Summer Statistics?

[mushymanrob]

well before i could reply to pauls post regarding 'rip summer', he closed it! :lol:

he cites that recorded statistics are the only way to truely record how good/bad the summer factually is.

but statistics are flawed.... why?.... well what if for eg, the rainfall came overnight? but the daytimes were largely sunny? wouldnt that skew peoples perception?

look at earlier this decade, temps were often up being muggy, but under very milky skies which statistically reduced actual sunshine levels!

the cet for summer 07 was rather high...but that was due to high nightime minima as opposed to high daytime maxima, a warm summer?...lol.

these examples highlight why imho statistics cannot be relied on to give a picture of what our summer weather is really like.

perhaps the most accurate way would to be statistics from say... 9 am - 9 pm... after all, what matters what the weather does when we are asleep!

[Thundery wintry showers]

I think most people would be happier with a wet summer if the rain fell predominantly overnight and the days were sunny. The statistics may not show that it was wet overnight but they'd show higher sunshine totals. Of course this would be less good for people whose outdoor work relies upon summers being relatively dry.

With regards the max/min temperature problem, the simple solution is to give data for maximum and minimum temperatures rather than just mean temperature. Similarly with rainfall, number of rain days or wet days is often a more representative measure than rainfall totals, as in some months you can get a large percentage of the total occurring within a single frontal zone or thunderstorm.

One point I thought of raising in the other thread was Kevin/Mr_Data's idea of a summer index. A summer index is also subject to personal opinions/preferences (with regards what variables you choose and what weights you give them) but it at least gives a solid statistical overview of how good or bad a summer was (at least in the opinion of whoever devises the index) rather than pure perception. Also I think Kevin's index is pretty representative of the "majority view", certainly more so than the cruder Davis index for southern England (for example it uses max temperature and rain days rather than mean temperature and rainfall total). Other people can use similar methodologies to devise their own indexes for their own locations.

[Methuselah]

well before i could reply to pauls post regarding 'rip summer', he closed it! :lol:

he cites that recorded statistics are the only way to truely record how good/bad the summer factually is.

but statistics are flawed.... why?.... well what if for eg, the rainfall came overnight? but the daytimes were largely sunny? wouldnt that skew peoples perception?

look at earlier this decade, temps were often up being muggy, but under very milky skies which statistically reduced actual sunshine levels!

the cet for summer 07 was rather high...but that was due to high nightime minima as opposed to high daytime maxima, a warm summer?...lol.

these examples highlight why imho statistics cannot be relied on to give a picture of what our summer weather is really like.

perhaps the most accurate way would to be statistics from say... 9 am - 9 pm... after all, what matters what the weather does when we are asleep!

No, mushy. Paul didn't close it, I did...

But, from the argument of statistics, statistics are what make history?? The summer here has been far, far, far better than the last two...But, that's a local thing too? IMO, we cannot do an RIP-job on the summer just yet???

So, let's have some statistics????

[Paul]

Think this is a relatively new page on the meto site - allows you to plot a range of anomaly maps - tis quite good i think:

http://www.metoffice.gov.uk/climate/uk/anomacts/

[mushymanrob]

I think most people would be happier with a wet summer if the rain fell predominantly overnight and the days were sunny. The statistics may not show that it was wet overnight but they'd show higher sunshine totals. Of course this would be less good for people whose outdoor work relies upon summers being relatively dry.

With regards the max/min temperature problem, the simple solution is to give data for maximum and minimum temperatures rather than just mean temperature. Similarly with rainfall, number of rain days or wet days is often a more representative measure than rainfall totals, as in some months you can get a large percentage of the total occurring within a single frontal zone or thunderstorm.

One point I thought of raising in the other thread was Kevin/Mr_Data's idea of a summer index. A summer index is also subject to personal opinions/preferences (with regards what variables you choose and what weights you give them) but it at least gives a solid statistical overview of how good or bad a summer was (at least in the opinion of whoever devises the index) rather than pure perception. Also I think Kevin's index is pretty representative of the "majority view", certainly more so than the cruder Davis index for southern England (for example it uses max temperature and rain days rather than mean temperature and rainfall total). Other people can use similar methodologies to devise their own indexes for their own locations.

but even refering to the max - min temps are not free from skewing reality.... what if its cloudy all day until 3 pm? then the sun comes out and temps reach 25c? statisticall thatd look like a 'good day' but in reality most of it was overcast!

yep, a good idea about index's.

No, mushy. Paul didn't close it, I did...

