12z Model Comparisons Results

[Thundery wintry showers]

28 March

Slack regime with very floppy low (1016mb) over southern and also northern Britain, northerlies for Ireland and southerlies for the far east. The north-east coast of England had its traditional sea fog.

GFS scores just 6/10; an Atlantic low was overdone at T+144 giving strong southerlies, T+120 had a slack pressure regime but with the low pressure to the south, and HP over central and northern Britain, but T+96 was very accurate.

UKMO was dodgy as usual at T+144; a large low over SW England, giving moderate SE winds for all. T+120 was accurate but T+96 had the main centre of low pressure too far east, giving northerlies for all. As with GFS only 6/10.

ECMWF had a deep Atlantic low off western Scotland at T+144 with strong SW winds in the north, T+120 was similar but with a less overdeepened low, then T+96 was reasonably accurate. Only 5/10.

NOGAPS was clueless; a low to the SW at T+144 giving ESE winds, a large low off W Scotland giving strong to gale force southerlies at T+120, and a low to the SE at T+96 giving NE'ly winds; only 2/10.

JMA had an overdone low to the north at T+144 and T+120, giving westerly winds for all; T+96 was reasonably accurate though; it was similar to ECMWF and scores 5/10.

GEM overdid LP to the west of Scotland at T+120 (T+144 unavailable) giving SW winds for most; at T+96 it had the main centre of low pressure too far east, giving northerlies for all. Only 3/10.

A disappointing set of outputs; UKMO and GFS continue to have problems, but following a very reliable spell, the ECMWF was even less good this time.

[Thundery wintry showers]

29 March

1012mb low covering the east coast of England, northerlies to the west, and north-easterlies over Scotland. It was dull and damp in north-east England, while western areas had some sunshine and wintry showers.

GFS was poor at T+144; a large high pressure system over the British Isles and no resemblance at all to the correct output. T+120 was slightly better but the low pressure was a good 500 miles too far east, with high pressure over the SW. T+96 was also poor, with a deep low over northern Scotland giving westerlies for all, and a nose of HP to our south. Only a 3/10 I'm afraid.

UKMO was better than GFS at T+144, but the low was too far to the SE with high pressure already nosing into north-western Britain. T+120 was about the same as GFS's T+120, the low too far east with high pressure already over SW Britain. T+96 was quite accurate though; the low slightly too far west this time, giving a slack airflow across Britain and northerlies only over Ireland. A 5/10, but not great.

ECMWF was poor at T+144; a large Scandinavian Low giving NW winds. T+120 was pretty recognisable though; the low slightly too far north, giving westerlies across the south, and northerlies across the north-west, and the HP to the west accurately placed. T+96 was very accurate; low slightly underdone though (central pressure >1012mb). Overall good enough for 7/10.

NOGAPS was poor at T+144; a deep low over W Ireland giving southerlies. T+120 was unavailable. T+96 wasn't too bad; the low over SE England underdone, but a fairly accurate output on the whole; a 5/10.

JMA was surprisingly good at T+144; the low slightly too far north, but not far off being correct. T+120 was much worse, with a very deep low off N Scotland giving strong westerlies. T+96 was extremely accurate. Overall, I think a 6/10 is fair; not as good as the ECM but better than any of the others.

GEM scores 7/10; T+144 unavailable, T+120 good but the low over SE England underdone, T+96 also good but the low over SE England slightly too far west, giving easterlies across the north. Both outputs were recognisable at least.

Overall, a decent performance by the ECMWF and also the GEM. GFS was notably poor.

[Thundery wintry showers]

30 March

1028mb high off N Scotland, 1012mb low in the Channel, and easterly winds.

GFS was good enough for a 6/10; T+144 had low pressure too far east giving a slack NE flow for all; T+120 had the low overdeepened and over southern England giving a more unsettled picture, then T+96 was very accurate.

UKMO was very poor at T+144; a high to the south, giving westerly winds of an anticyclonic nature. However UKMO was very accurate at T+120 and T+96, and so scores 7/10.

ECMWF also scores 7/10; at all three timeframes the low was slightly too far east giving an ENE flow, but it was reasonably close.

