Raw model output on the native 0.25° grid — no MOS, no bias correction, no post-processing. Scores understate operational forecast quality. Read the fairness statement →

CastCheck methodology v0.2 · data through 2026-08-30

San Francisco Intl · GraphCast (IFS init) · lead day 8

KSFO · graphcast_ifs · lead 8

Target day = initialization date + 8 days. This page is a permanent link: the numbers change as history accumulates, the address and the methodology version do not.

Scores use the latest model version segment, GRAP_v100, starting 2026-01-01 — days before that ran a different version of the model and are excluded rather than silently pooled. Of the scored days, 1 carry a QC flag on the observation.

JSON for this page CSV of the daily errors Cite this page

Every number for this combination

All four windows × both initializations × both interpolations × both variables. Errors are computed in °C and shown in °F; CI is a 95 % moving-block bootstrap interval. MAE debiased and skill, debiased remove the constant offset of the window first, so a station that the model always reads a degree too warm is not scored as if it had no skill — compare the two skill columns to see how much of the error is a fixed offset. n QC next to n counts days whose observation carries a QC flag.

VariableInitMethod Windown MAE °F MAE debiased Bias °F RMSE °F ±1 °F±2 °F±3 °F Skill Skill, debiased Period
tmax00Zbilinear30d22 1 QC6.76 [5.73, 7.96]2.51-6.25 [-7.87, -4.58]7.16 [6.06, 8.26]0% [0%, 0%]0%5%-0.93+0.282026-08-09 → 2026-08-30
tmax00Zbilinear90d22 1 QC6.76 [5.60, 8.28]2.51-6.25 [-8.27, -4.25]7.16 [5.87, 8.56]0% [0%, 0%]0%5%-0.93+0.282026-08-09 → 2026-08-30
tmax00Zbilinear365d28 1 QC5.70 [3.36, 7.50]2.89-5.28 [-7.28, -2.70]6.46 [4.37, 7.92]14% [0%, 46%]14%18%-0.61+0.172026-01-09 → 2026-08-30
tmax00Zbilinearall28 1 QC5.70 [2.90, 7.46]2.89-5.28 [-7.27, -2.43]6.46 [3.88, 7.85]14% [0%, 50%]14%18%-0.61+0.172026-01-09 → 2026-08-30
tmax00Znearest30d22 1 QC2.76 [1.81, 4.01]2.75-0.14 [-1.49, 2.14]4.02 [2.09, 5.99]23% [14%, 36%]50%68%+0.21+0.212026-08-09 → 2026-08-30
tmax00Znearest90d22 1 QC2.76 [1.73, 4.23]2.75-0.14 [-1.53, 2.98]4.02 [1.92, 6.47]23% [14%, 38%]50%68%+0.21+0.212026-08-09 → 2026-08-30
tmax00Znearest365d28 1 QC2.68 [1.90, 3.82]2.58-0.46 [-1.75, 1.56]3.83 [2.25, 5.67]29% [17%, 48%]50%68%+0.24+0.262026-01-09 → 2026-08-30
tmax00Znearestall28 1 QC2.68 [1.88, 3.76]2.58-0.46 [-1.78, 1.53]3.83 [2.17, 5.60]29% [17%, 50%]50%68%+0.24+0.262026-01-09 → 2026-08-30
tmax12Zbilinear30d22 1 QC6.11 [4.72, 7.63]2.99-5.77 [-7.57, -3.73]6.88 [5.31, 8.36]0% [0%, 0%]9%18%-0.74+0.142026-08-09 → 2026-08-30
