Often forgotten

The northern Isles are very often forgotten in weather forecasts on the media, and often weather there is much better than it is further south on the mainland, and that’s been true in recent weeks with some lovely sunny days in the Shetland Isles. Here are the daily sunshine totals and accumulations for the last month from Lerwick.

Sunshine at 03005 Lerwick - United Kingdom 82 AMSL [13 May - 13 June 2016]
Sunshine at Lerwick  [13 May – 13 June 2016]

As you can see over the last month there has been well over 200 hours of sunshine and four days had totals of over 15 hours, because that’s the other thing this time of year in the northern Isles, the days are very long and the nights short. The sunrises at Lerwick at around 3:39 AM and doesn’t set till around 22:34 PM in mid June, which gives a day length of almost 19 hours. Mind you it never gets desperately warm up there thanks to the cold seas that surround the islands, here are the hourly temperatures from the last week.

03005 Lerwick - United Kingdom 82 AMSL 6 June-12 June 2016
Temperature, dew point and humidity at Lerwick 6 June-12 June 2016

Unpredictable weather

Courtesy of Google
Courtesy of Google

I did find it quite perverse that Tomasz Schafernaker should start his forecast on the BBC news channel (1030 AM) by stating how “unpredictable” the weather over the UK would be during the coming week. I can see what he means, and why he said it, but the whole point of a weather forecast is that it is a prediction, and if it’s going to be so unpredictable why even bother trying to forecast it! I think the problem is psychological though. When a presenter (or anyone else for that matter) gazes on a series of forecast charts like the one below, and knowing full well he has less than a minute to put a bit of detail on it knows he’s on a hiding to nothing. What can he say, other than it will remain unsettled in most areas, with showers or longer periods of rain but there will be drier and brighter interludes. Sometimes the messenger has to bring bad as well has good news, something I think weather presenters of the past did a lot better than the ones of today, they seemed to have an emotional detachment to the bad (or good) weather they were forecasting. Perhaps it’s the media pressure they are under to deliver good weather, I know during my career one of the classic lines that I got from people outside the office was “all this bad weathers your fault” and I was just an observer!

Surface UK [T+66] 18 UTC on Tue, 14 June 2016 (courtesy of Ogimet)
Surface UK [T+66] 18 UTC on Tue, 14 June 2016 (courtesy of Ogimet)

Sometimes the graphics that they use doesn’t help, especially if they are showing rainfall accumulations over periods longer than an hour in showery conditions rather than spot intensities. The above forecast frame looks much worse than it does because the coloured rainfall contours are six-hour accumulations, so it appears like its wall to wall rain with no dry or sunny breaks between the showers. As far as I can see the BBC Weather presenters do tend to use this type of graphic, especially with forecast frames beyond T+36, and they are normally static which gives the impression that’s it’s going to be that way for much of the day.

Weather impacts on Queensferry crossing

Forth road bridges new and old (courtesy of Scottish government)
Forth road bridges new and old (courtesy of Scottish government)

I was intrigued to hear that the opening of the Queensferry crossing had been delayed for five months by high winds in April and May this year. I’m not going to get into the politics of the whole thing on how delays during just two months can set the whole project back five months defies any sensible kind of logic, but I won’t dwell on that side of things.

Forth Replacement Crossing (courtesy of Wikipedia)
Forth Replacement Crossing (courtesy of Wikipedia)

The weather station at Gogarbank

I am sure that there are any number of wind vanes and anemometers affixed to the bridge and the cranes that are helping build the bridge, but I am not privy to that data, or the anemograph traces that they produced during April and May of 2016, so all that I can do is look for the closest SYNOP station reporting hourly wind speeds and that happens to be from Gogarbank in Edinburgh, in fact if you look at the map below, Gogarbank is remarkably only just over 5 miles to the SE of south Queensferry.

Queensferry - Gogarbank (courtesy of Google maps)
Queensferry – Gogarbank (courtesy of Google maps)

Gogarbank is a strange location to site a new weather station because it looks like a lot of money has been spent setting up a duplicate station to the one that must have existed for many years at Edinburgh airport. It’s an estimated WMO class 4 site as far as temperatures are concerned, but must be more representative of temperatures that the airport was. I won’t go into the obvious question of why the Met Office duplicated all the sensors at Edinburgh airport with a new site just 1.7 miles to its S’SE – perhaps they plan to do the same at other airports around the UK such as Heathrow? Gogarbank is 57 metres above sea level and looks well exposed to most directions.

Edinburgh airport and Gogarbank
Edinburgh airport and Gogarbank (courtesy of Google)
03166 - Edinburgh Gogarbank
03166 – Edinburgh Gogarbank (courtesy of Google)

Below is a graphic showing how wind speed increases with height, and as you can see a 16 MPH surface wind would equate to a wind of over 30 MPH at crane height of 200 metres (x 1.875) . The safe working limits are between 25 and 31 MPH, so using the winds from Gogarbank we are looking at surface winds of no more than 13 to 17 MPH, either as a mean speed or in any gusts. I’m not going to try adjust the winds for height, but obviously they must be a little higher at Gogarbank than places closer to the Forth, but how do you factor in winds that are blowing off an estuary rather than over farmland and allow for all the added friction?