NOGAPS was unavailable at T+144; T+120 was good, then T+96 less good with the low too far east giving a NE flow; overall 6/10.

JMA scores only 4/10. T+144 had an Arctic northerly with a low to the east; T+120 and T+96 had the low too far east, giving a NNE flow for Britain.

GEM scores 8/10 on the evidence of T+120 and T+96, both of which were excellent, though slight quibbles about low positioning; the low tended to be slightly too far east though less so than with ECMWF.

Overall a surprisingly good output from the GEM; all of the models apart from JMA did okay.

[Thundery wintry showers]

31 March

1032mb high covering Scotland, with a rather nippy easterly affecting most of England & Wales.

GFS was good enough for 8/10 this time; high over Scotland at all three timeframes and easterlies over the south. The reason why it lost a couple of marks is because the lows to the north of the high were positioned somewhat haphazardly and the lows to the south likewise, but for Britain the outputs were very good.

UKMO was reasonable at T+144 but with the high too far north giving easterlies for all; T+120 and T+96 were almost identical to GFS; thus I award UKMO a 7/10 because of the T+144 being less good than that of GFS.

ECMWF scores 8/10; the high too far south at T+144 resulting in only moderate easterlies for the southern half of England, then T+120 and T+96 were accurate.

NOGAPS scores 7/10; quite good, but at T+144 the high was too far south, and at T+120 too far west giving more of a NE flow over the south.

JMA scores only 6/10; T+144 was poor with the high too far south and Atlantic systems pushing into northern Scotland, then T+120 and T+96 were good, but had the high too far west giving a NE flow over the southeast.

GEM available only at T+120 when it was quite good, but had the high too far west giving NE winds over the south.

Overall a pretty good set of outputs.

[Thundery wintry showers]

1 April

1036mb flat high over Scotland with strong east winds across the southern half of Britain.

GFS scores another 8/10; haphazard positioning of low to our S at T+144, giving more of an ESE flow for the south, then very accurate T+120 and T+96; Scotland was correctly under high pressure at all three timeframes.

UKMO had high pressure a long way too far east at T+144, situated over Scandinavia, with easterly winds for all; SE'ly in the far northeast of Scotland. However it was good at T+120 and T+96. I give this 7/10 because the T+144 was less good than that of GFS; same assessment for the second day running.

ECMWF scores 8/10; high slightly too far south at T+144, but accurate at T+120 and T+96.

NOGAPS poor at T+144 with the high a long way too far south, giving an anticyclonic-north-westerly flow. T+120 unavailable, then T+96 was accurate; overall it scores a 6/10.

JMA had the high too far south at T+144, giving westerlies over Scotland. T+120 and T+96 were reasonably accurate though; overall 7/10.

GEM available only at T+120 and T+96, when it was spot on; however the same could also be said of the other models, so it's difficult to justify giving it a higher mark (the others were marked down because of dodgy T+144 outputs). I'll award GEM 8/10 as well.

Overall, some dodgy outputs at T+144 but remarkable accuracy at T+120 and T+96. GFS, ECM and GEM were equally good.

[Thundery wintry showers]

2 April

1036mb high, covering the northwest of Scotland, an east-north-east flow elsewhere, turning to northerly in the far northeast.

GFS scores 7/10; it had the high over northern Britain at all three timeframes, but positioned too far east, giving easterlies over the south, and no hint of northerlies coming into the far northeast. High too far east by about 300 miles.

UKMO had the high too far south at T+144, with ENE winds only over the extreme south. However, T+120 and T+96 were extremely accurate, so it scores 8/10 overall.

ECMWF scores 8/10; generally accurate, high slightly too far east at all three timeframes but less so than GFS.

NOGAPS was only available at T+96 so no mark. High pressure was too far east.

JMA had the high too far south at T+144 by a long way, with westerlies over Scotland, but T+120 and T+96 were accurate; overall 7/10.

GEM scores 7/10; high way too far east at T+120 and T+96 but otherwise accurate.

