For tomorrow at Lakenheath, GFS wins; ECMWF AIFS saw the wet week from 6 days ahead.
Rainfall totals and rain timing are different tests. This report checks the models on both.
From 17 August 2026 to 23 August 2026, Lakenheath recorded 22.2 millimetres of rain. At least one millimetre fell on 4 days; 21 August 2026 was the wettest with 13.5 millimetres.
For the question of whether it will rain tomorrow, GFS scored best over the whole test period at 66.5. Further out, ECMWF AIFS was the most consistent: from 6 days ahead, its weekly estimate stayed within a factor of two. One day ahead it expected 19.9 millimetres.
At 9 days ahead, GFS forecast 14.8 millimetres against 22.2 measured. It was the only model whose run covered the whole week at that distance. One day ahead, estimates still ranged from 2.7 by GFS to 19.9 by ECMWF AIFS.
| Notice | ECMWF AIFS Europe, AI | ECMWF IFS Europe | GFS United States | ICON Germany | UKMO United Kingdom |
|---|---|---|---|---|---|
| 9 days ahead | 9.0139 h · to 23 August | 0.4139 h · to 23 August | 14.82 wet days | Not available | Not available |
| 8 days ahead | 23.35 wet days | 18.32 wet days | 5.92 wet days | Not available | Not available |
| 7 days ahead | 7.81 wet day | 15.94 wet days | 10.31 wet day | 0.014 h · to 17 August | 0.02 h · to 17 August |
| 6 days ahead | 40.36 wet days | 15.35 wet days | 4.82 wet days | 2.338 h · to 18 August | 2.526 h · to 18 August |
| 5 days ahead | 27.04 wet days | 23.07 wet days | 17.34 wet days | 0.162 h · to 19 August | 2.250 h · to 19 August |
| 4 days ahead | 14.03 wet days | 22.73 wet days | 1.51 wet day | 3.485 h · to 20 August | 0.373 h · to 20 August |
| 3 days ahead | 12.65 wet days | 7.62 wet days | 6.41 wet day | 4.0103 h · to 21 August | 2.491 h · to 21 August |
| 2 days ahead | 23.87 wet days | 10.83 wet days | 13.63 wet days | 5.4127 h · to 22 August | 7.4115 h · to 22 August |
| 1 day ahead | 19.96 wet days | 9.54 wet days | 2.71 wet day | 18.5150 h · to 23 August | 14.9139 h · to 23 August |
| Within a factor of 2 from | 6 days ahead | never | never | does not reach | does not reach |
The wettest day, hour by hour
A good weekly total does not mean the showers fell in the right hours. On 21 August 2026, Lakenheath measured 13.5 millimetres across 10 of the 23 hours. ECMWF AIFS was the most useful at short range: one day ahead it caught 9 of those 10 wet hours with 2 false alarms.
| Notice | ECMWF AIFS Europe, AI | ECMWF IFS Europe | GFS United States | ICON Germany | UKMO United Kingdom |
|---|---|---|---|---|---|
| 7 days ahead | 9+2 false | 0+0 false | 1+2 false | beyond horizon | beyond horizon |
| 6 days ahead | 10+7 false | 7+2 false | 3+4 false | 0+0 false | 6+0 false |
| 5 days ahead | 10+7 false | 8+0 false | 0+0 false | 9+7 false | 4+0 false |
| 4 days ahead | 10+7 false | 4+0 false | 0+3 false | 6+0 false | 1+5 false |
| 3 days ahead | 10+7 false | 6+1 false | 3+0 false | 3+0 false | 4+2 false |
| 2 days ahead | 10+7 false | 8+0 false | 0+0 false | 8+0 false | 4+0 false |
| 1 day ahead | 9+2 false | 6+0 false | 6+0 false | 5+1 false | 2+0 false |
Which model timed tomorrow's rain best?
Across all 10 days, GFS offered the best balance between missed rain and false alarms, with a score of 66.5. ECMWF AIFS followed closely at 65.9. UKMO finished last at 30.5: it found many wet hours but raised 9 false alarms. For rain at Lakenheath one day ahead, GFS is therefore the best choice in this test.
| Model | Rain score | Caught | Missed | False alarm | Hours called wet |
|---|---|---|---|---|---|
| GFS United States | 22 | 20 | 2 | 22 | |
| ECMWF AIFS Europe, AI | 34 | 8 | 44 | 99 | |
| ECMWF IFS Europe | 23 | 19 | 11 | 40 | |
| ICON Germany | 15 | 27 | 4 | 15 | |
| UKMO United Kingdom | 9 | 33 | 9 | 19 |
About this comparison
The selected wet week runs from 17 August 2026 to 23 August 2026. The hourly comparison uses its wettest day, 21 August 2026 with 13.5 millimetres. Forecasts come from the public models' 00 UTC runs and are compared with the gauge at Lakenheath.
How rain is judged
An hour counts as wet from 0.1 millimetres. A forecast may be 1 hour early or late; when the gauge has no reading, that hour is left out for every model. This is a test at one gauge in Lakenheath, not a verdict on a whole country.
At Lakenheath, the scheduled METAR at 55 minutes past the hour carries a P group for rain during the preceding hour. It is compared with the model hour ending five minutes later. No P group in a scheduled report means dry; no scheduled report means unknown, and that hour is left out.
We compare archived model output with independent station observations. Data is at data.json, the full method at methodology and corrections at corrections.
The live version of this report is an app
After The Forecast comes out on a schedule. VeriSky keeps the rolling scores current for the places you care about and shows how models change with forecast distance.
Free standings · No account · No cookies