12z Model Comparisons Results

[The PIT]

Best thing about it, some on here think T240 is reliable

[Thundery wintry showers]

I should eventually be able to get the regime-specific details for the models, but I'll have to tinker about with the file, because Excel won't let me enter more than a certain number of values into a formula!

1032mb high to the SW with an anticyclonic north-westerly flow around the high.

GFS scores a 7/10- T+144 and T+120 were both very good, but it lost marks due to the dodgy T+96, which had the high accurately placed, but lows somewhat overdeepend around Iceland giving very strong westerlies.

UKMO scores 8/10- T+144 had the high too far northeast, giving slack north-westerlies over Britain, but T+120 and T+96 were both very good.

ECMWF scores 7/10- T+144 and T+120 both had the high too prominent and a bit too far NE, then T+96 was very good.

NOGAPS unavailable.

JMA scores 6/10- high too far north at T+144 and to a lesser extent T+120 giving northerlies, but T+96 was accurate.

GEM scores 7/10- high too far north at T+120 with westerlies mainly just over Scotland but T+96 was very good.

Overall a good set of outputs, a tendency to place the high too far north at T+144.

[Thundery wintry showers]

1 September

1028mb high to the SW, 992mb low north of Scotland, and strong westerlies. Many southern and eastern areas had a sunny day.

GFS scores a 7/10- unavailable at T+144, T+120 had the high a bit too far west, giving more of a NW flow over Britain, then T+96 was very good.

UKMO scores 8/10- T+144 was dodgy with the high slightly too far north giving slacker winds for most, but T+120 and T+96 were both very good.

ECMWF scores 8/10- it was practically the same as the UKMO at all three timeframes.

NOGAPS scores 7/10- high slightly too far west at T+120, but accurate at T+96. I think T+144 will be available for NOGAPS over the next couple of days.

JMA scores only 5/10- T+120 was good, but T+144 and T+96 were both poor, high pressure too far north and east, giving a very anticyclonic outlook.

GEM scores 6/10- T+120 was very good but T+96 positioned the high too far north giving too anticyclonic an outlook.

More 'traditional' scores over the past few days then, with the Euro models just shaving it.

[Thundery wintry showers]

2 September

988mb Scandinavian Low, 1028mb high to the SW, a WNW flow over Britain. Rain spread south, and was followed by much colder, brighter weather.

GFS scores 7/10- it was okay, but at T+144 and T+120, high was too far north, giving more of a N or NW flow than a WNW flow- typical of the GFS's overdoing of northerlies.

UKMO scores 8/10- T+144 overdid the high a bit, T+120 and T+96 were then very accurate.

ECMWF scores 6/10- same problem as GFS but more marked, with a NNW flow over Britain at all three timeframes.

NOGAPS scores 6/10- high pressure way too dominant at T+144, but reasonably accurate at T+120 and T+96.

JMA scores 7/10- T+144 and T+96 were very good but T+120 had low pressure much overdeepened to the NE, giving strong NW winds.

GEM scores 6/10- decent pattern but high pressure too strong and too far north/east at both T+120 and T+96.

Another decent set of outputs.

[Thundery wintry showers]

3 September

A northerly blast with Britain on the western periphery of it, so a cold bright day for most with just isolated showers. GFS and ECMWF unavailable at T+144.

GFS was very poor, scoring only 4/10. The northerly was vastly underdone at both timeframes; at T+120 high pressure was in charge, while at T+96 high pressure was to the south, giving westerlies for all.

UKMO scores 7/10, as it was poor at T+144 (high pressure in charge) but T+120 was reasonable, northerly only slightly underdone, and T+96 was spot on.

ECMWF scores 7/10 also as it had Britain under north-westerlies at T+120, but not far off being accurate, then T+96 was spot on.

NOGAPS scores 7/10 as well- superb at T+144, but poor T+120 (high over the south) then good T+96.

JMA scores 6/10- similar to NOGAPS, but with the northerly overdone at T+144, giving strong northerlies.

GEM scores 6/10- poor T+120, similar to GFS with high pressure over the south, but spot on at T+96.

Overall, it seems that most of the models actually underdid the northerly.

[Thundery wintry showers]

4 September

High pressure at 1028mb covering southern Britain, with westerlies over the north.

GFS scores 6/10- low pressure over Scandinavia at T+144 giving north-westerlies, T+120 quite good but had the high too far south giving westerlies for all, then T+96 if anything had the high slightly too far north, covering most of the country. Not terrible, but not good.

