LRF or attempting them

[bluearmy]

http://www.cpc.ncep.noaa.gov/products/pred...y/814day.03.gif

a tad more promising than for a while JH. although we still have no positive anomoly over us

a modicom of synergy with late July output (or am i being rather optimistic?)

[johnholmes]

well the July one had actual closed 500mb low and high circulations which is not the case so far, but we can live in hope!

[johnholmes]

No 25 is available below; again apologies for it being a link but there are quite a lot of charts in and I'm pushed for time; apologies to those with slow downloads, any problems please pm me and I'll send an abbreviated version.

25_Update_on_Tuesday_19_August_to_Net_Wx.doc

for those who either have no broadband or simply don't want to read the full lrf here is the summary

For the period 31 August-7 September my punt would be for:-

Mixed signals with the odds just about in favour of not much settled weather for anything more than a day or so, in brief. However the signals, to me, do suggest that there could be a relatively shortish (3-4 days) going into early September with high pressure being more dominant than low pressure. As to where that high pressure might locate I would plump for to the north or north east. This would give some areas above average temperatures by day but probably coolish nights and little in the way of rain. Confidence in this is barely more than 50% though.

Not an inspiring summary but I prefer to be honest about things rather than pretending. Perhaps there will be clearer signals by the time of the next update, Tuesday 2 September. I’ll try and remember to validate this on 31 August and 7 September.

Edited August 20, 2008 by johnholmes

[johnholmes]

No 26 for Tuesday 2 September for 7-14 September

The averages across the UK for this period are:-

The Northern Isles about 13C with Scotland and N Ireland 17C, England and Wales 19C and London about 20C

Looking at the various inputs which may affect the forecast summary

ENSO

No change from their last comment

NAO

The Sunday output is on the left and the Monday/Tuesday issue is on the right

As can be seen, whilst they differ in some aspects, they both, as have all of them over the past week or so, show a marked fall from the current level and show great uncertainty by the period end, although most tend to support the idea of it being –ve.

Hurricane situation

NOAA have yet (26/08) to make any mention of Gustav so any forecast for the period shown has no input about the possible impact it may have. Bertha caused a change in pattern; the last one did not, fizzling out in the southern states. This one?

NOAA comments

Over several days they have suggested that a ridge is likely to be a major factor over the eastern fringes of the States. Looking at both their and ECMWF 500 anomaly 500mb charts this seems to be what the two models show in most of their runs over the past few days. (See comments re hurricane situation as of Tuesday 27 August). Their scores for the models does seem a bit odd when looking at the differences between the major models they use. They rate them 4/5 down to 3/5 by day 14.

NOAA 500mb anomaly charts

Below are the last two 500mb anomaly charts from them, from 31/08-4/09 out to 2/09-8/09 on the right

Both show, and have for several days, +ve areas over the Greenland area almost up to the pole. The largish –ve area west of the UK on the first chart is just about there, a little further south on the second chart. A feature of comparing the two is that the anomalies are almost always smaller on the later chart.

ECMWF and GFS 500mb anomaly charts

This latest chart from this morning takes us to 5 September.

The charts over the past few days do show a progressive nature with now no sign of the largish +ve area, on either chart, to the west and north west of the UK that they had shown several days ago. Each shows the +ve area over the eastern States mentioned by NOAA, neither have any 500mb indication of the latest Tropical storm Gustav, apart from a suggestion of a trough to the east of the States which MIGHT be it. It seems unlikely though, so none of the extended upper air charts has yet, apparently, taken this into account? They do both show +ve anomalies though to the north.

Basic ECMWF and GFS trend for H+240 period (around 5-6 September), with GFS out to T+384(11 September)

The overall pattern around 240 shows, as is often the case, differences in where the main 500mb trough will be and its depth and orientation. GFS appears to be in favour of a 500mb trough being to the east of the UK by the start of the period whilst ECMWF mostly favours the main trough being to our west. So little continuity with the two models there. GFS, looking at the 3 500mb anomaly charts, seems a bit too quick so possibly the ECMWF may be nearer the mark. As to what happens further out then its very unclear at the moment. By the end of its run GFS shows another 500mb trough taking up position just west of the UK. Most charts through the period have surface high pressure at times to the north west of the UK, along with the usual Azores area and also over Europe at times. Low pressure being largely in the Atlantic and moving on its normal track up into the Norwegian Sea.