Forth wind speed profile (courtesy of Scottish government)
Forth wind speed profile (courtesy of Scottish government)
Weather impacts on the Queensferry crossing
Weather impacts on the Queensferry crossing

Above is what the Forth Crossing Bridge Constructors [FCBC] have said about the delays caused by the wind and weather during April and May of 2016. And below is the anemograph trace from hourly SYNOP data that I’ve put together for the two months in question. Just note that the anemographs show wind speeds in knots and not MPH, so the limits using these graphs are roughly 11 to 15 knots, which is just a little bit less than Beaufort force five.

Winds in April 2016

03166 Edinburgh Gogarbank - United Kingdom 57 AMSL 1 April-30 April 2016
03166 Edinburgh Gogarbank – United Kingdom 57 AMSL 1 April-30 April 2016

April was I would have thought a very typical month with winds generally from the west but with easterlies for short spells (but not exclusively) between the 2nd-4th, 10th-15th, 21st and the 23rd. At a quick glance I would have thought that the 1st, 6th, 11th-12th, 17th-18th, 25th-26th, and the 29th would have been out of limits, at total of around 9 days.

Winds in May 2016

03166 Edinburgh Gogarbank - United Kingdom 57 AMSL 1 May-31 May 2016
03166 Edinburgh Gogarbank – United Kingdom 57 AMSL 1 May-31 May 2016

In May there were prolonged spells of easterlies between the 6th-14th and the 23rd-31st. I should imagine an easterly and the longer sea fetch may have caused then more problems than other directions, but I could be wrong. Wind speed were out of limits at a guess for the 1st-5th, 7th, 11-12th and the 19th-21st, a total of 10 days by my rough and ready reckoning.

In conclusion

So my figures are for 9 and 10 days out of limits are a little lower than FCBC’s 13 and 12 days, a little lower but no real conspiracy theory. Reasons for the differences are the fact that my wind speed data is for a site 5 miles to the S’SE and over land and not water, and of course the fact that I don’t have to worry about anyone’s life when I do my guessing. All that I can say is with these low limits, is that they have done a pretty good job of building a fantastic bridge with 35,000 tonnes of steel and 150,000 tonnes of concrete in the timescales that they’ve achieved, I can’t wait to drive over it in the future.

Storm Troupers parts II and III

After my initial review of the first part of the documentary “Storm Troupers: the Fight to Forecast the Weather review”, I was quite looking forward to the second and third parts, and overall I found both to be very informative. In part II, it was fascinating to hear that despite the introduction of the Norwegian cyclone model just after World War I, the United States did not formally analyze fronts on surface analyses until late 1942, and how the D-day forecast came about, despite the best efforts of the American forecasters, led by Irving P Krick almost leading them astray.

In part III there were some interesting images of past weather disasters at the Met Office. These of course included the evening forecast of ‘hurricane, what hurricane?’ from Michael Fish, and a rather squeamish performance by Ian McCaskill in what looks like a live interview interrogation on the one o’clock BBC news by Michael Buerk the next day.

October storm 1987 washup with Ian McCasKill
October storm 1987 wash up with Ian McCaskill (Courtesy of the BBC)

I had a lot of sympathy for Michael ‘hurricane’ Fish, for what the program called the most notorious forecast ever in the UK, because technically he was quite correct because it wasn’t a hurricane. He – or should that the be the model – got it wrong regarding the extent of the hurricane force winds in that the vigorous low tracked further north along the south coast so that the strongest winds were not just limited to northern France as expected.

Hurricane what hurricane with Michael Fish
Hurricane what hurricane with Michael Fish (Courtesy of the BBC)

The other publicity disaster was of course the ‘Barbecue summer’ fiasco of summer 2009, and the program included footage of Darren Bett making a right mug of himself for posterity. This was all probably due to some over energetic member of the press team trying to spice up some unexciting phrases in the three-month outlook like ‘temperatures slightly above normal’ and ‘rainfall slightly below normal’. Of course the corporate answer to this is stop publishing seasonal forecasts and promote the man responsible out of the press office as quickly as possible. Perhaps just as in the October storm and Michael Fish, we are all too guilty of shooting the messenger, when we really should be shooting the supercomputer that produced the forecast!

Barbecue Summer with Darren Bett (Courtesy of the BBC)
Barbecue Summer with Darren Bett (Courtesy of the BBC)

Then there was a look at the first NWP forecast for the Met Office for 00 UTC on 3rd November 1965. I don’t think that I can fully agree with Alok Jha description of a forecast of ‘calm and settled weather for most of the country’ from this first NWP chart for 00 UTC, I would have thought that there were at least moderate to fresh northwesterly winds across the country on that day, with showers in the north, but of course pressure must have been rising quite smartly as the low moved away into the Baltic, and winds would have been decreasing through the day, so maybe a score draw on that point.

