This is an in-depth Met Office UK Weather forecast for the next week and beyond. Will there be any more snow this week? What will happen this weekend and beyond?
Bringing you this deep dive is Met Office meteorologist Aidan McGivern.
Bringing you this deep dive is Met Office meteorologist Aidan McGivern.
Category
🗞
NewsTranscript
00:00:00To say that I'm excited about this week's deep dive would be a massive
00:00:04understatement. Just ask any one of my colleagues, I've been talking about this
00:00:09for some time because in this week's deep dive I'm going to be joined by Rob
00:00:14Neill, a colleague here at the Met Office and we're going to be talking about a
00:00:17very exciting new visualisation tool for forecasting the medium range, for
00:00:23forecasting beyond a week ahead and for summarising what these forecasts mean
00:00:29for you. So we're going to be unveiling that in this week's deep dive which is a
00:00:35tool that we've been developing for some time, you could say months, could say
00:00:39several years as far as Rob's concerned. But before we get into all of that
00:00:45exciting stuff there's plenty of weather to talk about this week. Thank you for
00:00:50joining me on the Met Office deep dive, we do these every Tuesday and it's
00:00:55always a pleasure to see your comments and your questions and we try and turn
00:01:00some of those questions into topics for future deep dives so please do comment,
00:01:04hit like, hit subscribe, tell your friends about this if they don't know about it
00:01:08already. Now it's been an interesting few days if not longer for the weather of
00:01:14course we had the very mild weather over Christmas then it turned colder and some
00:01:19places have seen significant snow in the last few days. This image here shows the
00:01:25satellite and radar from the last few days, the UK is here outlined in black
00:01:30and what you can see playing through at the moment is Sunday into Monday. A
00:01:34weather system moved north across the UK of course brought significant snow,
00:01:38significant disruption in places. Snow across many central and northern parts of
00:01:43the UK but also heavy rain in the south and we've seen an awful lot of
00:01:47disruption from hazards such as snowfall and flooding. It has turned a bit quieter
00:01:54for now but the cold air has returned across the country and we're likely to
00:02:00see some more snow in places over the next few days so that's what I'm going
00:02:04to cover right now. Here's the general setup, Tuesday one o'clock we've got low
00:02:10pressure over Scandinavia, we've got another low in the Atlantic heading
00:02:15towards Europe and the UK there under the influence of cold northerly winds so
00:02:23just playing this forward for now and what you can see is that the whole of
00:02:27the UK is covered by cold air and as we go into the start of Wednesday a very
00:02:32cold start to the day across the whole of the country but it's this low that is
00:02:38of interest to people living in the south of England because this is likely
00:02:43to bring some precipitation, some rain in from the Atlantic and as that bumps into
00:02:48the cold air there's the possibility of a few flakes of snow but an awful lot of
00:02:53uncertainty even though it's just 24 hours away at the time of recording. The
00:02:58precise track of that low and by precise I mean 10 miles here or there will make
00:03:05a big difference in terms of rain, sleet and snow for southern England on
00:03:08Wednesday afternoon into the evening and the precise intensity of the
00:03:13precipitation, the wind direction, it's on a knife edge in terms of weather places
00:03:18in the south will get rain or sleet or snow or nothing at all and so I'm going to be
00:03:23covering that in just a moment. That's the general setup but meanwhile the rest
00:03:27of the UK covered by cold winds continuing to see sleet, snow showers
00:03:33especially across northern parts of Scotland but also parts of Northern
00:03:37Ireland into Northwest England and North Wales that's been the case through
00:03:40Tuesday it will continue to be the case over the next few days. Then heading into
00:03:46Thursday, well that low pulls away taking its milder wetter weather with us
00:03:52with it. High pressure then builds into the northwest and we've still got this
00:03:57cold northerly across the UK but it becomes a little more settled with that
00:04:01high temporarily moving across the country. Then another interesting setup
00:04:08heading into Friday another low from the Atlantic swings in and again bumping
00:04:14into the cold air with its band of precipitation. Once again the possibility
00:04:20of some snow in places but if you're after snow I wouldn't get too excited
00:04:25about this system unless you're in a more northerly location in the UK
00:04:30because further south it's going to be well much more patchy much more marginal
00:04:35but a change is introduced by that low it moves in and it eventually albeit
00:04:42slowly through the weekend brings much milder air in and then by the end of the
00:04:48weekend we're into this kind of setup we've got high pressure to the southeast
00:04:53and we've got this very mild southwesterly airflow outbreaks of rain
00:04:58in places particularly towards the northwest. That's a brief summary of the
00:05:02UK's weather over the next few days but of course we can also visualize it from
00:05:06this perspective all four panels on the screen at once and actually let's rewind
00:05:12a tad to take us off to the start of Wednesday. This is dawn on Wednesday so
00:05:20we've got the low sliding in its weather fronts angled so that they're likely to
00:05:25bring this band of precipitation in from the channel but for most of the UK
00:05:31Wednesday starts off with plenty of sunshine temperatures meanwhile widely
00:05:38below freezing. We'll be looking at those in just a moment in more detail but widely below
00:05:43freezing a very cold start to the day winds easing overnight of course it's a windy
00:05:47day across the north of Scotland at the moment enhanced wind chill but through
00:05:52Wednesday watch as this precipitation moves in in fact let's turning that off
00:05:58to get a bigger picture idea and zooming in to focus on southern England and you
00:06:08can see at midday we've got outbreaks of rain pushing into the south but as
00:06:14those outbreaks of rain mix with the cold air we've got these whites
00:06:18appearing and just suggestive there of a mix of rain sleet and snow but of course
00:06:27it will depend on whether the winds are coming from the south coming more from a
00:06:31maritime direction bringing milder air in from the sea or whether they're coming a
00:06:35bit more as you can see they're coming from the east so they're coming a bit
00:06:39more from the land that would make things a touch colder it would also
00:06:43depend on how heavy this precipitation is because of course we're on the
00:06:47northern flank of that low on the northern flank of the band of rain so
00:06:51it's likely to be light but if it turns a bit heavier it will cool the atmosphere
00:06:56a little more and result in a bit more snow and equally it will depend on the
00:07:02track of that low if it's a little further south won't get very much of
00:07:06that precipitation at all if it's a bit further north will get a bit more so the
00:07:11ingredients are finely balanced and it's just one of those situations that is so