Overall Summary for the period

At the moment I would favour some kind of blocking to the north or north west. This would favour temperatures around or, at times, below normal. Rather unsettled, with at times, as a ridge from the Azores high tries to build in, a fairly typical north west-south east split. More unsettled and rather windy perhaps at times in the north west. Less unsettled for the south east although even here it will not be a completely settled spell other than maybe 2-3 days at some stage in the period.

Verification is due 7 and 14 September (will be done 11 September if I remember as I'm away from 13th)

No further updates until either Tuesday 23rd or Wednesday 24 September. This will then be well inside the autumn period.

[johnholmes]

I'm not doing my lrf outputs until I come back off holiday late September.

However, looking at outputs from the various sources I use does seem to suggest that a fairly major block may develop to our north east.

NAO supports this, and that forecast is pretty dependent on what is happening around the Tropics. The various centres that output northern hemisphere 500mb charts or their anomaly charts also suggest something along those lines.

Turning to the oft looked at GFS and ECMWF 500mb-msl chart at T+240, they both show surface HP in about that position with a surface flow east of south.

What we don't know is just how much effect the various Tropical Storms/Hurricanes may have on this. They do create a huge surge of energy from the Tropics into the areas to the north and can totally upset any forecast, so with that caveat its best to wait and see. But there is, I would say, an increasing possibility of the block I mentioned earlier around the middle of the month

Edited September 4, 2008 by johnholmes

[The PIT]

I'm not doing my lrf outputs until I come back off holiday late September.

However, looking at outputs from the various sources I use does seem to suggest that a fairly major block may develop to our north east.

NAO supports this, and that forecast is pretty dependent on what is happening around the Tropics. The various centres that output northern hemisphere 500mb charts or their anomaly charts also suggest something along those lines.

Turning to the oft looked at GFS and ECMWF 500mb-msl chart at T+240, they both show surface HP in about that position with a surface flow east of south.

What we don't know is just how much effect the various Tropical Storms/Hurricanes may have on this. They do create a huge surge of energy from the Tropics into the areas to the north and can totally upset any forecast, so with that caveat its best to wait and see. But there is, I would say, an increasing possibility of the block I mentioned earlier around the middle of the month

I suspect as long they don't track this way they won't have much direct influence more a subtle one. Bertha threw up a nice ridge and gave us a bit of summer.

[TEITS]

I'm not doing my lrf outputs until I come back off holiday late September.

However, looking at outputs from the various sources I use does seem to suggest that a fairly major block may develop to our north east.

NAO supports this, and that forecast is pretty dependent on what is happening around the Tropics. The various centres that output northern hemisphere 500mb charts or their anomaly charts also suggest something along those lines.

Turning to the oft looked at GFS and ECMWF 500mb-msl chart at T+240, they both show surface HP in about that position with a surface flow east of south.

Forgive me for asking John but im confused about your comment "NAO supports this".

I have always read and learn't that the NAO isn't a driver of the weather and is merely a measure of the pressure systems between the Azores and Iceland. So if the ECM/GFS are indicating HP over Iceland region, LP over Azores then the NAO forecast is likely to be negative. If on the other hand the reverse is true then the NAO is likely to be forecast as positive.

I know forecasters use the SSTS to forecast the NAO when prodcucing a LRF but again this isn't really forecasting the NAO. They are using the SSTS to see where LP, HP systems are likely to be which of course would then alter the phase of the NAO.

Sorry if I have misread your post but it appears to suggest the NAO is the reason for the major block to our NE when its actually the other way round. I might have got this completely wrong and I apologise if I have.