But, from the argument of statistics, statistics are what make history?? The summer here has been far, far, far better than the last two...But, that's a local thing too? IMO, we cannot do an RIP-job on the summer just yet???

So, let's have some statistics????

oh apologies to paul

no probs

i know you have to have statistics, but statistics shouldnt be taken as reality... see my reply above regarding late afternoon sun..

statistically i bet theres more people moaning about this summer then the previous ones! :lol:

[Derp Patrol]

I do agree with your thoughts on the day/night conditions and how when recorded they can be misunderstood. I never really use the mean temp to say how good a summer has been, because what use is it if you have muggy nights, and cooler days? I also think that you have to take in all of the statistics to say how good a summer has been, a warm summer is of no use to me and I would guess the majority of people if it is wet and cloudy the whole time. So although this summer so far has been pretty average regarding temperatures, in my area at least it has been poor thus far because of high rainfall and cloud cover.

[Snowyowl9]

Statistics don`t tell the whole story by how much atlantic winds for the feel factor there might be blue skies/sun but it felt cool,and most of July was like that atlantic source and lots of blustery frequent showers at one point,with little sun more like autumn,besides earlier more S-ly.

Today on the other hand is much better light S/SW wind blue skies cumulus warmest day since July 4th at 21.4c

Last July average max here was 19.7c

This July cooler at 19c

2007 was 17.9c

Edited August 5, 2009 by Snowyowl9

[Nick.F]

An index is an interesting idea, but once again it would be a subjective measure, even if it were made up from empirical data.

I personally love warm muggy nights, and they are one of the best things in my "measure" of a good summer, other people can't abide them, so what "weight" to we give to the Minimum Night Temperature (that would be included in an Index, I assume). If we gave it a low weight it wouldn't reflect my view of warm nights as an important factor in a good summer. If we gave it a high weight in the index, others would be disaffected.

A prime example of how Indices do not reflect a balanced view of data, is the CPI (Consumer Prices Index) which is one of the prime indicators for a measure of how prices increase/decrease, and are used to determine economic policy and wages. This index is made up of a "basket" of goods and services chosen to be indicative of things that average households spend their money on. If you look at that list of things in that "basket", I defy anyone to say that is really typical. So the index means very little, other than to those who need to use it for their own purposes.

Indices can be useful-ish, but are rarely balanced. Bare statistical data doesn't lie, but indices rarely tell the whole truth and are just a convenience. Perhaps that's what we are looking for.

Apologies for wittering on, but an escapee ex-data analyst and statistician here.

[Thundery wintry showers]

My personal method is to give each day of the month a "good weather percentage", add up the scores and divide by the total number of days in the month. Practically, however, this would be of far less use to anyone else since it is purely measuring my own personal opinion, and it is also influenced by whereabouts I am at the time. For example June 2009 in Exeter scored 70%, whereas if I'd been in Tyne & Wear it's unlikely that the month would have even managed 60%.

The original Manchester index for reference is 10*(mean max temperature + (sunshine/67) - (rain days/ 8 )). If I was to modify that to give something that accurately reflected my summer weather preferences it would look more like this: 10*(mean max temperature + (sunshine/50) + (convective rain days/ 8 ) + (thunder days/2) - (non-convective rain days/ 8 )). That formula wouldn't be workable because there is no reliable way to distinguish days of convective vs non-convective rainfall.

However I find that the Manchester index does correlate pretty well with my perception of how good or bad a summer month is.

[Summer of 95]

I've posted before about how much I think bare monthly statistics of temp, rain and to some extent sunshine mask a month's true character; to get to the bottom of it you need to use more detail; day and night temps, rain days (that one I think is really important), and number of days with over say 10 or 12 hours sunshine (a month with 20-odd days that were sunny between 5 and 8 am, and from 7pm till dusk would show well over 150 hours but would be widely perceived as cloudy, especially by those who work outside from 9 to 5).

Regarding rain, some time ago I worked out a simple way to gauge how wet a month feels to me in a given location; (Total rainfall x Number of days >0.1mm)/(Number of days in month).

So say we had one July that was sunny and dry but with a couple of big thunderstorms each dropping 50mm, with only 2 more days each with 1mm, we get 102mm x 4 days = 408, divided by 31 gives an index of 13.16.

By contrast, An unsettled July month with 15 showery days each having 5mm, 4 more drizzly ones with 0.5mm, and one real washout day with steady rain giving 25mm also records 102mm, but the dampness index is 102 x 20 days = 2040, divided by 31 equals 65.8; so it "feels" about 5 times as wet as the month above.