The models kept positioning our high too far east, particularly in the case of GFS and GEM, while UKMO and JMA had the high too far south at T+144, though UKMO's T+120 and T+96 outputs were more accurate than those of any other model.

[Thundery wintry showers]

3 April

1036mb high covering north-western Britain, a north-easterly wind across most parts bringing in a lot of cloud.

GFS scores an 8/10 as it consistently had the high centred to the NW of Scotland and had easterly winds covering the south, it was docked a couple of marks for having the high too far south, restricting the easterly flow to southern England, at all three timeframes (mostly T+144).

UKMO had highest pressure too far south at T+144 giving anticyclonic/northerly winds, and the easterly flow having cleared the south. However, T+120 and T+96 were both very accurate- more so than GFS, so it also scores 8/10.

ECMWF scores only 7/10; the high was too far west at T+144 giving a stiff north/north-easterly airflow, but it was accurate at T+120 and T+96.

NOGAPS scores only 6/10; at T+144 and T+96 (T+120 unavailable) it had the high too far south, giving northerlies for most.

JMA scores 7/10; high too far NW at T+144, too far south at T+120, accurate at T+96.

GEM also scores 7/10; high too far south at T+120 but accurate at T+96, though with high slightly too far east, and more of an E'ly rather than NE'ly for the south.

Overall a reasonable set of outputs; UKMO and GFS were the most accurate.

And finally, here's an updated file showing the results:

model_comparisons.xls

[johnholmes]

tks for all that work Ian.

Over 12 months of data for us all to sift through.

I still find it pretty amazing that GFS remains in front of the UK Met O for this area. I wonder how well the UK Met O does over N America?

[Thundery wintry showers]

4 April

1028mb high to the west, covering Ireland, an anticyclonic/northerly drift for eastern areas. Most places had a very sunny day.

GFS scores an 8/10; it was pretty good. T+144 had the high too far west with a more pronounced NW flow for eastern areas, with a cold Arctic airflow penetrating close to Scotland, but T+120 and T+96 were both extremely accurate.

UKMO was similar to GFS at T+144- high slightly too far west, Arctic northerlies sweeping south towards Scotland. T+120 and T+96 were very accurate, so also an 8/10.

ECMWF scores only 7/10. T+144 had a cold Arctic northerly flow over Britain with the high a good 300 miles too far west; the high was only marginally too far west at T+120, though, and T+96 was very accurate.

NOGAPS scores 7/10 on the basis of T+120 and T+96- T+120 had the high too far west giving a stiff northerly breeze for eastern areas (though not especially cold) and T+96 likewise, but the high was only very slightly out in positioning.

JMA scores 8/10; high slightly too far south at T+144 giving westerlies across Scotland, but accurate at T+120 and T+96.

GEM also scores 8/10; reasonably accurate at both T+120 and T+96, but at T+96 the high was if anything slightly too far east, so no northerly flow in the east.

Overall a very good set of outputs, even allowing for the fact that with high pressure close by, there is a greater margin for error.

[Thundery wintry showers]

5 April

1024mb high covering western Britain with a slight NNW drift for eastern areas. For most parts it was a very warm, sunny day, with 20C exceeded quite widely, and 19C widely reached even in north-east England.

GFS scores 8/10; it had the high centred to the west of Britain at all three timeframes, but T+144 and T+120 had the high slightly too far east, with no north-westerly drift for eastern areas, and T+96 had the high slightly too far south with westerlies over Scotland.

UKMO scores 8/10; it was almost identical to GFS at all three timeframes, except T+120 where the high was slightly too far south, rather than east.

ECMWF scores 7/10; it had a chilly Arctic northerly covering eastern areas at T+144 with high pressure over the west, but was accurate at T+120 and T+96.

NOGAPS; almost identical to ECMWF at all three timeframes, so scores 7/10.

JMA was, again, almost identical to ECMWF and NOGAPS, with the chilly northerly for eastern areas at T+144 and then accurate outputs at T+120 and T+96; a 7/10.

GEM available only at T+120 so no mark; it had the high too far west with a chilly northerly for eastern areas.

Overall, a good set of outputs; GFS and UKMO edged it because they were also reasonably accurate at T+144, whereas the other models had a bit of a cold northerly blowing for eastern areas.