The first NWP forecast chart (00 UTC 3 November 1965)
The first NWP forecast chart (00 UTC 3 November 1965) Courtesy of the BBC

It might be me, but I did detect a whiff of the same intellectual arrogance coming from Dame Julia Slingo as I did back in 1987 when Michael Fish arrogantly brushed asides the concerns of some woman phoning the BBC worried about an impending hurricane. I think that it might have been a little over the top when she compared the development of Numerical Weather Prediction [NWP] with the mapping of the human genome. The cost of computing has fallen over the decades since the 1950’s as the power of the processors has increased enormously, it was an inevitability that someone use computers to refine and finish the work that Lewis Fry Richardson had started. Now that we realise the effect that chaos and uncertainty has on NWP forecasting, and the limitations of forecasts beyond five days we are still a long way from cracking the forecasting of weather in mid-range let alone the long-range even with the use of ensemble probabilistic techniques rather than the old single deterministic models. On a number of occasions last winter, the Met Office were still dithering about the exact track and severity of a low less than 36 hours ahead of a possible named storm event. I guess this must occur when the synoptic situation is particularly chaotic, and the various ensemble members are split on the possible outcomes. If we could weave our way through all this chaos – which is probably totally impossible (the example of the double pendulum that Alok Jha demonstrated was very good) – and generate accurate forecasts out to ten days or more, then I might agree with her.

So you thought it’s been a warm start to June?

You may have thought it’s been quite a warm start to June, but it pales into insignificance when you look back at the Central England Temperatures [CET] for the first eight days of June over the last 244 years. So far the mean anomaly for the first eight days of June 2016 has been +2.15°C above average, making it the 32nd warmest start since 1772. The warmest first eight days of June occurred in 1982 when the mean anomaly was +5.68°C, over three degrees higher. I suppose we can console ourselves that 1976 had a cooler start than 2016 has had, but then again there looks like there is a distinct change of type in the offing if the NWP models are anything to go by.

Warmest 1-8 June 1772-2016
Warmest 1-8 June 1772-2016

Interestingly there has been a very, very slight cooling trend for the mean temperature for the first eight days of June in the last 244 years.

CET 1-8 June 1772-2016
CET 1-8 June 1772-2016

The latest climate statistics

I’ve written some additional code for my SYNOP application to generate images of various climate statistics derived from the SYNOP reports that I download from the internet. As a starter, I thought that I would list the thirty warmest, coldest, wettest and sunniest SYNOP stations from WMO block #03, which comprises over 150 stations across the United Kingdom and Irish Republic. I’ll generate and then upload the images in my application to my website, and then display the small images in widgets down the sidebar, as you will most likely have noticed. To be honest it would probably be more efficient if I wrote a bespoke widget that took a CSV file of places and values and displayed them in a ranked tabulated list, but for the moment this solution will have to do, because I have no experience of writing WordPress widgets in Javascript and PHP. All I’ve got to do now is to remember to run the code each day to update the images!

Wettest place
Warmest Place
Coldest Place
Sunniest Place

Recent sunshine totals

Daily sunshine 24 May - 6 June 2016
Daily sunshine 24 May – 6 June 2016

Excuse the rant

I saw one of the BBC forecasters today show some sunshine statistics for the UK and the stark difference between the totals from the east and the west. That got me to thinking that I’ve never really done very much with sunshine stats from SYNOP observations, so I set myself a little task to knock up an application to do just that this afternoon. <rant>The reason why I haven’t done it before is probably due to the lack of sunshine reporting stations in the UK which are released by the Met Office! It might have its headquarters in Devon, but there is just one sunshine reporting station in the whole of Cornwall and Devon, have a look at the coverage in France and Germany and see the difference. There are many climatological stations out there that the Met Office are just never going to release the data for – this side of hell freezing over – which I think is a very great shame, in fact it’s worse than that, it’s a scandal, after all it’s our – the public of this country’s – data. My plea to the Met Office is to please release as much climatological data as you can, you might not want to do much with it, but I certainly can. The sad thing is that I wouldn’t even be able to process this data and present you with these maps, charts and tables if it wasn’t for a Spanish web site, if I depended on the Met Office I would have no observational SYNOP data, they would rather just sit on it </rant>.

The sunshine data

Anyway getting down from my soapbox and back to the sunshine totals from the last two weeks, below is a tabulated ranked list of the 52 sites across the UK, and as you can see Tiree tops the list with over 170 hours in 14 days, or 12.2 hours per day and 72.6 of the theoretical maximum available – if my astronomical functions are working correctly. Bottom of the list is poor old Leconfield on Humberside with just 21.3 hours in the two weeks, or just over 20% of the possible maximum. Having said that there are a great many cloudy stations at the bottom of the list, and all in the east of the country, all in all a very interesting spell of weather. Most of the southern and central European countries support the reporting of daily sunshine data in their SYNOP/BufR observations. The trick is because of the various time zones across Europe there are some countries that report sunshine at midnight rather than 06 UTC, so I still have some work to do for the more eastern countries to get the maximum coverage.

Daily Bright Sunshine 24 May 2016 - 6 June 2016
Daily Bright Sunshine 24 May 2016 – 6 June 2016