[johnholmes]

Not misread perhaps I did not make it clear enough, NAO is ONE of the factors I look at as does anyone else attempting any type of lrf. It is as you say a function of pressure differences. The question of course is not that but how they come to be there. The NAO shows what is there, or predicted to be there by the GFS/ECMWF/Met O models but it is partly(NAO) as a result of what is happening world wide that the NAO goes -ve or +ve.

Not well explained as I'm rushing off to catch a train. If its not correct when I look again during tomorrow I'll attempt a better explanation.

Just as a comment

items like QBO, ENSO and other items which you see on posts from GP are all part of the 'puzzle' of trying to get some idea of how the larger scale troughs and ridges in the upper atmosphere may be created.

Edited September 4, 2008 by johnholmes

[TEITS]

Thanks for replying John.

I must admit I try and understand some of GPs post but unfortunately do not have the intelligence to understand some of his posts. I have tried gain further knowledge using the internet but much of it doesn't seem to sink in. The climate is full of puzzles which are difficult to place together and I don't even understand some of the pieces!

[johnholmes]

I'll come back with fuller and I hope better explanation of some of the things later today

[johnholmes]

I am in the process of putting my idea of what is involved with doing an lrf, largely from our perspective as amateurs but with links into the professional centres. If GP would like to add to it when I post it, correcting any mistakes I've made(!) I would be delighted. After further rewriting it can then be put in our Guides section for easy and ready reference for all of us.

and briefly again about NAO; what we are looking for is the ability to predict whether it will be -ve or +ve up to several weeks in advance. To be able to do this we need to know what is happening on a far wider scale than we normally see in the routine 4x daily GFS outputs.

Predicting whether it will be -ve or +ve is just as fraught as day to day forecasting for the UK. Take a look at this link, book mark it, and save it, say 2-3 times per week. Then have a look after

http://www.cpc.ncep.noaa.gov/products/prec...k/pna/nao.shtml

a couple of weeks to see how much it can change, just as the models we look at do.

Edited September 5, 2008 by johnholmes

[Lettucing Gutted]

I suspect as long they don't track this way they won't have much direct influence more a subtle one. Bertha threw up a nice ridge and gave us a bit of summer.

Do hurricanes "throw up" upper level ridges due to latent heat release in the upper troposphere?

Edited September 5, 2008 by Craig Evans

[johnholmes]

at times yes, but difficult to be precise about just what happens, each is different depending on a mass of variables.

But simple answer yes

[johnholmes]

as promised, an attempt to put many of the terms used in lrf work together, links to a whole variety of first class sites.

first class material with the links for many tea and coffee breaks.

Please, those with some understanding of the subject, probably higher than mine, do add your bit. As I've posted at the end I hope eventually, incorporating those corrections/additions to my input, will allow me to put the whole amended post into our Guides section.

Fairly large file so give it time to download if not on broadband.

Long_Range_Forecasting_by_we_amateurs_explain_links_etc.doc

Edited September 6, 2008 by johnholmes

[acbrixton]

John, many thanks for this.

regards

ACB

[johnholmes]

I'll restart these tomorrow. Bear in mind though I have no data from any source for the first one whilst I've been away so only charts around today and tomorrow will be available for any 'stab in the dark'.

Looking at the post above for an idea of the lrf into September whilst away and the block seems to have been there much of the time. Mind you that is just a quick look at how things are so those that follow these things closely may wish to put me right about that.

Anyway into autumn and I'll do a weekly lrf update for 2-3 weeks in advance, usually each Wednesday.

[Iceberg]

John a quick question, the PNA is probably more positive than it's been all year. With largely Neutral NAO what kind of effect might this have ?.

Cheers

[johnholmes]

finally found time to put charts and text together

here with the restart of lrf's from me, usually up to 2 weeks ahead, sometimes a bit more, sometimes a bit less.

Ice, I hope the text will answer your query above?

No 27 for 3 October to 10 October(? beyond?)

The average afternoon temperatures for the country are shown below

11C in the Northern Isles to 14+C over Scotland and Northern Ireland, with 16-17C over England and Wales and about 18C for London.