UKMO scores 8/10- high pressure too dominant at T+144, too far north, but T+120 and T+96 were both excellent.

ECMWF scores 7/10- T+144 was the same as for UKMO with a high over the whole country, T+120 was excellent, but T+96 had the high slightly too far north again.

NOGAPS scores 8/10- T+144 unavailable, T+120 had high slightly overdone, T+96 was excellent, but both had the high accurately placed.

JMA scores only 3/10 as it must have overdone the northerly. High pressure was out to the west at all three timeframes with a chilly northerly flow over eastern Britain.

GEM scores 5/10- not as bad as GEM but high pressure was too far west at T+120 and T+96 with northerlies clipping the east coast of England.

Overall, a very good output from UKMO, and also NOGAPS although T+144 was not available for NOGAPS. The only poor runs were from JMA and GEM, although GFS was rather shaky.

JACKONE has kindly turned the Model Comparisons.xls file into a more user-friendly format so I should get it updated sooner than I would otherwise have.

[Thundery wintry showers]

5 September

1032mb high over SW England, with some warm sunny weather in places.

GFS scores a 7/10- T+144 was dodgy with the high too far east and south, and low pressure affecting northern Scotland but T+120 and T+96 were both very good.

UKMO scores 9/10- it was pretty much spot on at all three timeframes.

ECMWF scores 8/10- high overdone at T+144 but not as bad as the GFS T+144, then T+120 and T+96 were very good.

NOGAPS scores 8/10- high accurately placed at T+120 and T+96, but a bit overdone at T+120.

JMA scores 7/10- dodgy T+144 with high centred over the Continent, but T+120 and T+96 accurate.

GEM scores a 9/10 because T+120 and T+96 were both excellent.

I suppose that when high pressure is in charge there is more margin for error, but in any case, the models did very well.

[The PIT]

Well with hints of a possible change to more mobile conditions it'll be interesting to see if the Models start to struggle again.

[Thundery wintry showers]

6 September

1032mb high centred over western Britain with anticyclonic conditions for all.

GFS scores 8/10- reasonably accurate at T+144 and T+120 but with high slightly elongated, but T+96 was spot on.

UKMO scores only 6/10 as the high was too far south at all three timeframes- especially T+120 and T+96- with westerlies covering most of Britain, and probably a cloudier outlook as a result.

ECMWF scores 7/10- high too far south at T+144 and T+120, much as UKMO, but T+96 was accurate.

NOGAPS scores 8/10 as the high was reasonably accurately placed at both T+120 and T+96, but slightly too far south at T+120.

JMA scores 6/10- T+144 had a strong Scandinavian High, though it at least had high pressure over Britain, then T+120 had low pressure over northern Scotland, giving strong westerlies, then T+96 was very accurate. Overall not impressive.

GEM scores 6/10 also- high too far south at T+120 and T+96 giving westerlies.

Overall a tendency for the models to position the high a little too far south. GFS was the only model not to do this.

[Thundery wintry showers]

7 September

1032mb high over south-western Britain, and an anticyclonic/westerly flow, northerly in eastern Britain. It was a dry day, but not everywhere was sunny.

GFS scores 7/10- T+144 had high pressure in charge but low pressure to the far north underdone, T+120 and T+96 had the high too far south, and too far north, respectively, but not changing the general outlook.

UKMO scores 9/10- very slight quibbles with high positioning but generally speaking all three timeframes were spot on.

ECMWF scores 8/10- T+144 was spot on but T+120 and T+96 both had the high slightly out of place.

NOGAPS scores 7/10- high out of place at T+120 and T+96 but otherwise good.

JMA scores 7/10- T+144 had high too far west giving northerlies for the east, but T+120 and T+96 both very accurate.

GEM scores 8/10- slight high positioning quibble at T+120, accurate T+96.

Overall the models did very well.

[Thundery wintry showers]

8 September

1032mb high to the west of Ireland with slack anticyclonic/northerly flow, although it was warm everywhere. Dry weather was guaranteed; sunshine was abundant in some places but not others.

GFS scores 7/10 again- T+144 had the high too far west giving a pronounced northerly flow in eastern areas, T+120 had the high marginally too far east, and T+96 was perfect.

UKMO scores 8/10- high slightly too far east at T+144, covering the entire country, T+120 and T+96 were then very accurate.

ECMWF also scores 8/10- high slightly too far east at T+144 and T+120, accurate at T+96.

NOGAPS scores 8/10- high accurately placed at all three timeframes, but extending a bit too far north at T+120.

JMA scores 7/10- high too far south at T+144 with westerlies over Scotland, T+120 and T+96 were good with high marginally too far east.