00:07:18marginal that we often see in the south of England a difference in temperature
00:07:23of half a degree here and there which could be caused by heavier
00:07:26precipitation hence the cooling of the atmosphere more slight change in wind
00:07:31direction a slight difference in the track of that low a lot hinges on the
00:07:37flavor of that precipitation in the south not forgetting of course there is
00:07:41more substantial snow further north in fact going back to Tuesday afternoon
00:07:46we've seen frequent snow showers into the north of Scotland Northern Ireland
00:07:50into North West England North Wales and significant snow continuing to build up
00:07:54here after some decent snow over the weekend and the start of this week so at
00:08:02the moment the Met Office do have a number of weather warnings in force and
00:08:07actually going back to Tuesday's weather warnings we've got this snow warning
00:08:11across the north and west of Scotland Northern Ireland snow and ice hazards
00:08:15and in these areas there's the risk of two to five centimeters of snow building
00:08:21up with showers so very hit-and-miss very localized amounts and differences
00:08:26from here and there but over higher parts above 200 meters or so for
00:08:30northern Scotland the potential for more than 10 centimeters of snow to build up
00:08:35through Tuesday and similar for Wednesday as we go into Wednesday
00:08:40actually let's rewind a bit because we've still got the start of that day
00:08:44that's snow and ice warning in force but then we've got this snow warning in the
00:08:49south of England now when you read the text of the snow warning in the south of
00:08:53England it mentions widely two to five centimeters and a risk of 10 centimeters
00:08:59in places but it's worth reading as well about the likelihood now the warning
00:09:06talks about a low likelihood of significant impacts from those depths of
00:09:12snow because that's a worst-case scenario so we're talking about worst
00:09:17case scenario two to five centimeters ten centimeters over hills in the south
00:09:21and that would depend on a particular track of the low particular intensity of
00:09:25the precipitation and so on more likely more likely is a couple of centimeters
00:09:32over Dartmoor and zero to one centimeters in patches across southern
00:09:39counties of England so that's the more likely scenario but the warning of
00:09:43course covers a lower likelihood of more significant impacts so it's worth
00:09:48bearing that in mind warnings take into account not only the impacts which of
00:09:53course in southern England you get significant impacts a lot of people
00:09:56living in these areas a lot of busy roads coming at the end of the day on
00:10:00Wednesday rush hour and so on you'd get significant impacts from two to five
00:10:04centimeters widely 10 centimeters over hills but these warnings also take into
00:10:09account the likelihood of these impacts and so this one mentions a low
00:10:12likelihood of these kinds of impacts with a more likely scenario where we
00:10:18get significantly less snow but like I say conditions are finely balanced
00:10:23higher likelihood that these snow amounts will continue across the north
00:10:27and west of Scotland and Northern Ireland and similar conditions across
00:10:31northern parts of the UK through to the end of the week so let's take a look in
00:10:36a bit more detail actually first of all I just want to talk about temperatures
00:10:40over the next few days as you can see the max temperature over the next few
00:10:43days is staying significantly below normal for the time of year we're
00:10:47talking about one to four Celsius typically and when you add on the
00:10:51windchill believe me this is Tuesday afternoon and these are the feels like
00:10:55temperatures comparing them to the actual temperatures these natural
00:10:58temperatures one to four Celsius feels like minus four minus six for Shetland
00:11:03where it's particularly windy likewise going to Wednesday those feel like
00:11:07temperatures are significantly below freezing in the north not quite as low
00:11:10because the winds aren't quite as strong similar on Thursday the winds are coming
00:11:15down a little by the end of the week however overnight temperatures are going
00:11:20to be particularly significant so we start off Wednesday for example with
00:11:27temperatures of minus eight minus seven just running my pen across the map there
00:11:33anywhere where we've got lying snow we could see minus double digits but more
00:11:40widely minus one to minus four or five Celsius across the UK so we're starting
00:11:45off very cold we've got that low moving in it bumps into the cold air what does
00:11:51it do we've shown these graphics before in the deep dive this shows the freezing
00:11:57levels the height at which the temperature goes above zero in the
00:12:01atmosphere of course typically as you go higher up in the atmosphere temperature
00:12:05drops away this is the height at which temperature falls below freezing and
00:12:09normally when we're talking about a height of 200 meters or less anything
00:12:14that's falling out the atmosphere will be coming down as snow so if it's one
00:12:20degree at the surface but the freezing level is at 100 meters you can normally
00:12:25expect snow when it gets to about 400 meters these blues or 600 meters it
00:12:30typically will depend more on your elevation and also on the intensity of
00:12:35the precipitation so very heavy rainfall will cool the atmosphere enough so that
00:12:41you can talk about snow when the freezing levels at 400 meters or if
00:12:46you're in the south it's interesting because you know you've got places like
00:12:50Dartmoor Salisbury Plain that are a bit higher so a slightly higher freezing
00:12:56level will make the difference and looking at Wednesday 6 a.m. and what you
00:13:02can see is as that low moves in it's got this significant gradient between
00:13:07freezing level of 400 to 600 meters across the south and above 1400 meters
00:13:15so anything here across Brittany will be rain and then moving forward you can see
00:13:21that how marginal it becomes and how much of a knife-edge situation it is
00:13:27because of that sharp gradient there and actually the air is cooling through the
00:13:32afternoon so that's across parts of the southwest into South Wales the freezing
00:13:37level is falling in this scenario but of course it will depend on the track of
00:13:41the low and the intensity of the precipitation and just to put the
00:13:46precipitation on here and you can see that where you've got the very cold air
00:13:51it's just on the edge of where the precipitation is is moving in so that
00:13:57basically whoo press too many buttons that basically highlights how much of a
00:14:04knife-edge it is very mild air there to the south slicing the channel in two and
00:14:10then air that's cold enough perhaps certainly over hills for some falling
00:14:16snow so just to reemphasize most likely we'll get some patchy sleet and snow at
00:14:24lower levels in places along the south of England with the more significant few
00:14:30centimeters of snow over the likes of Dartmoor and perhaps some other higher
00:14:34parts Exmoor if it goes that far north perhaps Swanswee Plain and so on but
00:14:39there's a lower chance that would result in much more significant impacts where
00:14:43you get widely 2 to 5 centimeters of snow and more locally 10 centimeters
