for most places. A lack of frost. February: A continuation of January, until the final third, when a monster Scandi high sets in. Bitter easterlies will meet Atlantic storms to create blizzard conditions in some areas, with drifts up the sides of houses. March: The Scandi block persists, bringing more snow and bitterly cold temperatures. The final third of the month sees a thaw set in. April: A typical April with changeable conditions, sunshine and showers, hail at times. Occasional warm shots. May: Cool, cloudy, and damp, much like May 2013. June: Starting off very cool from a stream of northerlies, but warm shots later on. July: Mixed and overall average in terms of temperatures, sunshine, and rainfall. August: Warmer than average and thundery in many places. Wetter than average. September: Warm to start and a freak storm mid-month, before returning to more normal late-September conditions. October: Anticyclonic and cool, the first frosts of the season in many places. Sunnier and drier than average, with some notably cold minima. November: Another anticyclonic month, similar to November 2016. Plenty of frost. December: Atlantic kicks in, storms abound. Wet, windy, and very mild. Total guesswork, but probably as good as most other LRFs.

The provisional first-half (1st-15th) CET for this month (7.0C) ties with the years 1776, 1866, 1913, and 1931. The second halves of these months were 2.0C, 5.4C, 3.3C, and 3.7C respectively. Things can turn around, there is still some hope for a chilly spell before the New Year.

It always surprises me when people are desperate for snow even during cold, frosty periods of weather, like the late November/ early December spell. I for one always welcome waking up to -5C and a thick frost.

Happens often, people's predictions are too readily swayed by the first few days. July 2015 was a classic example of this.

As bonkers as December 2015 was, in terms of standard deviations from the series mean, May 1833 (15.1C), and October 1740 (5.3C) beat it. Taking the means and standard deviations for May, October, and December for the entire series, excluding the "bonkers" months (as anomalous) in each case, gives May: Mean = 11.21C, SD = 1.13 October: 9.73C, 1.30C December: 4.08C, 1.73C May 1833 is thus 3.45 SDs from the mean, October 1740 is 3.33 SDs from the mean, and December 2015 is 3.25 SDs from the mean. Having said that, May 1833 holds only two date records (15th at 20.0C, 17th at 19.7C). Unfortunately, no data exist for October 1740. Edit: I totally missed the Junes of 1676 (18.0C) and 1846 (18.2C), which can also claim to beat December 2015 in this regard, at 3.51 and 3.72 SDs above the mean respectively (here, both months were excluded when calculating the SD). April 2011 (11.8C) also comes close, at 3.22 SDs away, as does January 1795 (-3.1C), at 3.24 SDs away. There may well be more but these seem like the obvious candidates to me.

Best month of the year for daylight hours, enjoy it while it lasts!

Yesterday's provisional daily value of 1.0C is the coldest November day since 2010. Certainly felt cold!

Six years ago today I woke up to a white blanket of snow. This morning, it was a white blanket of a different kind: a thick frost. A lovely end to Autumn 2016.

A remarkable statistic from last year's December was that the second half (16th-31st) saw a mean of 10.4C. That's a whopping 1.8C up on the next warmest, 1987 (8.6C); 1.2C warmer than the warmest first half, 1898 (9.2C); 0.8C warmer than the warmest second half of November, 1994 (9.6C); 0.3C warmer than the warmest November, 1994 (10.1C); Only 6 first halves of November have ever equalled or beaten it; Only 5 first halves of April have ever equalled or beaten it.

I'll go reverse-Craig: -2.0C.

What's going on? Do we abandon our previous theories for the record high ice extent/ area? The ozone hole is still there and now we're at record lows...