Met office Hadley              16.2c.             Anomaly        2.6c.  Provisional to 28th.

Metcheck                           16.20c.          Anomaly         2.48c.

Netweather                        16.83c.          Anomaly         3.14c.

Peasedown St John         16.2c           Anomaly           1.1c.    

Phew - looking alright this month.

For my location, I've been tracking the mean temperatures-to-date since mid-month in comparison with the record-warm September of 2006. Some of the model runs prior to the hot spell made much more of it and had 2016 leaving 2006 in the dust (see that faint brown line that represents the incredible GFS 00z op run of 8th September? Mind-blowing, that was). The reality was so much tamer (yet still exceptional with several date records set) that this year was left with more work to do in order to snatch the record - and amazingly enough, it currently appears that it will manage to pretty much tie with 2006, provided tonight comes in at around 9*C and tomorrow around 18*C, which is what GFS has consistently predicted for many days now. 

In fact, using those GFS numbers combined with observation data up to and including today, the final value is 16.62*C for 2016, which compares well with the 16.62*C of September 2006... seriously, I'm not kidding!

The slightest of departures from the GFS numbers will make or break it. Last night was exactly as predicted but today climbed over half a degree above the GFS figure. I worry about tonight though; the model has in recent times seemed to struggle to capture the full extent of cooling when there is a lot of clear sky (or thin high cloud cover) and a bit of a breeze.

I am so weirdly excited but concerned right now 

Phew - looking alright this month.

For my location, I've been tracking the mean temperatures-to-date since mid-month in comparison with the record-warm September of 2006. Some of the model runs prior to the hot spell made much more of it and had 2016 leaving 2006 in the dust (see that faint brown line that represents the incredible GFS 00z op run of 8th September? Mind-blowing, that was). The reality was so much tamer (yet still exceptional with several date records set) that this year was left with more work to do in order to snatch the record - and amazingly enough, it currently appears that it will manage to pretty much tie with 2006, provided tonight comes in at around 9*C and tomorrow around 18*C, which is what GFS has consistently predicted for many days now. 

In fact, using those GFS numbers combined with observation data up to and including today, the final value is 16.62*C for 2016, which compares well with the 16.62*C of September 2006... seriously, I'm not kidding!

The slightest of departures from the GFS numbers will make or break it. Last night was exactly as predicted but today climbed over half a degree above the GFS figure. I worry about tonight though; the model has in recent times seemed to struggle to capture the full extent of cooling when there is a lot of clear sky (or thin high cloud cover) and a bit of a breeze.

I am so weirdly excited but concerned right now 

Great post there. Excellent. :-)

We've also done well here in the Itchen Valley: currently (to 28th Sep) 0.5C warmer max and (to 29th) 0.4C warmer min than 2006, so the last two days will be interesting. Reached 20C here today.

Phew - looking alright this month.

For my location, I've been tracking the mean temperatures-to-date since mid-month in comparison with the record-warm September of 2006. Some of the model runs prior to the hot spell made much more of it and had 2016 leaving 2006 in the dust (see that faint brown line that represents the incredible GFS 00z op run of 8th September? Mind-blowing, that was). The reality was so much tamer (yet still exceptional with several date records set) that this year was left with more work to do in order to snatch the record - and amazingly enough, it currently appears that it will manage to pretty much tie with 2006, provided tonight comes in at around 9*C and tomorrow around 18*C, which is what GFS has consistently predicted for many days now. 

In fact, using those GFS numbers combined with observation data up to and including today, the final value is 16.62*C for 2016, which compares well with the 16.62*C of September 2006... seriously, I'm not kidding!

The slightest of departures from the GFS numbers will make or break it. Last night was exactly as predicted but today climbed over half a degree above the GFS figure. I worry about tonight though; the model has in recent times seemed to struggle to capture the full extent of cooling when there is a lot of clear sky (or thin high cloud cover) and a bit of a breeze.