00:14:49over higher parts so that's something that we're keeping a close eye on and
00:14:54while I'm on this graphic actually it's a slightly different graphic it's this
00:14:58one just want to show the setup for coming into the weekend after Wednesday's
00:15:06low slides away and we're going to Thursday and Friday we continue to see
00:15:10some snow showers across northern parts of the UK and you can see how northern
00:15:13parts of the UK are generally 200 meter freezing level or lower but in the south
00:15:20by Friday this is Friday dawn here's Plymouth Cornwall Southwest Ireland the
00:15:27next lows moving in and that again brings milder Atlantic weather with it
00:15:32and on the boundary we've got once more another band of precipitation and you
00:15:39can see it's quite heavy actually for Cornwall for western parts of Devon
00:15:42but the heavier rain there is also tied in with the much milder air and it's on
00:15:51the leading edge where we could see some snow across parts the West Country
00:15:56perhaps into the Midlands South Wales just fast forward this and what you can
00:16:02see there midday the milder air is winning it's moving in by this stage
00:16:07it's into Dorset Somerset Pembrokeshire but on the leading edge there this is
00:16:12where there may well be another transient but patchy spell of snow but
00:16:18what looks the case at the moment is that way the where the air is cold
00:16:23enough the precipitation is lighter and more patchy the heavier precip mostly
00:16:29affects Cornwall during the morning and then it kind of fizzles away and in fact
00:16:33playing that forward this is through Friday afternoon much of this fizzles
00:16:39away however this is the latest model run from the Met Office six hours earlier
00:16:44the previous model run had a bit more snow across parts of the Midlands into
00:16:49parts of Wales and the West Country so certainly not a done deal yet it's
00:16:55looking like a trend towards less snow but something that we need to keep a
00:17:00close eye on finally before we get on to the exciting stuff in the medium range
00:17:07just to look at the progression as we go into the weekend and beyond now
00:17:14we've got that weather front moving in as I say some heavy rain for Cornwall
00:17:18early Friday a patchy rain that tends to fizzle away as it runs into high
00:17:23pressure there and moves north but at the same time we've got milder air
00:17:30coming in now through the weekend that milder air tends to take over quite
00:17:36quickly in the west and the southwest but it takes much longer further north
00:17:40and east and just looking at the snow depth for example this is the start of
00:17:49Saturday and what you can see is that there's a little bit of a patch there
00:17:56over Dartmoor likely to be disappearing patches over parts of Wales more
00:18:01significant snow over North Wales, Northern England into parts of Scotland
00:18:05so we're likely to continue to see lying snow if not some fording snow
00:18:09across the north and east and fast forwarding that you can see how it
00:18:15eventually kind of melts away at lower levels but we keep some hill snow going
00:18:19and the temperature rise across the northeast of the UK the east and the
00:18:25northeast is likely to be relatively slow which if you compare Plymouth for
00:18:33example here's the temperature change over the next few days quite chilly 5
00:18:38degrees there for Wednesday 0 degrees in the morning up to 8 degrees overnight by
00:18:43the start of next week 10 degrees by day Norwich meanwhile it's a slower
00:18:47temperature trend as you can see struggling to get above the mid single
00:18:52figures by day into the start of the weekend likewise Aberdeen the temperature
00:18:56trend is an upward temperature trend there's this same trend everywhere but
00:19:01it just takes a bit longer for the likes of Edinburgh for example as well
00:19:05it's not until Monday that we're back into the double figures in Edinburgh on
00:19:10Monday and it will be it will be mild by the time we get to Monday because when
00:19:16you look at the bigger picture we've got this warm front moving in we've got
00:19:19these mild southwesterly winds moving in and we've got this high pressure which
00:19:24is close to the south and southeast of the country and that high pressure is
00:19:29crucial for determining next week's weather so what do we need to look at to
00:19:38decide next week's weather well one thing we could look at is this computer
00:19:43model simulation from the European model this is the number 12 weather simulation
00:19:48from the European model the UK is here I know it'd be difficult to see on the
00:19:53screen but the UK is here the red and blue lines are basically cold fronts and
00:19:57warm fronts in this visualization and then we've got the high pressure I
00:20:02mentioned that is close to the southeast of the UK at the end of that sequence
00:20:07for Sunday and Monday and what this is showing is high pressure over Denmark
00:20:14into the the Baltic and it's got these southeasterly winds for UK although for
00:20:22Scotland it's southwesterly so it'd be mild but generally settled with some
00:20:26weak fronts over the country so that's one computer model simulation for next
00:20:30Wednesday however another simulation for next Wednesday has that high pressure
00:20:34centered over the UK mostly dry perhaps some overnight frosts and just some more
00:20:40breezy weather there in the far north of Scotland milder here another simulation
00:20:46from the European model shows the same high but over Iberia over western France
00:20:52and more of a mild west of the airflow over the UK so when we're looking at
00:20:57these simulations and we don't just run the model simulation once when we're
00:21:03looking at the forecast we run them multiple times in fact the European
00:21:09model is run 51 times and this is midday next Wednesday and it's showing all 51
00:21:16outcomes the reason we run the models multiple times this is explained in
00:21:22depth by Alex Deacon in a recent video that we can provide a link to on
00:21:27ensemble forecasting and in a nutshell it's because the forecast is subject to
00:21:34chaos theory so the idea that a subtle change at the start of a forecast will
00:21:38make a big difference in a week's time or two weeks time and so on you might
00:21:42have heard of butterflies flapping their wings and setting off tornadoes in
00:21:46Texas that's essentially the idea that you can tweak the forecast very slightly
00:21:51at the start and things can rapidly become very different by the time we get
00:21:56to a week's time and that's often what we see we look at these multiple
00:22:00simulations and it's too much information sometimes 51 different
00:22:05projections for where that high pressure is next Wednesday now one way of making
00:22:10sense of that is to count them up there's one there that has high pressure
00:22:13over the UK there's another one here to the south another one there to the south
00:22:18so that's two to the south one over the UK I could do that for all 51 simulations
00:22:24but then we've got to bear in mind that it's not just the European model the
00:22:26Met Office has an ensemble producing supercomputer as well with 34 or 36 I
00:22:34should say members coming out of the Mogreps simulations and 40 coming out of
00:22:41Glossy and then the Americans also run multiple simulations and it all gets a