As posted above there is not a lot of data available to me as I've been away but here we go with the first one for autumn.

ENSO

I copy part of the latest text on 22 September

The most recent weekly SST anomaly in the NINO3.4 region is -0.01C, indicating continuing neutral conditions in the tropical Pacific. How long will conditions remain neutral, and what are the probabilities for developing an El Niño or re-developing a La Niña later in this year? The low-level easterly wind anomalies in the central and western Pacific that were reinforcing the shallow thermocline anomalies reappeared in August. Their main impact was to erode the warm anomalies that had been sitting in the eastern equatorial Pacific. Those easterly wind anomalies were likely associated with an MJO event, as they have subsequently disappeared. Most of the ocean-atmosphere indicators suggest that ENSO-neutral conditions are likely to continue not only during the Sep-Oct-Nov season in progress, but through most of the remainder of 2008 and early 2009.

Currently, the models see the development of El Niño conditions more likely than the re-emergence of La Niña conditions, although the model probabilities for El Nino reach only about 30% through most of the forecast period. The sub-set of the models that do see the possibility of developing El Niño conditions are primarily the dynamical models, rather than the statistical models. All things considered, there is an approximately 5% likelihood for either La Niña or El Nino conditions to develop during the Sep-Oct-Nov season, and a 90% probability for ENSO-neutral conditions.

NAO

It shows a rather up and down idea of what the pressure field across the North Atlantic may be. An overall trend of falling is pretty evident. That said, as has been pointed out above the PNA(Pacific North American Pattern) is more positive than for a long time. I have to admit to having little idea just how much the two can be seen to correlate. Obviously they do, in that what happens in the Pacific in turn affects what later happens in the Atlantic. A point I often make when someone tries to dissociate one factor from another. Just for comparison I've put the 11September output alongside.

I do this to re emphasise my point of some time ago that ENSO, QBO, PNA or NAO are NOT necessarily correct and thus the task of trying to decide which bit of data to accept and which to discard is very much a balancing act. Again to repeat I'm a newcomer to this so its all a learning area for me.

NOAA comments

They are reasonably happy with their neck of the woods out to day 14, although they have reservations regarding Tropical storm development, commenting that it appears favourable through the period.

Their 500mb chart in that period is shown below

The 500mb anomaly charts for the period 1-8 October are indicated below. This does show something of a pattern change, both from its 30 Sept- 4 October and even more so from the last two I have seen. The +ve area is much smaller near the UK and does not link to any other +ve area, as it did previously. Then it was a large +ve area Greenland/Iceland into the UK.

That from ECWMF and GFS for a similar period is below; this appears to be suggesting –ve values (on the ECMWF version) for the UK area with GFS of a similar idea.

If we look at the two main models for T+240=4 October(12z today), ECMWF on the left and GFS on the right

Both have high pressure near/just south to south east of the country with low pressure in the Iceland/Greenland regions. Obviously one cannot expect the two to be identical but there is some similarity, or I think so. Turning to the GFS at 384=10 October, then it shows the first major autumn storm, if it was to be correct that is?

my apologies for some being printed over a previous charts - saving time again!

So having looked at what data is around for me what do I suggest for the period of the forecast, 3-10 October?

SUMMARY

I would suggest that surface high pressure will be near the south of the country with lows running from fairly well south in the Atlantic up between Iceland and Scotland. Thus a north west-south east split seems the most likely. Breezy and with rain at times but never cold for the north west. For areas further south then more settled for much of the time, possibly a touch too much breeze for any fog problems during any clear overnight spells. Mostly dry, especially for the south east and with the flow from a point south of west, then fairly mild/rather warm. Ground, let alone air frost, seems unlikely other than possibly in any temporary ridge behind a low pressure system for a few Scottish Glens.

By the end of the period, into the second week of October then it looks much less definite. I take the point from NOAA about Tropical storm developments and looking at what the equatorial area seems to be suggesting. Overall I would suspect becoming more unsettled for all areas with a possibility, no more than that, of a fairly major low (from Tropical beginnings) approaching from the south west.

