[Storm and Convective Discussion 10th June Onwards]

Met Office messed up proper with this one.

 

 

I criticised the MetO for their convective forecasting the other day, and today's yellow zone is but further evidence of their weakness in this area of forecasting. 

 

Not to dwell too much...but it's interesting you both have this view..because from my point of view it was extremely well forecast, considering the complexity of the set-up and multiple trigger mechanisms. However, unfortunately it seems this info was not disseminated to the public in a way that reflects this...

[Storm and Convective Discussion 10th June Onwards]

No offence Mark but what, if anything does that mean in plain simple English ?

Apologies, UKPP = UK post processed, a suite we have here at the Met Office, which means our model output from the Unified Model has additional algorithms and corrections applied to the raw data to better fit reality. At the shortest lead times (the next 6 hours) Nowcast data will be blended, then UKV (which is the Met Office high resolution model for the UK, with a resolution of 1.5km), up to Euro4 for the longest lead time (the Met Office 4km gridded model covering the Europe domain). So what I was referring to earlier was post processed output from our highest resolution model, regarding the potential for isolated storms in the south east, and possibly more significant convective activity in Lincolnshire tomorrow evening.Post processed data at short lead times (close to a weather event) for site specific forecasting, are generally regarded as the most reasonable prediction of what will occur. Broadscale considerations are still taken into account though, as sometimes UKPP looks too much like reality as opposed to forecast!

[Storm and Convective Discussion 10th June Onwards]

I guess it's a case of wait and see! So the UKPP is showing lower values of instability when compared the NMM-6 charts I've posted? Are you able to post any of the UKPP charts by any chance....Be interesting to see for comparison, what grid resolution is the UKPP running at, 1.5km?

 

As I alluded before it's hard to compare as it outputs dilute CAPE, which is cloud base to cloud top. It's actually scaled to 2km, for rather boring and fiddley reasons. Unfortunately probably cannot post charts, so I won't haha. But interesting to see 3hr ppn output (18z-21z) as well from UKV (have not looked at anything else that's in public domain, maybe other output agrees?) with lincolnshire looking most at risk and isol SE/London by 2100z tomorrow. Considering no dynamic activity tomorrow, these can be assumed convective so further support potential for some storms about!

[Storm and Convective Discussion 10th June Onwards]

Although not as high as the SB instability some moderate ML CAPE on offer around 700-1100 j/kg. I'm taking all of this data from the hi res NMM-6, so it's a a bit one sided as I haven't looked at the other models! But at this short range I tend to stick with the the NMM-6  Quite surprised when I saw the potential this evening! 

Yup, could get some moderate values, though not comparable to the storms earlier in the week though! I'm looking at UKPP, which is post processed data to match location better from UKV at this lead time.

[Model Output Discussion 31/05/14 onwards.]

Not sure if common knowledge or not...but GM (UKMO) breaks down blocks faster than it should as a function of the statistics that output the verification index..so in short, more often than not, it gets it right, as more often than not, HP does not dominate and GM pushes us back into a zonal pattern consistent with climatological mean. (the model looks better, scores higher, we get it right most of the time). So, if GM keeps block for a long time, this is quite interesting as it is going against its internal revert to zonal flow after 3-4 days...so we can gain good confidence from this. I would be confident in 10 day output at this point in time.This is reflected in UK Outlook for Monday 16 Jun 2014 to Wednesday 25 Jun 2014 on MO website.

[Storm and Convective Discussion 10th June Onwards]

Our post processing spits out dilute CAPE, and has values of ~700 max (generally 400-500J/kg) in south, central and isol northern UK later tomorrow afternoon, so not directly comparable, but perhaps not as severe as the graphic above posted by Liam. I'd say though there is a great degree of interest (and has been since Monday) in the destabilising effects the over-running trough will have for profiles in south and central UK tomorrow evening. Maybe even Wales with orog uplift late afternoon...As the upper short-wave trundles through cold advection could erode warm noses from the 15C theta-w plume below, with isol thunderstorms popping up into the late evening. So CBs not ruled out tomorrow!(also mid-level instability can result in Ac castellanus => lightning! though don't get your hopes up too much as detail is not pinned down and won't be until short lead time)

[Model Output Discussion 31/05/14 onwards.]