GEM scores 8/10- high too far east at T+120, but accurate at T+96.

Again, a good set of outputs.

[Thundery wintry showers]

9 September

1028mb high off south-western Britain, slack northerlies over the south, strong WNW winds over the north.

GFS scores 7/10- a decent T+144 but with high pressure slightly too far south giving WNW winds for all, and high pressure too far north at both T+120 and T+96 giving too anticyclonic an outlook, but not bad.

UKMO scores 8/10- high alignment issues at T+144 and T+120 but generally as good as GFS, T+96 was spot on.

ECMWF scores 7/10- a decent T+144 with low pressure to our N slightly too far west, HP over the south, westerlies in the north, high was too far north at T+120, T+96 was very good, but overall not quite as good as UKMO.

NOGAPS scores 7/10 also- T+144 had high pressure too far north, covering Britain- a worse output than any of the big three, but T+120 and T+96 were both very accurate.

JMA scores 6/10- high pressure way too far north at both T+144 and T+120, accurate at T+96.

GEM scores 7/10- T+120 and T+96 were pretty good but both had high pressure too far out to the west with low pressure dominant over Scotland.

Overall another reasonable set of outputs.

[johnholmes]

interesting Ian, in as much, that the UK Met is nicely in the lead this month.

Its been a month so far dominated by HP to the west or sw of the UK.

Is that something you have spotted it seems better than the other two with before?

Edited September 10, 2007 by johnholmes

[Thundery wintry showers]

No, I actually recall the GFS doing best with this setup in the past!

[johnholmes]

so much for a pattern emerging!

[Thundery wintry showers]

10 September

1028mb high off SW England, NNW winds covering the north and east.

GFS scores 7/10- T+144 was poor with high pressure covering southern England, but T+120 and T+96 were reasonably accurate, slightly underdoing the northerly flow.

UKMO scores 7/10- same as GFS, T+144 had high pressure over the south but T+120 and T+96 were accurate.

ECMWF scores only 5/10 because although T+144 was reasonably accurate, the northerly flow was somewhat overdone at T+120 and T+96, with low pressure in the North Sea and strong northerlies.

NOGAPS gets no mark as it was unavailable except at T+120, when the NNW flow was underdone.

JMA scores 5/10- high pressure over the south at T+144, and NNW winds covering Britain at T+120 and T+96, but the northerlies somewhat overdone, much as in ECMWF.

GEM scores 8/10- high pressure slightly too close at T+96, otherwise spot on at both T+120 and T+96.

So surprisingly the GEM did particularly well. ECMWF and JMA both overdid the northerlies and as a result didn't score very well.

[Thundery wintry showers]

11 September

1028mb high off south-western Britain, 992mb low east of Iceland, an anticyclonic/WNW flow in the north, northerly in the southeast.

GFS scores 8/10- a good output, with low pressure too far east at T+144, covering the south, but otherwise good, and very accurate at T+120 and T+96.

UKMO also scores 8/10- high marginally too far east at T+120, accurate at T+144 and T+96, a good output again.

ECMWF scores 7/10, as it was accurate at T+120 and T+96 but overdid the northerly at T+144, giving strong northerly winds over eastern England.

NOGAPS scores only 6/10- high too far south at T+144, T+120 unavailable, then high too far north and east at T+96.

JMA scores 5/10- high too far north at all three timeframes with chilly north-easterly winds over south-eastern Britain at both T+144 and T+120 in particular; a dodgy output.

GEM scores 7/10- high too far east at T+120, but accurate at T+96.

Overall, problems with NOGAPS and JMA, with GFS and UKMO the most accurate of the models. ECM seems to have overdone most of the northerlies so far.

[Thundery wintry showers]

12 September

1028mb high over England and Wales, light winds, light westerlies over Scotland. Dry with variable cloud cover.

GFS scores another 8/10- high underdone at T+144 slightly, but otherwise very accurate, then spot on at T+120 and T+96.

UKMO scores 7/10 as the T+144 was poor, the high to the south of Britain giving westerly winds for all, but T+120 and T+96 were very good.

ECMWF scores 8/10- T+144 and T+120 were spot on, but T+96 had the high slightly underdone, like the GFS at T+144.

NOGAPS scores 7/10- T+144 unavailable, T+120 and T+96 had high pressure too far south, covering southern Britain and giving westerlies over the northern two-thirds.

JMA scores 8/10- high slightly too far east at T+144, very good at T+1230 and T+96.