This tends to go in the face of what the NAO seems to be suggesting with a fairly –ve spell being predicted for late on in the period. Once again conflicting advice and currently I plump for what I have above. Another week may show me either right or wrong.

The next update is due on Wednesday 1 October for the period 10-17 October. I will try to remember to do a check on the current one on 3 and 10 October.

Edited September 25, 2008 by johnholmes

[johnholmes]

No 28 issued on Wednesday 1 October for 10-17 October 2008

Afternoon temperatures for the country for this period show 10C for the Northern Isles, 13-14C for Northern Ireland and Scotland, about 15C for England and Wales and nearer 16C for London.

Looking at the data seen over the past few days shows

ENSO and other tropical items

There is no update from the one I posted in the last lrf issued 22 September. see below

Currently, the models see the development of El Niño conditions more likely than the re-emergence of La Niña conditions, although the model probabilities for El Nino reach only about 30% through most of the forecast period. The sub-set of the models that do see the possibility of developing El Niño conditions are primarily the dynamical models, rather than the statistical models. All things considered, there is an approximately 5% likelihood for either La Niña or El Nino conditions to develop during the Sep-Oct-Nov season, and a 90% probability for ENSO-neutral conditions.

NAO

Below is the latest; it shows more signs of a recovery towards, or even above, the zero line than previous outputs although still mainly below zero.

NOAA comments

Over the past few days they are not very confident with the outputs into their middle term, 10-14 days ahead, rating the models overall at 1-2/5.

NOAA 500mb charts

1. 500mb for this period; this shows little sign of anything but an overall westerly flow.

2. 500mb anomaly charts

1 October

for the period into mid October this shows a –ve area just west of the UK and +ve over Iberia/southern Europe and also over eastern USA. This has moved slowly over the past day or so, being further west a couple of days ago, so seems fairly reasonable.

ECMWF-GFS 500mb anomaly chart for the same period; showing, over the past few days the growing –ve area over the UK/west and also parts of north west Europe, along with the +ve area for the USA.

12z today GFS and ECMWF for T+240(10 October), with both showing a similar upper air pattern but with surface variation.

12z today GFS for 384 hours (17 October)

Summing this up

Most data, actual and forecast shows an unsettled theme for the first part at least with a marked 500mb trough over or mainly west of the UK. Beyond that and there is little guidance. I would suggest that the upper air wavelength of the main troughs and ridges is probably slowly mobile. This would allow the upper high and surface high over eastern USA to migrate slowly into the Atlantic. However the NAO is not totally supportive of this idea.

Summary of the general weather pattern for the period shown

I would think we will remain in an unsettled period with a marked 500mb trough over/ just west of the UK. Some models suggest this major trough to link right across the pole into Asia. A spell of surface lows, generally tracking close to or over the country. Obviously rather windy at times, especially for western and northern areas. Bands of rain and showers, again more of this for the areas just mentioned. The south east should be only bothered now and then by this but there will be rain or showers at times and breezy conditions also. Whilst winds blow from a southerly point then above average temperatures. At this time of year as the lows track east, then, depending on the time of day, any temporary ridge that develops is likely to give ground frost and possibly slight air frost as far south as the Midlands, possibly further south. In any temporary northerly turn to the winds then snow or sleet over the higher hill tops, especially for Scotland.

Later in the period there are even more mixed signals. However, on balance I would suggest that as the wave length of the main upper troughs/ridges becomes slightly progressive, then a spell of more 'mellow mists and fruitfulness' type of weather could develop. It could also eventually turn out warmer rather than colder as the high may track across the country and into northern Europe allowing a flow from the south not the north for a time.

However, it's a long way off even for this thread so best we just check back on the 10th and again on the 17th to see how it has panned out.

Next update No 29 will be on Wednesday 12 October for 17-24 October.

GFS 240 is in different format hence black appearance

Edited October 1, 2008 by johnholmes

[pottyprof]

Interesting read John.. Thanks again!!