It looks like most of this warmer weather, if it comes, will be held mostly to the E, with western areas seeing more of a wetter outlook. Past Saturday solutions start to become varied, with ECM and GFS pulling a plume up from the continent. This could bring warm, or even, hot conditions (outside of the climatological mean). Confidence is reasonably high, with MOGREPS on board too. However, depending on what happens on the continent, and what any engagement with this plume might result it, is still uncertain. Whether or not the vortex off the SW of the UK into the weekend retrogresses will have something to say about what happens past Sunday..though currently all models look to be holding it to the west bringing a S'ly flow into Britain and further potential plumes.T+240 signal and beyond still has a very weak anticyclonic bias, so any fine and dry conditions are likely not to be prolonged, with a westerly flow favoured.

[Convective / Storm Discussion - 21st May 2014 onwards]

Hi AJ, if you scroll back through the posts you'll see Nick F has done an analysis as too has BrickFielder.

 

Hi Harry, I noticed you asked me in the previous thread about which model was showing high values of CAPE...today, again, UKV was putting >1500Jkg-1 of "dilute CAPE" (apologies if I failed to mention this before, not quite the same as CAPE, but still these values aren't often seen) across areas of the SE and London that are quite concerning, for tomorrow afternoon. Dependant on surface heating in the morning tomorrow (which should be relatively sunny in that area of the country), some deeply convective t-storms may be triggered with localised accumulations of up to 40mm/24hr for some isolated areas.It's a complex set-up at the moment, what with tonight's frontal zone containing embedded instability moving northwards from France into the SW..MCS not off the cards in SE as high WBPT air advected northwards...then you've got a separate conveyor all together that is beginning to interact with the first feature. Behind this will be massively unstable air!

[Convective / Storm Discussion - 20th May 2014 onwards]

Thursday looks promising for those after storms...latest models putting CAPE at 1500Jkg-1 plus in some areas. Will be worth watching tomorrow, and seeing what the chief says. Also would not be surprised to see a warning from ESTOFEX. In the meantime, enjoy tomorrow's sun whilst it lasts!

[MetOffice Operational Forecaster Technical Test?]

Also they are keen to stress it's "operational meteorologist" as more emphasis is put on the advisory role; forecaster ≠ meteorologist!

[MetOffice Operational Forecaster Technical Test?]

Hi all,I have a query for anybody who has ever been through the interview process before. How difficult is the assessment? Would you recommend revising any sections in particular? Is it 50.50 physics/maths?And any idea what the group assessment involves? Hints or tips!?Thanks!

Hi Matty,Apologies if this reply is a little late! I imagine by now you have already had your assessment day and exam (as I saw a lot of you around the building a few weeks back), so know exactly how much you needed to know! (or didn't know!) If you are selected for interview please don't hesitate to contact me for some advice on how to prepare for that. :-)

 

All the best,

Mark

[Model Output Discussion 31/03/14 onwards.]

It does look as though the weather will rapidly deteriorate after the weekend, which for most will be the warmest of the year so far owing to a fair amount of warm air advection. At the moment there is a degree of uncertainty around whether or not some pretty heavy isolated storms may break out in the S/SE by Saturday, so it might not be BBQ weather for all!Regarding post Monday though there is still a large disparity between the models, so confidence is low as to what to expect... EC det allows for cyclogenesis in the bay of biscay, which would lead to most of next week being pretty wet and windy. MOGREPS is too spread the the moment to have anything more than low confidence in the solutions (though some do allow for building a high for the bank holiday weekend).

[Model Output Discussion 31/03/14 onwards.]

Thank you knocker and johnholmes for your time and replies  

They have indeed helped. Always a learning curve the weather isnt it. Very fun though!

 

As mentioned already it isn't easy looking at simple synoptic charts of parameters such as pressure or temperature to try and identify fronts. However, if you have access to charts with station circles and begin to plot a few simple variables, such as isobars, isodrosotherms (areas of equal dewpoint - very useful as different air masses will have different humidities, and a front divides airmasses!) and try and spot wind veer...combine that with simple knowledge of clouds and types of weather you'd expect along different fronts and you'd actually not do too badly identifying where they are!

 

However, it's much easier to look at the charts with these frontal features already positioned for you

[Model Output Discussion 31/03/14 onwards.]

It just goes to show how a week away is still almost impossible to predict with any accuracy even in 2014. The GFS tonight is nothing like the charts I posted yesterday, and as SS says above, the GEM has thrown out a completely different run from this morning.

 

Do you think we will ever be able to predict a week out as accurately as we do for 24 hours today within our lifetimes?

 

It could be that more processing power is needed, but i just have this thought in my head that, perhaps the atmosphere and the Earth is just far too complex and fluid to ever be able to predict!