[Thundery wintry showers]

6 April

Large 1024mb high centred over south-western Britain. A slight north-westerly drift and cooler air over Scotland and north-east England, which brought a surprisingly cloudy day to those areas; some warm sunshine was experienced elsewhere.

GFS scores 7/10; it was quite poor at T+144 with a south-westerly flow and rain approaching western Scotland, but T+120 was reasonably accurate, with the high only slightly too far south; T+96 was almost perfect.

UKMO was worse than GFS at T+144; a large low off north-west Scotland giving a cyclonic/south-westerly flow. T+120 and T+96 were similar to GFS, but again T+120 was worse with the high a good 300 miles too far SW. Overall a 6/10.

ECMWF scores 8/10; a slack area of high pressure at T+144 with slack area of relatively low pressure across Scotland (1015-1020mb), then very accurate at T+120 and T+96.

NOGAPS scores only 5/10; high pressure over the south at all three timeframes but too far south, with T+144 and T+96 in particular having a large low off NW Scotland giving strong west/south-westerly winds away from southern England.

JMA scores 7/10; poor T+144 with south-westerlies and a low coming into W Scotland, but accurate at T+120 and T+96.

GEM available only at T+96, when it was pretty good.

Overall a good set of outputs, though with a tendency to overdo low pressure to the NW. The ECM was slightly more accurate than the others.

[Thundery wintry showers]

7 April

1028mb high centred over southern Britain, a slight westerly flow over Scotland; dull in the far north but for most it was sunny and warm.

GFS scores 8/10; very good at all three timeframes, T+144 had the high slightly too far south giving more pronounced westerlies for northern Scotland.

UKMO's T+144 was okay with the high slightly too far west, but an Arctic northerly was shown as penetrating into northern Scotland, so the low positioning to the north was all wrong. T+120 was also a bit out; high a good couple of hundred miles too far south, with westerlies for all; T+96 was accurate. Overall only a 6/10.

ECMWF scores 8/10; practically identical to GFS, high slightly too far south T+144, near-perfect T+120 and T+96.

NOGAPS had the high too far south at all three timeframes, especially T+120 when a pronounced westerly flow was shown for all. Overall, about as good/bad as UKMO so scores 6/10.

JMA had the high well to the south at T+144 giving a strong westerly flow for all and low pressure just north of Scotland; T+120 was better but still had the high slightly too far south; T+96 very good. Another 6/10.

GEM was very good at T+120 and T+96 and so scores 8/10; high slightly too far south at T+120.

GFS, ECMWF and GEM were all very good; the other three less so.

[Thundery wintry showers]

8 April

1028mb high off SW England, high pressure over the south but a strong westerly flow in the north, somewhat taking the edge off the high temperatures.

GFS scores 7/10; low pressure too far south at T+144 giving strong westerlies for all but the far south, T+120 was pretty good but with low overdeepened; T+96 was very accurate.

UKMO had the high too far north at T+144 with little or no wind for all. T+120 and T+96 were pretty good but with the Atlantic low slightly too far west; overall also a 7/10.

ECMWF scores 8/10; accurate at all three timeframes, but T+144 had the low to the north overdeepened.

NOGAPS had the low too far south at T+144 giving westerly winds; accurate at T+120, then high if anything slightly too far north at T+96, overall 7/10.

JMA scores only 6/10; a cyclonic/NW flow at T+144, better at T+120 but with low to north overdone, T+96 was then accurate.

GEM available only at T+120, when it was pretty good; no mark.

Overall, ECMWF slightly outshaded the other models.

[Thundery wintry showers]

9 April

1024mb high off SW England, and an anticyclonic WNW flow over Britain. Low off N Scotland.

GFS was a dodgy output this time and scores only 6/10. T+144 had the high too far north, lows to the north underdone, and a slack westerly flow. T+120 had the low north of Scotland a long way too far west, although the pattern over Britain was close to being correct. T+96 was accurate.

UKMO had the high too far north at T+144, with westerlies only over the far north. T+120 was reasonably accurate but with the low off N Scotland slightly too far east; T+96 was very good. Overall 7/10.