00:22:47bit too much so rather than counting these up by eye well Rob Neal here at
00:22:53the Met Office came up with a much better solution and he's here now to
00:22:58talk about it and how we can use Rob's tool for helping people determine the
00:23:04medium-range forecast absolutely welcome to the deep dive Rob thank you
00:23:08very much great to be here now we we actually go quite far back you and I
00:23:13we had a very similar start to our career didn't we we did because what
00:23:18what was your degree so I did a geography undergraduate degree at the
00:23:22University of Birmingham and then following straight on from that did a
00:23:25one-year masters in meteorology also at Birmingham yeah and we we just crossed
00:23:31over didn't we because I did the same master's degree a year earlier and then
00:23:37we both started the Met Office at roughly the same time but doing very
00:23:42different things yeah so I started in 2008 straight on the graduate scheme as
00:23:47a scientist and I joined into one of the many post-processing teams that we have
00:23:52here and in those teams we essentially take the raw forecast output from the
00:23:58Met Office supercomputer plus forecast output from other models around the
00:24:01world and create forecast products from it to be used by operational
00:24:06meteorologists and the public really or anyone who wants to get hold of these
00:24:10products so that they can interpret the forecast and make decisions from that
00:24:15yeah and we're getting more and more of information aren't we from supercomputers
00:24:20which is great we've got more and more stuff to use but it's just overwhelming
00:24:25the amount of information you could look at so your job basically is to make
00:24:30that information much more available and much more accessible it is and in fact
00:24:36when I joined the office in 2008 the the Met Office ensemble system MoGreps had
00:24:42only just become operational the year before I think it's fair to say back
00:24:46then ensembles was still a fairly new thing but obviously since then we now
00:24:51have access to hundreds of forecast simulations every day not just from
00:24:55MoGreps but from ECMWF from NSEP in the US and we've got short-range models and
00:25:01long-range models and essentially it's information overload for the forecasters
00:25:05because as you say it's go for all these different simulations by hand and try
00:25:10and work out probabilities manually would just take you forever and ever and
00:25:13ever so actually post-processing is a great way to turn this into something
00:25:18useful for the end-user. So talk us through the first thing you did to make
00:25:23this information more useful for people. Yeah so the first thing we did was to
00:25:28create a set of 30 climatological weather pattern definitions which you
00:25:33see here primarily to use these in the medium to extended range forecasting
00:25:39context so we wanted to produce these to be used primarily as a forecasting tool
00:25:43so that we can understand the probability of these different weather
00:25:48patterns occurring out into the future and the premise behind that was that we
00:25:53know that the skill in predicting specific weather parameters such as
00:25:57temperature and particularly rainfall is quite low beyond a few days in advance
00:26:02but we know that predicting a large-scale pressure pattern like what
00:26:06you see here is much more skillful one to two weeks in advance maybe even
00:26:10further in advance than that so if we can forecast the probability of one of
00:26:15these weather patterns occurring we can then make sensible decisions as to what
00:26:19the weather conditions are going to be at the ground for example weather
00:26:23pattern 30 down here that's a very cyclonic westerly type very stormy very
00:26:28deep area of low pressure so we would intuitively expect that to be a very
00:26:36windy pattern we would expect it to be warmer than average as you've got that
00:26:40flow off the Atlantic and we'd expect it to be wet particularly across western
00:26:45parts of the country so without knowing anything about without knowing anything
00:26:49about the weather the rainfall or the temperature direct from the model all
00:26:53we're doing here is taking the pressure field from the model turning that into a
00:26:57weather pattern forecast and then we can make sensible assumptions as to what the
00:27:02weather conditions and maybe the impacts are going to be at surface I mean we
00:27:06talk about these kinds of weather patterns a lot we were doing it before
00:27:11you did this analysis we talk about Scandinavian high so high pressure of a
00:27:16Scandinavia a lot of people particularly meteorologists and weather
00:27:19enthusiasts know that high pressure over Scandinavia in the winter can bring cold
00:27:23easterly winds equally in the summer you might enjoy an Azores high extension so
00:27:30that's where you get the semi-permanent high pressure of the Azores extending
00:27:34towards the UK to bring some pleasant weather in the summer so we talk about
00:27:37these but I guess you formalized them didn't you how did you do that well to
00:27:42produce the weather patterns we applied a clustering machine learning technique
00:27:46where we took a hundred and fifty four years of gridded sea level pressure
00:27:50observations across the UK and surrounding area which is like about
00:27:5550,000 days over that hundred and fifty four year period and then applied the
00:28:00clustering algorithm to it and what it basically did was group days together
00:28:04which had similar synoptic set up so like it groups all the westerly days
00:28:08together it groups all the the days of a cyclonic northerly it groups the
00:28:12rigid Azores high days together and so we ended up with these 30 types and it
00:28:17was partly an objective method and partly subjective so the objective part
00:28:22was just running this clustering algorithm where we tell it what domain
00:28:26to do the clustering over we tell it how many weather patterns we want it to
00:28:29produce and we hit run and then it produces our set of 30 weather patterns
00:28:32the subjective part was then to show these weather patterns to the
00:28:36operational meteorologists at the Met Office and get their feedback on them
00:28:40because we went through quite a few iterations to produce these weather
00:28:43patterns so for example the first set of weather patterns were very westerly
00:28:48dominated we had very few northerly or easterly types in there and we
00:28:53discovered this was because our clustering domain was too far over the
00:28:55North Atlantic so we restricted the domain over the North Atlantic and
00:29:00extended it a bit further east and a bit further north and we were then able to
00:29:03capture more of those so more rarer but high-impact weather patterns so we
00:29:08didn't want to just get there sort of the climatological types in there we
00:29:11wanted to get some of the more rare extreme types in there so that this tool
00:29:16is more useful for forecasting high-impact weather yeah and what I find
00:29:21interesting about these is that they they change through the year don't they
00:29:24so you've got not necessarily 30 weather types it's 30 times 12 because you've
00:29:29got 30 you've got slightly different flavors for each of these weather
00:29:32patterns for each month yes that's right yeah for example if you you can pick up