 

 

Well 30 years ago the 24 hour forecast was as accurate now as we can go out to 4 days. So that is some achievement in itself! In the near future though, more computing power would not give us anywhere near the accuracy we have for 24 hours for a week out. A lot of other things need to be improved before that!Incidentally, within the next year or so, we will indeed have more computing power here at the Met Office (as a new supercomputer is soon to be purchased, the most expensive yet!)

[Model Output Discussion - 10th Jan Onwards]

Mods, if you move this post to another thread can you PM and tell me where to so that I can check back for replies please.

 

Ok, genuine questions ...

 

Why do different models offer different solutions? Why do models rarely agree until almost T24? Why do they get so many details wrong the further out they go?

 

The reason I ask is that with modern technology these days it appears (to me anyway) that forecasting (certainly long range in Winter) is no further forward than it was say a decade or so ago.

 

Why do I not see any tangible improvement in forecasting the weather, particularly but not exclusively, during Winter (I must admit I only really have an interest in Winter or extreme weather). How can various models predict varying solutions until almost T24?

 

Is this an investment issue?

 

Furthermore, if FI is almost futile, why do the models go this far out? What is the point? Would it not be better to produce more accurate modelling say to 5/7 days rather than say 10 or more?

 

These questions aren't borne out of frustration at lack of cold/snow, more why do we seem to be no further forward on long range forecasting than I think we were roughly 10 years ago. Or are we further forward and if we are, then why do we see models flip about like a fish in hot water?

 

Thanks.

(purely biased towards MetO)For the Met Office, in terms of skill and verification improvements please consult these links;UK NWP Index - http://www.metoffice.gov.uk/research/weather/numerical-modelling/verification/uk-nwp-indexGlobal NWP Index - http://www.metoffice.gov.uk/research/weather/numerical-modelling/verification/global-nwp-indexForecast accuracy and how this is measured - http://www.metoffice.gov.uk/about-us/who/accuracy/forecasts & also http://www.metoffice.gov.uk/media/pdf/d/h/13_0068_Accuracy_v09.pdf This is not neccesarily the thread to discuss how models reach their solutions and the how the nature of NWP inevitably will result in greater degrees of uncertainty relating to certain synoptic setups. I'm sure there is an area of this forum where those topics can be discussed! (and would be more than happy to be directed there).Relating to the last decade, the Met Office have made huge advances in their forecasting skill (and are rightfully the joint best operational weather service globally!), and are regularly set targets to reach (achieved and tracked by using KPIs and BPMs to retain their world class status).Briefly in terms of investment, you may remember that even during government science budget freezes, MetO still gets a new supercomputer at the cost of tens of million, every few years!

[Model Output Discussion 17th Dec.2013-12z onwards.]

Seeing as we are discussing the chances of snow, here is some supplementary info regarding various parameters and their relationships with the likelihood of snow... (all from the Met Office, for educational purposes)..Probability of snow________________90%____70%___50%___30%___10%Surface temp ( C ) _________________+0.3____+1.2___+1.6___+2.3____+3.9Height of 0C isotherm agl (hPa)______12______25____35_____45______611000-500hPa thickness (gpm) ______5180___5238__5258___5292____5334 ______ (the "528 line" falls around 40%)(mutually exclusive)Based on WHEN precip begins, as mentioned above latent cooling of the atmosphere will lower the freezing level once the snow starts falling/melting.The wet-bulb freezing level is very important as well...   Mainly snow Readily turns to snow   Mainly rain   Snow very rareHeight of theta-w freezing level agl < 1000ft (~300m)               <2000ft (~600m)             >= 2000ft       >= 3000ft  (~1000m) It's also important to watch out for cold surface air undercutting warm air.A general rule of thumb for the UK is that an 850hpa theta-w of 2C or below can allow for snowfall down to penetrate down to all elevations (with a large number of caveats obviously). Heavy instability and persistent precip with an initial wet-bulb temperature level of as high as +3C can allow snow to reach the surface. 

[Cold Spell Next Week Discussion. (18 November) Onwards.]

When looking at upper levels for indications of snow, perhaps the 850hpa theta-w is worth paying more attention to rather than 850hpa temp or 500-1000hpa thickness. A general rule of thumb for the UK is that an 850hpa theta-w of 2C or below can allow for snowfall down to penetrate down to all elevations (with a large number of caveats obviously).Based on the 12Z run of GFS, we are likely to see most favourable theta-w temps on Tuesday evening.

However no significant precip is forecast for this period..maybe some flurries across Wales and the East Coast..

[The Alps Snow thread Season 2013-14]

Tignes was operating with 8 lifts open today, blue bird conditions and 220cm of snow at the top - some great conditions for this early in the season!