ECMWF scores only 6/10; high too far north at T+144, reasonably close at T+120 and T+96 but with the low off N Scotland a long way too far east, bringing cooler WNW winds of relatively northerly origin to all parts.

NOGAPS scores 6/10; high too far north at T+120, but well positioned at T+96 and T+144; however low positioning to the north was haphazard at all three timefreames.

JMA scores 6/10 also; a NW flow at T+144 giving cold Arctic air across Britain, then T+120 good but with overdeepened low to the north, T+96 then very accurate.

GEM unavailable except at T+96, when it had a cyclonic/north-westerly flow...

A so-so set of outputs, UKMO better than the others.

[Thundery wintry showers]

10 April

1024mb low off SW England, moderate/strong WNW flow for most, with 968mb low level with northern Scandinavia, to the north of Scotland.

GFS scores a 7/10; high too far west at T+144 with a cool NNW incursion shown, but accurate at T+120 and T+96; T+120 had our high slightly too far north.

UKMO was better than GFS at T+144; high slightly too far north though. T+120 and T+96 were both accurate, so 8/10 overall.

ECMWF, likewise, scores 8/10; high too far north at T+144, but accurate at T+120 and T+96.

NOGAPS also, surprisingly, scores 8/10, with the high too far north and lows to N underdone at T+144, but very accurate at T+120 and T+96.

JMA scores 7/10; high too far north at both T+144 and T+96 with little or no wind over England & Wales, but accurate at T+96.

GEM unavailable except at T+96 when the HP was too far north.

Tendency for the models to place high pressure too far north, the only particularly dodgy output was GFS at T+144. UKMO, ECMWF and NOGAPS were particularly good.

[Thundery wintry showers]

11 April

1024mb high centred over southern Britain and a slack westerly flow over Scotland; many places had yet another warm, dry, sunny day.

GFS scores 7/10; it had a high over Britain at all three timeframes but at T+144 it was too far east giving southerlies, and at T+120 too far west, giving an anticyclonic/northerly flow; T+96 was very good.

UKMO was good at T+144; high marginally too far NE, covering the whole country. T+120 and T+96 were very good also, though T+120 overdid low pressure to the north giving strong westerlies for northern Scotland. 8/10 overall.

ECMWF scores 7/10; it was consistent but the high was too far south at all three timeframes giving a fairly pronounced westerly flow across northern Britain.

NOGAPS scores 8/10; very accurate, though at T+144 the high was too far north, covering the whole country.

JMA scores 7/10; high too far east at T+144 giving SSW winds, T+120 very accurate, T+96 high slightly too far north, covering the whole country.

GEM unavailable.

Overall a good set of outputs.

[Thundery wintry showers]

12 April

1024mb high off eastern Britain with a very slack southerly flow.

GFS scores an 8/10; it was very good, high slightly too far south-west at T+144 giving weak westerlies over Scotland, then very accurate at T+120 and T+96.

UKMO also scores 8/10; high marginally too far south at all three timeframes giving slack SW flow for NW Scotland but otherwise very accurate.

ECMWF scores 9/10; almost spot on at all three timeframes, high perhaps slightly too far west at T+144, but not making any difference to the outcome anywhere.

NOGAPS scores 6/10; high too far east at T+144 giving SSW winds and low pressure coming into western Britain, T+120 and T+96 good but had the high too far SE giving SW winds in the north-west.

JMA scores 7/10; high too far west at T+144, and too far east at T+120 and T+96 giving SW winds in the far north-west, but otherwise accurate.

GEM scores 8/10 on the evidence of T+120 and T+96; very good, high slightly too far east with SW winds in the far NW.

A very good set of outputs.

[Thundery wintry showers]

13 April

1028mb SW Scandinavia High, with a slack east/ESE flow over Britain; the result was low cloud for many eastern areas.

GFS scores 8/10; it had the high marginally too far east generally, giving a more pronounced ESE flow than actually happened, but otherwise it was very good.

UKMO also scores 8/10; it had a southerly flow at T+144 with the high oriented slightly too far south, but T+120 and T+96 were very accurate.