I am so weirdly excited but concerned right now 

That's interesting because my stats show that this month has been substantially cooler than September 2006. I recorded a mean temperature of 17.89C in September 2006 which was warmer than the 17.71C recorded in August that year. There were 24 days in September which saw the temperature exceed 20C with 6 exceeding 25C.

For September 2016 my mean currently stands at 16.65C which is very close to your figure for this year but well down (more than 1C) below my 2006 figure. Only 16 days have reached 20C and 4 have exceeded 25C. 

So 2006 must have been significantly warmer inland towards the Thames Valley as opposed to down near the south coast.

Just checked my COL archives and most parts of Berkshire and north Hampshire saw means of 17.5C to 18C in September 2006. Places nearer the coast towards Portsmouth also saw very high figures - Porchester had 18.3C and Portsmouth Uni 18.4C. Don't have many stations near your location but there is one at West Totton where the mean was only 17.1C so quite a lot cooler. No doubt it was cooler still in rural areas around the New Forest. So I can well believe your 16.6C figure.

Latest data suggests we should just squeeze out a final CET of 16.0C this month.

By my estimated calculations the CET stands at 16.15C up to today. A slightly cooler day tomorrow (especially minimums tomorrow morning although 9C is still good for the time of year!) means the CET is expected to fall to 16.02C by this time tomorrow. 

So this would be only the second time we have seen 16C in September since 1949. 

 

Can you hear that? It's my chair clattering on the floor 

Until xioni2's late entry, that was the highest punt, as far as I can see? Here's hoping we're seeing clearly the way of the weather 

Thought I would go back and repost this as it is always interesting to look back and reflect. As I said at the time I was sorely tempted to go above 16C but decided to hold off. As it turns out my prediction of 15.9C was very close to the mark. A lot of luck in this of course as I was really basing it on a very warm first half which looked fairly nailed on and then guessing there would only be a gradual drop in the mean during the second half of the month.

I think September will end on 14C (+1.4C) here

I think September will end on 14C (+1.4C) here

Quite a difference to my location, where I expect the final figure to be around 17.6°c

Last nights chilly night might have knocked it down a bit.

Latest data suggests we should just squeeze out a final CET of 16.0C this month.

By my estimated calculations the CET stands at 16.15C up to today. A slightly cooler day tomorrow (especially minimums tomorrow morning although 9C is still good for the time of year!) means the CET is expected to fall to 16.02C by this time tomorrow. 

So this would be only the second time we have seen 16C in September since 1949. 

It will be amazing to end with a 16.0C or over. Impressive by any means.

Met office Hadley              16.2c.             Anomaly          2.6c.     Provisional  to 29th.

Metcheck                          16.08c.            Anomaly           2.36c.

Netweather                       16.79c.            Anomaly           3.1c.

Peasedown St John        16.2c.             Anomaly           1.1c.                        

 

That's interesting because my stats show that this month has been substantially cooler than September 2006. I recorded a mean temperature of 17.89C in September 2006 which was warmer than the 17.71C recorded in August that year. There were 24 days in September which saw the temperature exceed 20C with 6 exceeding 25C.

For September 2016 my mean currently stands at 16.65C which is very close to your figure for this year but well down (more than 1C) below my 2006 figure. Only 16 days have reached 20C and 4 have exceeded 25C. 

So 2006 must have been significantly warmer inland towards the Thames Valley as opposed to down near the south coast.

Just checked my COL archives and most parts of Berkshire and north Hampshire saw means of 17.5C to 18C in September 2006. Places nearer the coast towards Portsmouth also saw very high figures - Porchester had 18.3C and Portsmouth Uni 18.4C. Don't have many stations near your location but there is one at West Totton where the mean was only 17.1C so quite a lot cooler. No doubt it was cooler still in rural areas around the New Forest. So I can well believe your 16.6C figure.