GEM scores 7/10 also- high a bit out west at T+120, but accurately placed at T+96.

Overall a good set of outputs, but none were perfect.

[Thundery wintry showers]

13 September

992mb lows south and east of Iceland, 1024mb high over southern England, and WSW winds covering the northern half of Britain.

GFS scores a 7/10; T+144 had high pressure covering the whole of Britain but T+120 was good, low pressure slightly too far east to our north but otherwise accurate, T+96 spot on. Overall lost some marks for the poorish T+144.

UKMO also scores 7/10; T+144 had high pressure covering most of Britain, T+120 was similar but with high pressure only centred about 100 miles too far north, T+96 spot on.

ECMWF scores another 7/10. T+144 was very good, but T+120 had high pressure too far south and lows overdeepened giving strong westerlies in the north, T+96 slightly overdid high pressure to the east giving more of a SW flow. Again good but nowhere near perfect.

NOGAPS scores only 5/10; high significantly too far north at T+120 and T+96 and covering all of Britain.

JMA similarly scores 5/10. High too far north at T+144 and T+120, covering Britain, T+96 a bit better but still high extending up to S Scotland.

GEM scores 6/10. T+120 was poor with a Scandinavian Low and north-westerlies already dominating over Britain, but T+96, by contrast, was very good.

Overall a decent set of outputs from the big three but nothing spectacular. The three lesser models had troubles.

[Thundery wintry showers]

14 September

984mb low north-west of Scandinavia with north-westerly winds over Britain, with broken cloud cover, and a band of very light patchy rain moving SE. Coldest air headed out to the north and east of Britain.

GFS scores a 7/10; T+144 was poor with high pressure in charge, but T+120 and T+96 were reasonable with north-westerlies over Britain, though T+96 had the low slightly overdone.

UKMO scores 6/10; T+144 had high pressure in charge, T+120 was quite accurate though with the northerlies slightly overdone, T+96 had high pressure too far east to our south giving WNW winds.

ECMWF scores 6/10; T+144 and T+120 were both very good, but T+96 was poor, the Scandinavian Low a good few hundred miles too far west, giving strong westerlies.

NOGAPS scores 3/10 as at T+144 and T+120 it had high pressure over Britain, and at T+96 it had a low to our north giving strong westerlies, a bit like ECM's T+96.

JMA scores 5/10; low pressure was too far west at all three timeframes. T+144 and T+96 had high pressure over the south, T+120 had the north-westerlies spreading to most of Britain.

GEM scores only 4/10 as high pressure was too dominant over the south at T+120 and T+96 with northerlies only just coming into Scotland.

None of the models did particularly well. Surprisingly, with the exception of UKMO at T+120, the main problem was underdoing the northerly.

[Thundery wintry showers]

15 September

996mb low south of Iceland, WSW winds in the north, high pressure covering the extreme south.

GFS scores only 6/10- T+144 and T+120 had high pressure right over Britain, with northerlies in the east at T+120 also, but T+96 was spot on.

UKMO scores 5/10- high pressure was right over Britain at all three timeframes though the northerly wasn't overdone at T+120 to the same extent as GFS. Overall a slightly poorer output than GFS.

ECMWF scores 7/10 as high pressure was too far north at all three timeframes, but essentially over the south, so not as bad as UKMO or GFS.

NOGAPS scores 6/10- high pressure over Britain at T+144 and T+120, T+96 unavailable, but centred over the south at T+120.

JMA scores just 4/10 as northerlies were still ongoing at T+144 and T+120, high pressure over Britain at T+96.

GEM scores 6/10- high pressure too far north at T+120 and T+96, but centred mainly over the south.

Overall a tendency to position high pressure too far north and west, showing signs of overdoing the latter stages of the northerly.

16 September

992mb low west of Scandinavia giving strong, mild west to south-westerly flow, but with notably cold northerlies spreading south later in the day.

GFS scores 7/10 as low pressure was too far west at T+144, giving SW winds, but it was near-accurate at T+120 and T+96, especially T+96.

UKMO scores 7/10 also- almost identical to GFS.

ECMWF scores 9/10 as it was close to spot on at all three timeframes, with just minor quibbles with regards low positioning. An excellent set of outputs from ECM.

NOGAPS scores 7/10 as T+144 and T+96 had low pressure slightly too far west, but were otherwise good; T+120 unavailable.

JMA scores 6/10; T+120 was good, but T+144 and T+96 had low pressure too far west giving WSW winds generally, and northerlies not coming into the north.

GEM scores 7/10; T+120 good, T+96 had low slightly too far west giving WSW winds for all and northerlies not coming into the north.