ECMWF similar to UKMO; also 8/10, high slightly too far south at T+144 giving SW winds in the far north-west, but T+120 and T+96 were very accurate.

NOGAPS scores 8/10 also; identical to UKMO.

JMA scores 7/10; similar to GFS with high marginally too far east giving a relatively pronounced ESE flow, but at T+120 the high was significantly too far east, with a SSW flow covering western areas.

GEM unavailable except at T+120 when it had the high slightly too far east, with SSW winds in the west.

An excellent set of outputs.

[Thundery wintry showers]

14 April

High over southern Scandinavia, slack easterly flow covering the southern half of Britain, a lot of warm sunny weather across a large majority of the country, but there was fog on the east coast.

GFS was only good enough for 6/10 I'm afraid. T+144 had the high to the SE, giving WSW winds across the north; T+120 was dodgy with the high too far east giving a pronounced southerly flow, and T+96 was similar. However T+120 and T+96 were fairly accurate otherwise.

UKMO had the high too far south at T+144, with SW winds in the north. T+120 and T+96 were very accurate, with the high only marginally too far east, southerlies in the west, easterlies in the south (very slack airflows). Overall 8/10; it was much better than GFS.

ECMWF had the high too far east at T+144 giving southerly winds, but was every bit as good as the UKMO at T+120 and T+96; overall 8/10 also.

NOGAPS available only at T+144 and T+120, scoring 5/10 on the basis of those as the high was too far east on both occasions giving a SSW flow.

JMA was poor, scoring 4/10; a SSW flow at T+144 with high too far east, and then westerlies spreading east at T+120 and T+96.

GEM: only available at T+96 when the high was too far east, giving southerlies.

Overall, I can see why there was a tendency for the GFS to show a thundery breakdown. High pressure was too far east, giving southerly winds. The UKMO and ECMWF handled it much better.

[Thundery wintry showers]

15 April

Slack high pressure, 1024mb, covering most of Britain, a WSW flow for the far north of Scotland. T+96 outputs unavailable.

GFS scores 6/10; fairly accurate at T+144 with highs to the W and E and a slack area of neither low nor high pressure over Britain, but T+120 overdid the low to the NW, giving SW winds for most parts.

UKMO almost identical to GFS at T+144, but it was extremely accurate at T+120, and so scores 8/10.

ECMWF scores 7/10; high centred too far east at both timeframes giving a slack southerly flow, lows too far out to the west.

NOGAPS only available at T+144, with a slack SSW flow and a high to the east.

JMA had high pressure too far to the SE at T+144, with a SW flow for most; T+120 had the high off SW England with a wsterly flow; only 5/10.

GEM unavailable.

Overall, UKMO gave the best output. The next couple of days will be missing due to insufficient data but I expect summaries to restart on Thursday 19 April.

[Thundery wintry showers]

19 April now:

1020mb high over Ireland/to the west, a north-westerly flow over Britain, a cold front sweeping Arctic air south, but the coldest air only touching northern Scotland.

GFS was only good enough for 5/10; high too far to the SW at T+144, giving a cyclonic/WNW flow with northerlies coming into Scotland, but ironically this was its best output. T+120 and T+96 had high pressure a long way too far west, giving northerlies, and the cold front passing southwards far too quickly, wintry showers likely in Scotland.

UKMO was reasonable at T+144; high fairly accurately placed, northerlies over Britain but the coldest air only just coming into Scotland. T+120 was poor, with the high to the NW and a NNE flow for all; T+96 was then pretty accurate, though the high was too far west giving cyclonic/NW flow (similar to GFS at T+144). Overall not great, but better than GFS; a 6/10.

ECMWF also scores only 6/10; high too far North at T+144 and T+120 giving a slack easterly flow, but quite accurate at T+96.

NOGAPS scores just 4/10; a slack low over western Britain at T+144 and T+120, giving a cyclonic/southerly flow for most, but T+96 was fairly accurate, with high too far to the SW giving cyclonic/NW flow.

JMA also scores 4/10; high too far NW at T+144 giving cold NE winds, T+120 and T+96 had high pressure too dominant over the south with westerly winds over Scotland.