00:29:37weather pattern well 30 yeah we'll see that that is a very windy type so this
00:29:44is the the actual climatology the actual temperature and the wind for January so
00:29:50this time of year we can see it's a very windy we've got a very deep area of low
00:29:53pressure there but then as we move through the year we find the position of
00:29:57the low changes the direction of the wind changes slightly it might become
00:30:01slightly less windy the low might become a bit weaker as we go through
00:30:04into the summer months but then as we go back into the autumn and winter again it
00:30:08becomes deeper and more stormy and there's 30 weather patterns here and the
00:30:14numbered from 1 to 30 so the lower numbered patterns occur most often
00:30:18annually and the higher numbered patterns occur least often annually right
00:30:21so and also the high numbered patterns tend to be our more extreme high-impact
00:30:26so this weather pattern weather pattern number one is a relatively frequent one
00:30:32you've got high pressure over the Azores there you've got a bit of a ridge of
00:30:36high pressure over the UK and west to northwesterly winds that's how it looks
00:30:43in January skipping forward to July and it's the high moves slightly you've got
00:30:50a bit more of a northwesterly low pressure over Scandinavia that says the
00:30:53subtle changes there's September subtle changes to the year both in terms of the
00:30:57position of the high the wind direction they're all roughly the same aren't they
00:31:01for this weather pattern for each yeah when weather pattern one is a fairly
00:31:05benign weather pattern if we saw that forecast with high confidence we
00:31:09wouldn't be particularly worried in terms of high-impact weather and you
00:31:12know weather warnings and that sort of thing yeah but if we saw a high
00:31:17confidence forecast for one of the high numbered patterns particularly 30 or 29
00:31:20any of those really there they all have sort of deep lows coming in stormy
00:31:25conditions and these are rare these aren't forecast very often so we'll be
00:31:29thinking in terms of named storms if we saw this being forecast with with high
00:31:33confidence but it's also important to note that we have restricted this down
00:31:38to just 30 weather patterns so we won't always get a perfect match to the
00:31:43conditions in reality so for example is this weather pattern this is 29 this is
00:31:48weather pattern 29 if we had a high forecast confidence for weather pattern
00:31:5229 and we showed this climatology map for weather pattern 29 we're not saying
00:31:58this is the exact position of the low it could be a bit further north a bit
00:32:01further south east or west this is just the closest match that the NWP has to a
00:32:07climatological pattern at this stage but things could change but I
00:32:11think for a forecast two three weeks in advance that's absolutely fine yeah
00:32:16we're just trying to get a steer for what's going to happen in the future it
00:32:19gives us that feel of is it going to be stormy is it going to be dry is it going
00:32:23to be frosty is it going to be warm in terms of you know talking about warm
00:32:29cold the precipitation as well well I should mention that we do have
00:32:34precipitation data we just haven't got it to work yet it's under development a
00:32:37lot of this is still under development I just couldn't wait to show people
00:32:40because it's been a long time coming so precipitation data is coming for now
00:32:44it's temperature it's wind and it's pressure and in terms of the
00:32:48temperatures and the precipitation where did you get that data because you
00:32:52analyzed 154 years of weather patterns or daily weather scenarios to get the
00:32:59the pressure pattern where did you get the temperature from that's right yeah
00:33:03we didn't at the time we produce these weather patterns which was back in 2014
00:33:07sort of time we didn't have access to 150 odd years of sort of high resolution
00:33:12weather data in terms of temperature rainfall snowfall wind and so forth but
00:33:18we did have access to the to ECMWF's global reanalysis which is a really
00:33:23thorough sort of view of weather conditions over the whole globe you know
00:33:28on a daily basis over at least 50 years so we were able to get hold of that data
00:33:33and then calculate a climatology for each of the weather patterns when we
00:33:38produced the weather patterns the clustering algorithm produced two things
00:33:41it produced the weather pattern definition so those 30 maps that we saw
00:33:47at the beginning and it also produced a daily historical weather pattern
00:33:51classification data set which tells us what the weather pattern is or what the
00:33:56weather pattern was should I say on each day from the 1st of January 1850 up
00:34:00until well essentially near present so we've got this really long record of
00:34:04what the weather patterns were on each day and then from that we can take our
00:34:09global reanalysis from ECMWF and which is also on a daily interval and
00:34:14essentially merge those two data sets together to form a climatology for each
00:34:20of our 30 weather patterns and this is what we're seeing here this isn't
00:34:23actually forecast information this is what we're viewing here is the
00:34:27climatology for that particular weather pattern and that's how it's really
00:34:31clever stuff that's how you know that if I draw on the map here now this is
00:34:35weather pattern 17 so we've got high pressure there over Sweden we've got a
00:34:41bit of easterly wind it's quite a cold weather pattern for the UK and these are
00:34:45temperatures that typically at midday that you've seen come out of the
00:34:48reanalysis isn't it? That's right. For the locations I'm clicking on there
00:34:52you know you could go all the way to Europe for example to Norway but that's
00:34:58how you know these are the kinds of temperatures from this kind of weather
00:35:00pattern that you typically see at midday in January but also for the same weather
00:35:06pattern for July you get those kinds of weather patterns. Yeah that's a really
00:35:12interesting pattern because in the winter it's much much colder than
00:35:15average but in the summer it can be really hot so yeah it just shows that
00:35:19how one synoptic pattern can change in terms of the weather conditions
00:35:22depending on the time of year. And this this shows the temperature anomaly so I
00:35:26was just on the temperature forecast for that weather pattern this shows the
00:35:29temperature anomaly so just again putting a few numbers on the map there
00:35:33most places below average two and a half degrees below average there for
00:35:36southern parts of the UK but fast-forward to July and it's much
00:35:41warmer than average particularly in the West because of course you've got that
00:35:44cooling easterly breeze coming in from the North Sea so really a really useful
00:35:50way of summarizing the UK's weather you know what does high pressure to the east
00:35:55of the UK mean in July what does it mean in January that's what you've
00:35:59determined. Yeah it's great and essentially we're getting all this
00:36:02weather information from climatology data rather than actual forecast data
00:36:07from the NWP model the only NWP coming out that we're using is the actual
00:36:12pressure forecast from all the different ensemble member simulations.