[johnholmes]

my pleasure P, its good fun if a bit of a slog, be interesting in the 'cold' season, whenever that is, how it shows, so far its not bad, I would reckon on a 60-70% success rate since I started in late January.

[pottyprof]

my pleasure P, its good fun if a bit of a slog, be interesting in the 'cold' season, whenever that is, how it shows, so far its not bad, I would reckon on a 60-70% success rate since I started in late January.

Id say you've been pretty much there with the ideas even if at times the models have been unwilling to play ball mate..

I think for any LRF, 60% is very good at the lower end.. Its great to see how gut feeling and charts interact.. Just been reading GPs latest post and another super set of thinking there..

[johnholmes]

No 29 is issued below in the link

once again time constraints mean you have to download the file to see what is written - sorry about that

It covers the period 17-24 October

No_29_issue_on_8_October_for__17_as_file_on_nw.doc

[Rob]

more rain! great, NOT

[johnholmes]

No 30 lrf on 15 October for the period 24-31 October

Going through the various items used in this attempt to predict the likely weather pattern for the period shown.

Average temperatures for late October are shown below

They range from about 9C for the N Isles to 10-11 for N Ireland and Scotland, 11-13 for England and Wales (north-south), and 14 for the London area.

ENSO update is the same as the last one (18 September) when the following brief excerpt showed this:-

The majority of models also remain ENSO-neutral through 2008, while a couple models develop weak El Niño conditions. For the OND 2008 season, only 3 of 23 models (13%) predict El Niño conditions to develop, and none predict La Niña development. OND=October, November, December.

There are indications, no more to me, that events over the Tropics and even the Stratosphere may be showing rather different signals to recently. In the time scale of this lrf neither will have any real impact, if they are correct, on this forecast summary.

NAO

The two below show how the forecast has changed during the last week or so. +ve for much of the time with some indication that by the month end it MAY be tending –ve.

ECMWF and GFS 500mb anomaly charts over the past week. Both are fairly consistent in their version and, as is usually the case, they are a little different in some aspects but not by all that much. Both suggest, as can be seen below from this morning, a –ve/trough type feature off the eastern seaboard, and also across the polar region, with some extension towards the UK. GFS suggests more of a w of north flow than ECMWF by 25 October.

The NOAA version, with their anomaly charts, have developed more of a trough off the western seaboard over the past few days (now valid at 22-28 October), along with the one off the eastern seaboard; this with –ve showing over and north to north east of the UK along with +ve to the south west.

NOAA forecast comments over the past few days show a rather mixed view on the models, and not very high confidence in them, at least with the detail for their area of interest. The major troughs they seem fairly happy with.

GFS and ECMWF outputs over the past week in the T+240 scale and for GFS out to 384 hours?

They are both, again, struggling to some extent. Both, along with UK Met, in the nearer time frame expect at least one deepish low to track not that far from the north west of the UK. GFS has certainly moved away from its idea of the Euro high setting up shop in any major way. Nor are there any real signs of either suggesting a major surface high to our north, especially north west. Given the NAO 'signal' this should not surprise.

Summary

I would think that a fairly 'normal' late October is in prospect. That is unsettled for all areas with the south east again having less wind and rain than the north west. No sign of any major Euro high giving any longish settled spell for the south and east. At least one of the various lows coming out of the eastern seaboard during the period seem likely to be almost into the relatively major storm category for the far north and north west but not elsewhere. Rainfall totals look like being about average for most away from western/northern coasts and hills. Not much in the way of wintry type weather seems likely with only brief and temporary colder air for, mostly, the northern half. Nights of frost in the south look fairly rare and even mist and fog, thanks to the lows, probably below average for late October.

Right at the end of the lrf I have a suspicion from some of the above data that there MAY be, note the term(!), a northerly outbreak, but I would rate it as no higher than a 30% chance as of this evening.

The next update will be on 22 October No 31 and will cover the period 1-7 November.

Verification of todays lrf is due on 24 and 31 October.