GEM available only at T+96, when it was fairly accurate but had too strong a HP over the south.

Overall, upon resumption, these outputs were not as good as for earlier days. The input of cold air was generally overdone, particularly by the GFS.

[PersianPaladin]

Just want to say this is fantastic stuff. Keep up the good work Ian. I've been trawling through the spreadsheet you compiled and it really is quite thorough.

[Thundery wintry showers]

20 April

1024mb high off eastern Britain with southerly winds for all; 996mb Atlantic lows.

GFS scores 6/10; T+144 had the high in the Atlantic giving NW winds, T+120 had the high over southern Britain giving westerlies across the north, and T+96 was very accurate.

UKMO was better than GFS at T+144; high over S England (i.e. too far west) giving westerlies over N Scotland but at least it wasn't in the mid-Atlantic. T+120 dodgy; high too far north, giving easterly winds for most. T+96 was then reasonably accurate. Overall 7/10, but not too impressive.

ECMWF scores 8/10; high slightly too far north at T+144 giving more of a SE flow over the south, then T+120 and T+96 very accurate.

NOGAPS had high pressure too far north at T+144 giving easterly winds for all; T+120 had the high off W Scotland giving an anticyclonic/NE flow, T+96 was then fairly accurate but high was too far west, covering the whole country. Overall only 5/10.

JMA had the high too far north at T+144 and T+120, giving E and then SE winds, but T+96 was accurate; overall 6/10.

GEM was good enough for 6/10; T+120 had easterly winds over the south with the high too far north, T+96 was fairly accurate but had high too far west, covering the whole country.

Overall the ECM came out on top this time.

[Thundery wintry showers]

21 April

1024mb high off SE England, 992mb Atlantic low complex centred W of Iceland, and a SW flow over Britain.

GFS scores only 5/10; high way too far west at both T+144 and T+120, centred over E England with an anticyclonic/southerly flow over Britain and no hint of Atlantic systems, but T+96 was then reasonably accurate.

UKMO similarly poor at T+144; high over N Britain and slack easterly winds for south. T+120 and T+96 were okay with SW winds shown but low pressure too strong in the Atlantic; overall 6/10.

ECMWF scores 7/10; high too far west at T+144 and T+120 (though much less so than for GFS) with SW winds and Atlantic systems just approaching the west, then T+96 was very accurate.

NOGAPS scores 5/10; almost identical to GFS.

JMA scores 6/10; high too far north-west at T+144 giving SSE winds, then T+120 and T+96 weren't far away, but high was too far west, giving SSW winds.

GEM scores 7/10; high too far W at T+120, but very accurate at T+96.

Overall the models seemed to rather under-do the Atlantic, with high pressure holding on in the east for too long.

[Thundery wintry showers]

22 April

996mb low off N Scotland, SW flow over Britain, and a rather mild damp day for the north; dry and sunny in the southeast.

GFS scores 7/10; low pressure too far west at T+144 giving southerly winds, but T+120 and T+96 were reasonably accurate; T+120 had too strong a SW flow due to the lows being overdone.

UKMO had lows too far south at T+144 and a Scandinavian Low, with WSW winds in the south and slack winds in the north- rather poor. Westerlies at T+120 with low to N overdone, and T+96 was okay but too anticyclonic with HP dominant in the south; only 5/10.

ECMWF scores 9/10; it was almost spot on at all three timeframes, SW winds slightly too strong at T+120.

NOGAPS scores only 4/10; high pressure too far North at T+144 giving SW winds only in the far north, while T+120 had low pressure too far south, with low centred over Scandinavia and a cyclonic westerly flow. T+96 unavailable.

JMA scores 7/10; low pressure overdone at T+144 giving a strong cyclonic/WSW flow, but closer at T+120 with the low only slightly overdone and too far west, very accurate at T+96.

GEM scores 7/10 also; reasonably accurate at T+120 and T+96, but low pressure centred too far west giving more of a SSW flow.

Overall, the ECM was way ahead of any other model this time around, and the UKMO was surprisingly poor.