00:36:16Now final technical point before we get on to what the actual forecast is
00:36:21showing and that's that we've got eight regimes and 30 weather patterns can you
00:36:28talk through that? Yeah so initially we applied this clustering algorithm to
00:36:32produce the set of 30 weather patterns and then we felt it would be nice to try
00:36:38and condense those down further perhaps into into a set of eight which we have
00:36:41here and that was produced by essentially running some sort of
00:36:45clustering again to group to cluster those 30 into a smaller set of eight
00:36:49here. So what we've got here basically is say pattern 2 has low pressure to the
00:36:56north high pressure of the Azores and a westerly airflow of the UK and that's
00:37:00the common theme for patterns 4, 8, 20, 23, 26, 30 that's kind of an unsettled weather
00:37:09pattern westerlies could be quite stormy meanwhile pattern 8 there that only has
00:37:15weather pattern 10 and I'm guessing that's like an extension of the Azores
00:37:20high into the south and so what was it 10 number 10 so that's that's that kind
00:37:27of more benign weather pattern but the the other one which was pattern regime
00:37:332 you've got low pressure to the north all of these are showing similar sorts
00:37:40of themes with a west to southwesterly ever quite unsettled weather. And the set of
00:37:44eight become really useful when we start looking at it looking at these
00:37:47patterns in terms of the forecasting context particularly when we have a lot
00:37:50of ensemble spread which will come on to in a minute. In terms of the forecast so
00:37:55we start with all of these weather simulations from the European model from
00:37:59the Met Office models, GFS the American model, how many in total is the
00:38:05computer looking at? I think at the last count we have 259 ensemble
00:38:10members or forecast simulations to look at so that's a lot if you were to
00:38:14look through all those 259 individually as a forecaster you'd be there all day
00:38:19looking at them so this is where we need post-processing to sort of try and
00:38:23condense the forecast down for you in a way to save you time. So it's basically
00:38:28taking each of these and saying that looks like weather pattern 6 that looks
00:38:32like weather pattern 8 counting them up that's right and then you end up with a
00:38:36probability for each day don't you? Yeah the each ensemble member is assigned to a
00:38:41weather pattern based on the ensemble member and weather pattern pairing with
00:38:44the smallest overall distance as a like a domain average over Europe so that's
00:38:49repeated for every single ensemble member and then the probability of each
00:38:53weather pattern is based on the number of ensemble members assigned to each
00:38:57weather pattern and that's essentially what this graphic is showing here but
00:39:01here this is based on the set of eight weather regimes so the probabilities are
00:39:06initially based on the set of 30 weather patterns and then those
00:39:10probabilities are aggregated up according to the set of eights which we
00:39:14can see here. So let's let's go through this because this is a really useful
00:39:18tool for actually forecasting when I give you the the actual forecast for
00:39:23next week and the week after in just a moment what's this showing is it's the
00:39:27next two weeks the dates on the bottom there although we can go out to four or
00:39:31five weeks or so we've got bigger bigger charts that will will show off in the
00:39:36future and this is showing that probability of those different weather
00:39:40patterns isn't it for each of those days and the higher the probability the
00:39:44bigger the bar essentially. That's right yeah and it's also color-coded so we've
00:39:49got general sort of a set of blue weather regimes and a set of sort of
00:39:53orangey yellow colored weather regimes so very broadly speaking the blue
00:39:58coloring is is trying to draw your eye towards weather regimes that could be
00:40:02sort of more wet and more unsettled whereas the red colors try and draw your
00:40:07eye towards weather regimes that are more likely to be settled so high
00:40:10pressure dominating you know less rain and wind around you know very broadly
00:40:14speaking but there will be a lot of variability within there but it just
00:40:18helps particularly when you've got a lot of spread across the set of 30 weather
00:40:23patterns within the forecast then it helps to draw out a forecast signal
00:40:28where there might be quite a bit of uncertainty across all the members. So
00:40:31the most basic terms here from Friday from the weekend it's mostly reds
00:40:37oranges yellows and that would suggest that higher pressure is close to the UK
00:40:42it's a bit more settled compared with the next couple of days where low
00:40:46pressure is closer to the UK so that's the first thing to point out the weather
00:40:52is becoming a bit more settled because you've got the bigger bars here are
00:40:56associated with those red orange colors. Absolutely yeah I think it's a
00:41:01this is a great visualization for doing like a drill down approach you can start
00:41:05by looking at the bigger picture just like the difference between the blues
00:41:07and the reds and we can clearly see here as a transition from blue to red but
00:41:12then but then you can start drilling down to look at like sort of different
00:41:15flavors of reds into next week where there's quite a few different scenarios
00:41:20being shown and you can then start looking at the set of 30 as well if you
00:41:24want to get more information. Well we'll be talking about next Wednesday so this
00:41:27is Wednesday here and it's clear to see that the biggest bar is this color so
00:41:31what we could do is we can click on that bar and bring up what that looks like in
00:41:37terms of the weather pattern so this is a weather pattern in which high pressure
00:41:42is centered over the UK but it's mostly across southern parts of the UK and this
00:41:47shows the temperature anomaly so just to give you a reminder of the key that's
00:41:51the key so it looks like parts of northern Scotland are more than three
00:41:56degrees above average but southern UK is closer to average in fact. Will that
00:42:01work? Yeah we can click on the map and confirm that with the with the pen.
00:42:06Meanwhile there's also it's also apparent that there's this other weather
00:42:12pattern here that's a darker red that has a slightly different position of the
00:42:17high it's it's more towards Western Ireland and as a result more of the
00:42:22country is below average because of course at this time of year when you've
00:42:25got high pressure centered you've got the clear skies you've got the frost
00:42:28you've got temperatures kind of decreasing day by day so that's the
00:42:33second most likely outcome for next Wednesday isn't it? Yeah. And you know
00:42:38clicking on any of these you can see that they're all variations on a theme
00:42:45they're all dependent on exactly where that high is and that's where these
00:42:48weather patterns come in handy because they show you okay this scenario where
00:42:52you've got high pressure to the east it's a small probability but could
00:42:56happen high pressure to the east colder than average whilst the most likely
00:43:01they're high pressure to the south warmer than average across northern
00:43:05parts of the UK. Now as well as the different colors here which correspond
00:43:09to these types these regimes high pressure centered for example for this
00:43:14color they're also subdivided but these little black lines not sure if you can
00:43:19see them so that's just different I can click on the right place it's just a
00:43:24different flavor of that high pressure centered in this case it's a little bit
00:43:27further south so you've got first of all the reds oranges versus the blues then
00:43:34within the blues you've got four different types here and four different
00:43:37types of reds and oranges four different areas of high pressure location and then
00:43:41within each of these you've got the subdivisions in terms of those 30
00:43:44weather patterns so high pressure centered there are two coming out here
00:43:48there's there's this one which is much more likely across the south one that's
00:43:52less likely much further south. Yeah that's right it just it provides a
00:43:56really nice drill down approach starting with the reds and the blues and then
00:44:00drilling down to the set of eight and then drilling down further to the set of
00:44:03thirty depending on what level of detail you want to go to and it also depends on
00:44:07how uncertain or certain the forecast is as well. Now if you're not bamboozled or