Overall a decent set of outputs on the whole, except for the ECM which was excellent.

[Thundery wintry showers]

17 September

984mb Scandinavian Low, 1032mb high to the west. Quite an impressively potent northerly for the time of year I thought, despite it being very much a toppler.

GFS scores an 8/10 as T+144 and T+96 were both very accurate, a low in the North Sea marginally overdone, though T+120 had the northerly underdone a little.

UKMO scores only 6/10; the northerly underdone at T+144 with high pressure already knocking on our doorstep, and T+120 and T+96 had a low over southern Britain giving more of a north-easterly flow, and the coldest air's arrival being delayed.

ECMWF scores 8/10 as T+144 and T+96 were both very accurate; T+120 slightly less so with low pressure to the east underdone.

NOGAPS only available at T+120 so no mark. It had straight northerlies over Britain, so not bad.

JMA scores 4/10 as low pressure was too far west at T+144 giving WNW winds, and a low was over southern Britain at T+120, giving easterlies; T+96 unavailable.

GEM scores 6/10; T+120 had low pressure over East Anglia, but T+96 was quite accurate.

Overall, the northerly was well-handled by GFS and ECMWF, but less so by UKMO.

[Thundery wintry showers]

18 September

1032mb high off south-western Britain giving an anticyclonic north-westerly flow. It was a rather cold day for most, with temperatures struggling to recover from the northerly blast.

GFS scores only 5/10 because it overdid the northerly, especially at T+144 and T+96, with strong northerlies in the east, and warmer westerlies only just coming into the far north-west of the country. T+120 was better, but high pressure still too far north with northerlies overdone.

UKMO scores 7/10; northerly overdone at T+144, similar to GFS, but T+120 and T+96 had high pressure only slightly too far north, NNE winds in the SE, but north-westerlies elsewhere.

ECMWF scores 8/10; northerly slightly overdone at T+144 but very accurate at T+120 and T+96.

NOGAPS unavailable except T+144, when it overdid the northerly.

JMA also available only at T+144, when it had an easterly over Britain.

GEM completely unavailable; WZ problems prevented me from downloading the lesser models on 13/14 September, but the charts available on N-W had no such problems.

Overall, signs that the northerly was generally overdone in its latter stages (this also happened with the previous northerly). GFS's failing of positioning Atlantic high pressure too far north and prolonging northerly topplers was shown up clearly this time, while ECM handled the northerly most precisely.

[Thundery wintry showers]

19 September

984mb Icelandic Low, 1028mb high to the SW, and strong westerly winds over the British Isles, a typical 'westerly' autumn day really.

GFS scores 7/10- it was okay but unconvincing. T+144 and T+96 were reasonably accurate, but T+120 had the Icelandic Low a bit too far west, with SW winds over Britain and relatively high pressure clinging onto eastern areas.

UKMO also scores 7/10. T+144 unavailable, T+120 and T+96 fairly accurate but overdeepened low pressure over Iceland, giving very strong WSW winds, much stronger than actually happened. It's quite unlike UKMO to underdo Euro Highs and overdo Atlantic lows.

ECMWF scores 8/10 as it was more accurate generally at all three timeframes; low positioning wasn't consistent but the pattern over Britain was.

NOGAPS, JMA and GEM only available at T+96, with NOGAPS and GEM being quite accurate, and JMA positioning low pressure too far west.

NOGAPS, JMA and GEM should be available at two timeframes tomorrow. In the meantime, the main three models did reasonably, with ECM having a slight edge over UKMO and GFS.

[Thundery wintry showers]

20 September

980mb Icelandic Low, 1024mb high to the SE (Bartlett anyone?) with strong WSW winds over Britain.

GFS only scores 6/10- lows too far south and east, giving Britain a cyclonic/westerly flow and the Euro High underdone. T+96 was reasonably accurate though.

UKMO scores 7/10- unavailable at T+144, low pressure overdone at T+120, spot on at T+96.

ECMWF scores 9/10- almost spot on at all three timeframes, low pressure and high pressure accurately placed, T+144 had low pressure marginally overdone.

NOGAPS scores 7/10- low pressure overdone at T+120 giving a cyclonic westerly flow but T+96 was accurate.

JMA scores 7/10 also- T+120 was reasonably accurate but T+96 had low pressure slightly overdone.

GEM scores only 5/10 as low pressure was overdone at both T+120 and T+96, giving a WNW flow in the case of T+120.

Overall not an inspiring set of outputs, except for the ECM which was again outstanding.