The 5km gridded data from the Met Office for September 2006 shows how the heat was especially concentrated towards the SE and along the South Coast

When yours and Stormchasers likely locations (from your tag lines) are cross referenced to the Meto 5x5 grid codes you can see that you were exactly right and the temperature reduced quite quickly moving west

These are the 2006 UK and SE England temperature charts based on the UKCP09 MetO grid data

Thought I would go back and repost this as it is always interesting to look back and reflect. As I said at the time I was sorely tempted to go above 16C but decided to hold off. As it turns out my prediction of 15.9C was very close to the mark. A lot of luck in this of course as I was really basing it on a very warm first half which looked fairly nailed on and then guessing there would only be a gradual drop in the mean during the second half of the month.

Haha yes, I can't say my approach to the CET punt was much different to yours 

Thanks for those fascinating stats regarding how this year compares to 2006 across much of the S/SE. I have been tracking London Heathrow as well as my location and can confirm that this year will fall well short 2006 there (difference looks to be around -0.8*C).

From daily records, it seems among the most notably colder days in my locality were early and late in the month; 1st-3rd comes in at a mean of only 17.3*C here compared to 19.9*C at Heathrow, after which that location spends three days climbing several degrees higher than mine. At the other end, the final four days were maxing in the 17-18*C range by day here, while Heathrow was seeing 19-20*C.

Proof that if I want exceptionally high temperatures, I need to get away from this valley, which often seems to channel sea breezes along it and has a habit of being a spawning ground for lingering murkiness and low cloud well into the mornings - these being possible explanations for the relatively cool daytime highs here.

I'm hoping to be living not far from Reading before long now. Could end up near High Wycombe though, which is hardly a hot spot 

 

The 5km gridded data from the Met Office for September 2006 shows how the heat was especially concentrated towards the SE and along the South Coast

When yours and Stormchasers likely locations (from your tag lines) are cross referenced to the Meto 5x5 grid codes you can see that you were exactly right and the temperature reduced quite quickly moving west

These are the 2006 UK and SE England temperature charts based on the UKCP09 MetO grid data

 [Images cut to save scrolling time]

Wow, those are fantastic, thanks 

Intriguing to see that especially 'cool' spot just west of here - but that's an area of slightly higher ground, which is probably the main reason for that.

Interestingly, Sep 2006 was the 8th sunniest in my records so nothing special there, while Sep 2016 is looking to be the third dullest. This implies that a predominance of unusually warm airmasses, resulting in exceptionally high nighttime minimums, is the main player here. This is confirmed when looking at the mean minimum stats; 2006 came in at 12.67*C and this year will be 12.49*C or perhaps 12.50*C, and both these years are well above 2011, which is now in third place with its figure of 11.74*C.

Yet this theory threatens to be unhinged when looking at the mean maximum for this month; it looks to be 20.70 to 20.78*C depending on today's value, which will be around 0.15*C above the 2006 figure. That doesn't fit with cloudy skies - but really it's down to the mid-month hot spell, which was extreme enough to put 2016 in contention for the record when otherwise it would have had little chance.

The numbers are in, and 2016 has clinched it... by a thousandth of a degree Celsius.

I honestly have never seen something so improbable in the world of statistics.

The numbers are in, and 2016 has clinched it... by a thousandth of a degree Celsius.

I honestly have never seen something so improbable in the world of statistics.

I'm sure I would share your celebration if I knew what I was looking at SC. Sorry to be a pain but could you please explain as I'm probably missing out on something good!  

Caz- I think the 9 columns are: max/min/daily mean/running (cumulative) mean, 2006; then the same 2016; then the comparator. I think as these are rounded, they don't show the finer resolution mean, which obviously extends into the thousandths and beyond. I might stand corrected by Stormchaser however.

Final figure for September is 16.03C. So we did manage to stay above 16.

Daily HadCET mean.txt (official 2016 so far)

Congratulations to Tim S and ozone_arora who were the closest this month with 16.00C

16.2C

http://www.metoffice.gov.uk/hadobs/hadcet/cet_info_mean.html