00:44:14perhaps amazed enough let's just take a look at something that is particularly
00:44:23complicated stroke exciting I don't know depends on your point of view we'll go
00:44:28to next Wednesday again because we've been talking a lot about next Wednesday
00:44:31although the forecast for the next two three weeks is roughly the same coming
00:44:36out you know when you look at the the content that's coming out of Decider it's
00:44:40that high pressure that's close to the south of the UK most likely isn't it so
00:44:45what you're seeing here next Wednesday it's got the most likely weather pattern
00:44:51showing up on the map here okay that's high pressure to the south meanwhile
00:44:56what you've got on this graph is the different circles represent the
00:45:04likelihood of each of these patterns so again we've got the high pressure
00:45:10centered pattern that's a big circle because it's most likely and it's color
00:45:16matches the description that high pressure centered there are less likely
00:45:19weather patterns that are coming up which are smaller and there are some
00:45:23little dots that you might not be able to make out that are possible but much
00:45:27less likely so the bigger the circle here the more likely it is on this
00:45:31particular day which is Wednesday the 15th of January and its placement here
00:45:36depends on where I'm clicking on the map so I'm just going to hover over London
00:45:41there just to start off with and what that what this is showing is that its
00:45:46position on the graph shows higher than average pressure so average pressure is
00:45:51the line down the middle there right in the middle would be higher than normal
00:45:55pressure go left you've got lower than normal pressure go right you've got
00:45:59higher than normal pressure it's got higher than normal pressure for London
00:46:03for this circle for this high pressure centered pattern but all the other dots
00:46:08are also higher than normal pressure so whatever you're looking at here for
00:46:14London for next Wednesday all these weather patterns are the most likely
00:46:19weather patterns are generally showing higher than normal pressure meanwhile
00:46:22we've got warmer as you go further up and colder as you go further down
00:46:26average temperatures would be in the middle and that means that in this box
00:46:31you've got where you've got most of the weather patterns including the most
00:46:36likely it's higher pressure but colder which is what we're just talking about
00:46:40isn't it yeah higher pressure centered you've got colder than average
00:46:44temperatures you've got overnight frost perhaps yeah nothing extreme but that's
00:46:49what you'd mostly expect in the south but what's really useful with this is
00:46:53that you can move that dot and go further north for example into northern
00:47:00Scotland and what you can see is the most likely weather pattern moves into
00:47:04the warm category still higher than average pressure but it moves into the
00:47:08warm category a few of these other circles jostle around as well and just
00:47:13goes to show that different weather patterns influence the weather in
00:47:18different ways in the UK depending where you are so northern Scotland this kind
00:47:22of set up you've got the southwesterlies around the top of the
00:47:25high it's mild perhaps wet as well in the south it's more settled but it's
00:47:30drier so settled colder drier higher pressure but warmer up here yeah I think
00:47:38it's great and I think one weather pattern can be representative of a range
00:47:42of weather conditions as you say across the country and on this scatterplot here
00:47:46we could have potentially up to 30 dots one for each of the 30 weather patterns
00:47:51but clearly in this situation for this particular day next week only maybe 10
00:47:57to 15 of the weather patterns are actually being forecast out of the 30 so
00:48:01that's why I've got that number of dots there but the color coding helps draw
00:48:05your eye towards whether we're high pressure dominating or low pressure
00:48:08dominating next week and we can see here the yellows and reds there with the
00:48:13higher probabilities and then these little dots here these tend to be blue
00:48:17in color so it's a quite strong signal towards higher pressure next week and
00:48:22you know as you say for for the south of the UK colder than average as well using
00:48:26those climatologists from the weather patterns yeah I hope you're keeping up
00:48:29at home but just to summarize what this is showing because it's really
00:48:32impressive it follows on from Rob's work since 2012 and then the further
00:48:37visualization work that we've been doing in our team at the Met Office and also
00:48:42with our partners presentation cartography in New Zealand what this is
00:48:46showing is a summary of the expected weather for Wednesday the 15th of
00:48:50January from 259 weather simulations so simulations of the atmosphere from 259
00:48:56259 individual maps and what that's basically saying is that next Wednesday
00:49:02is a bit colder but higher pressure in the south average to warmer in the north
00:49:08and it it just summarizes so much information in this scatterplot it's
00:49:14going to be a really useful tool and you know a lot of this stuff is under
00:49:18development we can refine it further let us know how useful is for you and and if
00:49:23you find it confusing or if you find it find that you can understand it would be
00:49:29really really good to know that basically summarizes the current forecast
00:49:34although another very very quick and easy way of showing high versus low is
00:49:39this isn't it is yeah this is a really nice graphic to illustrate transitions
00:49:43from cyclonic to anticyclonic conditions so the blue shading here
00:49:48indicates that weather patterns with low pressure are more likely and then red
00:49:52shading indicates that weather patterns were sort of with high pressure are more
00:49:55likely and you can see here a clear transition between low and high pressure
00:49:59between later this week and this is calculated essentially by having two
00:50:04regimes so we take those 30 regimes and we aggregate them up to just two a set
00:50:09of we I'm adding up the probabilities for all the low pressure probabilities
00:50:14and then adding up all the probabilities for the set of high pressure weather
00:50:18patterns and then you can see the probabilities of the two showing they're
00:50:22quite neatly and not just for the most recent run which is the top row but also
00:50:26for previous runs that gives you an idea of the trend in the model output as well
00:50:30doesn't it yeah yeah and this is based on the UK so if I produce that plot for
00:50:34different countries around Europe then it would look different because obviously
00:50:36one weather pattern might be low pressure dominating over the UK but it
00:50:39could be potentially high pressure dominating over like Spain or further
00:50:43south the final thing to look at well three final things to look at there's
00:50:49there's this which is the confidence isn't it that it shows whether the
00:50:56current forecast is more or less confident than average that's right yeah
00:51:00it's using this forecast confidence index which ranges between 0 and 1 and
00:51:06it shows the forecast confidence over time so the solid line is the forecast
00:51:10confidence index for the latest forecast cycle so a score of 1 would mean that
00:51:16all of our 259 ensemble members are assigned to exactly the same weather
00:51:20pattern so a hundred percent confident of a particular weather pattern
00:51:23occurring a score of 0 would mean that all our forecast members are equally
00:51:28distributed across the set of 30 weather patterns so in other words we
00:51:31haven't got a clue what's gonna happen and you and then what we also have on
00:51:35here is like a forecast confidence climatology based on historical forecast
00:51:41runs so our climatology here as it says in the title is actually based on the
00:51:44last 13,000 four cycles blimey so we've got this quite so it's constantly being
00:51:49updated it's meant to be on a five-year rolling cycle yeah so it should always
00:51:53be relevant to the latest sort of model configurations so we've also got the
00:51:58forecast confidence climatology plotted on here we've got the average which is
00:52:01shown by the dashed line so when our forecast confidence confidence index for
00:52:06the latest run is above average we can see this green shading here if it's
00:52:10below average we get red shading and then we've got the 10th and 90th
00:52:14percentiles and also the historical maximum and minimum so you can see on