[johnholmes]

seems not a bad idea at first sight

most full climat stations do have data 09-21z and 21-09z available; not sure how its accessed on the net though.

[J10]

It is very difficult to get one measure of summer weather that accurately sums up how good or bad a summer is, IMO Sunshine amouts and Max Temps are the 2 positive factors, with the amount of rainfall the major negative factor. When the rain occurs is another point, rain that falls in the night is much better than rain that falls in the day, and also the weekend weather can skew perceptions of a summer.

So that if you have a summer of good weekends and bad weekdays most people would rate that higher than a summer of bad weekends and good workdays.

[Eugene]

Here's some shocking stats for you, birmingham hasn't recorded any temps above 22C since July 1st :lol:

Wow even the last two summers haven't managed that and put the way above average rainfall amounts on top of that and i think i'ts been a worse July 1st to August 7th period than 2007/08.

http://www.weatheronline.co.uk/cgi-bin/detail?SID=03534fe7cf0bcdc4d9a3df209d1361b84b4d0&PROVIDER=anwendung&ART=temperatur&CONT=ukuk&WMO=03534&LANG=en

Goto history and change it to 8 weeks to see the max temperature graph.

[Weather-history]

Here's some shocking stats for you, birmingham hasn't recorded any temps above 22C since July 1st :lol:

.

I think you are mistaken there Eugene, because Manchester recorded 27C on the 2nd of July

According to this site, no temperature above 22C since 4th of July

http://www.tutiempo.net/en/Climate/Birmingham_Airport/07-2009/35340.htm

Edited August 7, 2009 by Mr_Data

[Derp Patrol]

A month without a temperature above 22C is truly shocking for Birmingam, down here it's been mainly in the low 20S, but we still haven't recorded a temperature above 25C for a month, which is very rare indeed.

[Nick.F]

Sorry kind of a cross post, but interesting nevertheless

Cloudy Morning, sunnier afternoon, max of 24. That's all but 4 of the last 58 days reaching 20+, and all but 1 of those 4 within 0.5 of 20 (all but one of last 58 days at 19.5+)

58 days since our last sunless day.

another weird stat only once since 27 March (that being 27 Apr) has pressure dropped below 1000mb.

mean daily temp been above 10 since 5th May.

only 7 days since Jan 01 have we recorded 10mm of rain or more. Only 4 since 2nd Feb. 122/219 days this year no rain recorded (55.7%), 158/219 days 1mm or less (72.1%) - so seemingly quite dry.

[Snowyowl9]

Just had the longest settled spell of the whole summer with 5 dry sunny warm days until last nights rain,4 days in late June,it can what it wants now,but it looks like a typical 90`s type spell of weather coming now with a few days of HP,then a few days of LP with rain/showers then back to more HP.

Edited August 10, 2009 by Snowyowl9

[Thundery wintry showers]

http://www.climate-uk.com/

August so far has been somewhat less wet than July but temperatures and sunshine have been more or less the same. I imagine the regional breakdown might look a bit different though.

[Weather-history]

http://www.climate-uk.com/

August so far has been somewhat less wet than July but temperatures and sunshine have been more or less the same. I imagine the regional breakdown might look a bit different though.

Hadley tends to be wetter than Philip's and is 32mm up to the 9th

No chance of this being a drier than average summer south of the Border now overall.

Already 109% of the 1961-90 summer average rainfall and we still got the rest of August to go yet.

This summer is already wetter than these summers that are often regarded as poor

1962, 1967, 1972, 1978, 1979, 1981,

1979 and 1981 are interesting, 1981 is in the top 20 driest summers on record for England and Wales. July 2009 was only 5mm away from overtaking the summer 1981 rainfall figure.

1979 had a dry June and July.

[Nick.F]

from 01 Jun to today we are at 121mm (average 148mm) so at 81.7% average rainfall. Remainder of august needs to give 27mm of rain to hit average.

If we include May it is at 80% of average.

Not been very wet here.

Edited August 11, 2009 by Nick.F

[Nick.F]

Stats update for NE Essex

2 sunless days since 1st May, 65 days since last sunless day

77 dry or trace rain (less than 1mm) since 1st May - 72.8% dry

65 Consecutive days of 19.5+ temps - all but 4 of the last 65 days reaching 20+, and all but 1 of those 4 within 0.5 of 20

A perfect British summer continues