00:52:20occasions when we're really uncertain at some point in the past four or five
00:52:25years there's been a lot of spread even in the first couple of days yeah I
00:52:29suspect that was like one or two cases where there's a lot of
00:52:32uncertainty going on. How does this tool perform? It's summed up
00:52:41here isn't it? Yeah so that just going back to that forecast confidence index
00:52:45that's not a measure of forecast skill it just tells you how much ensemble
00:52:49spread there is because there could be very little ensemble spread but that
00:52:53doesn't mean it's a skillful forecast whereas this plot this is a verification
00:52:57plot for weather regime 2 and this is our most skillful weather regime to
00:53:04forecast out of the set of eight weather regimes that we've got. It looks like
00:53:07this doesn't it? Here it is low pressure to the north high pressure over the
00:53:11Azores, westerlies for the UK basically. Yes that's right yeah yeah so it's a
00:53:16skill score using the Bryce skill score which is a common probabilistic
00:53:22verification skill score so as you'd expect skill score sort of drops off
00:53:26over time we're going out to 15 days here on the x-axis and on the y-axis
00:53:32we've got the forecast skill so when you when you get a skill of zero that means
00:53:38our forecast is no more skillful than just using climatology right so anything
00:53:43above zero is good. So first point to make is that yeah forecast skill drops
00:53:49over time as you'd expect and it gets close to zero by the end of week two but
00:53:55this is using a strict verification measure where for example if weather
00:53:59regime 2 is forecast with a high probability in say 12 days time but it
00:54:05actually occurs in 13 days time instead I would argue that's still a very good
00:54:08forecast still a useful forecast but using the the method here that would
00:54:13mark that as a false alarm or miss right so it's quite a strict verification
00:54:17method so I suspect the skill we're getting from this tool we've been
00:54:20showing today is better than what's being illustrated in this graph but it
00:54:24still gives us a good feel. The other point to make is we've got lots of lines
00:54:28on here and these are for all the different ensemble models that we're
00:54:32using within this forecasting tool so we've got MoGrexG at the shown by the
00:54:36black line that's this one here and that only goes out to seven days we've then
00:54:41got the ECMWF medium range ensemble shown by the blue line here which goes
00:54:45out to 15 days and then we've got other models here like the USGEFS ensemble
00:54:51with its 31 members and the ECMWF extended range ensemble with 101
00:54:56members plus the Met Office glossy global seasonal forecast system but we
00:55:01also have these extra lines here called a multi-model version of it where we're
00:55:07taking all these ensemble members from all these different global centres and
00:55:10just sort of putting them into one pot essentially and then using those to
00:55:14calculate the probabilities. So we've got two different versions of the
00:55:17multi-model one where we've given each ensemble equal weights so ECMWF even
00:55:23though it has more members than MoGrex they both those two models both
00:55:27have the same weighting and then we tried it using by weighting the models
00:55:31according to their ensemble size so in that case ECMWF would get a much bigger
00:55:35weighting. So that's why we've got these two lines at the top the sort of the
00:55:38yellow and the greeny line but we found that they they're over plotting over the
00:55:42top of each other essentially so there's not much difference in terms of
00:55:45the weightings of the models doesn't really matter but the most important
00:55:48thing to say is that the multi-model performs at least as well as if not
00:55:53better than the best performing individual model so this is evidence
00:55:57that actually taking a multi-model a multi-ensemble approach is a very
00:56:01sensible way of looking at the forecast particularly in the medium range to get
00:56:05a steer of what's around the corner essentially. Yeah so taking this new
00:56:10tool of yours 259 weather simulations summarising them and matching them to
00:56:15weather patterns that the UK normally experiences is a great way of a
00:56:21summarising all the output from all those models, being I think the most
00:56:26accurate representation of the future forecast and c finding out exactly what
00:56:30that means for different locations in parts of the UK. Yeah and the number of
00:56:34ensemble members we have is bound to increase in the future and but that
00:56:39won't affect how we use this tool because we can increase it to maybe a
00:56:42thousand members in the future and it won't change how we use it and how we
00:56:45interpret it because that's all hidden behind the scenes. Before talking about
00:56:50the future you mentioned that back since 1854 you've got an idea of the weather
00:56:57pattern for every single day up until the near past and you've come up with
00:57:03some interesting trends in terms of the occurrence of these weather patterns
00:57:07over time. That's right yeah we've got 30 plots on here, one for each of the 30
00:57:11weather patterns and they're showing the frequency occurrence of each of the
00:57:15weather patterns over the last 150 or so years. So we can see like with weather
00:57:20pattern one there's been a bit of variability from one decade to the next
00:57:24and then over the more recent years it's been sort of tailing off in
00:57:27frequency and you could say the same for patterns two, three and four. Perhaps when
00:57:32you look at some of these patterns in the middle it's more sort of variability
00:57:36from one decade to another but it's sort of on a cycle. So we've got these lower
00:57:41numbered patterns which there's potentially a trend showing up here now
00:57:45in the more recent sort of 30 years or so where they're reducing in
00:57:49frequency and these are more sort of benign sort of day-to-day weather
00:57:52patterns and that's happening because some of the higher numbered weather
00:57:57patterns are potentially increasing in frequency over the most recent years.
00:58:01For example weather pattern 23, maybe even 22, maybe even 30 are the most sort of
00:58:08stormy weather pattern. So it's just interesting to look back over time and
00:58:12we can see there are cycles but there's potentially also trends going on as well
00:58:17and James Pope, one of our climate scientists, has done some work recently
00:58:21where he's assigned the UK Ensemble out to 2100 to these 30 weather patterns and
00:58:30he's been able to see how the frequency of these 30 weather patterns may change
00:58:34going 50, 100 years into the future and that's really interesting work.
00:58:38It is interesting because you know we talk about climate change, how it affects the UK,
00:58:42there's the average warming and the average increase in rainfall and so on
00:58:47but then there's also this idea that these particular weather patterns are
00:58:52changing. We've seen it, some decadal variation but also some longer term
00:58:58trends. We've seen it happen, how much will it happen in the next hundred years?
00:59:02That's what James Pope has been looking at and actually in a future deep dive,
00:59:06I've not asked him yet but hopefully he'll say yes, we'll get James on here to
00:59:10explain more about what he's found about how these weather patterns are changing
00:59:15with climate change and how they're likely to change in the next hundred
00:59:18years or so but I'm sure that's more than enough from the two of us for now.
00:59:23I don't know if we've gone over Alex Deakin's record for the longest deep
00:59:25dive, hopefully, but yeah it was just a topic that we really had to cover in
00:59:31great detail because it's fascinating to me at least but it's really useful as
00:59:36well, the work that Rob's been doing over the last 10 to 12 years and now we can
00:59:41finally visualise it and use it more and more in our forecast here at the Met
00:59:45Office. So thank you for sticking with this and I hope you find it useful. Let
00:59:48us know in the comments what you think and we'll be back with the 10-day trend
00:59:53tomorrow, that will be Alex Deakin and then the live on YouTube of course that
00:59:58we do every Friday at 12.15 so join us for that. But thank you Rob for
01:00:03explaining in great detail your amazing work on all of this. Yeah well thanks for
01:00:08inviting me on, it's been great to be able to explain it to a wider audience
01:00:10and I look forward to seeing it used more in the future. Yeah, thank you very
01:00:14much and bye bye for now.