tmax12Zbilinear90d22 1 QC6.11 [4.38, 8.09]2.99-5.77 [-8.08, -3.29]6.88 [4.96, 8.56]0% [0%, 0%]9%18%-0.74+0.142026-08-09 → 2026-08-30
tmax12Zbilinear365d27 1 QC5.29 [2.92, 7.20]3.12-4.95 [-7.20, -2.17]6.30 [3.80, 8.07]11% [0%, 38%]19%30%-0.44+0.132026-01-09 → 2026-08-30
tmax12Zbilinearall27 1 QC5.29 [2.56, 7.16]3.12-4.95 [-7.09, -2.01]6.30 [3.27, 7.95]11% [0%, 43%]19%30%-0.44+0.132026-01-09 → 2026-08-30
tmax12Znearest30d22 1 QC3.21 [2.22, 4.82]3.25+0.78 [-0.94, 3.49]4.22 [2.77, 5.89]18% [5%, 29%]45%59%+0.08+0.062026-08-09 → 2026-08-30
tmax12Znearest90d22 1 QC3.21 [2.16, 5.36]3.25+0.78 [-1.02, 4.54]4.22 [2.64, 6.35]18% [0%, 30%]45%59%+0.08+0.062026-08-09 → 2026-08-30
tmax12Znearest365d27 1 QC2.96 [1.98, 4.49]2.98+0.41 [-1.18, 2.95]3.97 [2.55, 5.65]26% [8%, 47%]48%63%+0.19+0.172026-01-09 → 2026-08-30
tmax12Znearestall27 1 QC2.96 [1.93, 4.49]2.98+0.41 [-1.18, 3.00]3.97 [2.55, 5.60]26% [11%, 50%]48%63%+0.19+0.172026-01-09 → 2026-08-30
tmin00Zbilinear30d22 1 QC1.49 [0.87, 2.27]1.33-0.96 [-2.06, 0.08]1.89 [1.06, 2.57]45% [18%, 67%]73%77%-0.72-0.502026-08-09 → 2026-08-30
tmin00Zbilinear90d22 1 QC1.49 [0.85, 2.49]1.33-0.96 [-2.35, 0.38]1.89 [1.04, 2.70]45% [10%, 67%]73%77%-0.72-0.502026-08-09 → 2026-08-30
tmin00Zbilinear365d28 1 QC1.89 [1.06, 3.05]1.89-0.03 [-1.72, 2.08]2.34 [1.33, 3.27]36% [0%, 61%]57%71%+0.35+0.452026-01-09 → 2026-08-30
tmin00Zbilinearall28 1 QC1.89 [1.03, 3.12]1.89-0.03 [-1.62, 2.45]2.34 [1.29, 3.33]36% [0%, 61%]57%71%+0.35+0.452026-01-09 → 2026-08-30
tmin00Znearest30d22 1 QC1.51 [1.22, 1.80]1.20-0.89 [-1.39, -0.17]1.71 [1.35, 1.99]27% [15%, 38%]68%95%-0.75-0.352026-08-09 → 2026-08-30
tmin00Znearest90d22 1 QC1.51 [1.19, 1.84]1.20-0.89 [-1.46, 0.07]1.71 [1.28, 2.01]27% [13%, 38%]68%95%-0.75-0.352026-08-09 → 2026-08-30
tmin00Znearest365d28 1 QC1.45 [1.12, 1.75]1.10-0.96 [-1.41, -0.27]1.67 [1.33, 1.96]36% [21%, 57%]68%96%+0.50+0.682026-01-09 → 2026-08-30
tmin00Znearestall28 1 QC1.45 [1.12, 1.70]1.10-0.96 [-1.40, -0.28]1.67 [1.31, 1.93]36% [21%, 59%]68%96%+0.50+0.682026-01-09 → 2026-08-30
tmin12Zbilinear30d22 1 QC1.54 [0.99, 2.39]1.44-0.95 [-2.11, 0.14]1.96 [1.25, 2.70]45% [23%, 60%]64%77%-0.79-0.632026-08-09 → 2026-08-30
tmin12Zbilinear90d22 1 QC1.54 [0.98, 2.61]1.44-0.95 [-2.46, 0.46]1.96 [1.23, 2.84]45% [17%, 62%]64%77%-0.79-0.632026-08-09 → 2026-08-30
tmin12Zbilinear365d27 1 QC1.76 [1.04, 2.73]1.75-0.27 [-1.77, 1.54]2.19 [1.31, 2.97]41% [17%, 59%]56%74%+0.35+0.462026-01-09 → 2026-08-30
tmin12Zbilinearall27 1 QC1.76 [1.09, 2.77]1.75-0.27 [-1.72, 1.85]2.19 [1.32, 3.00]41% [16%, 57%]56%74%+0.35+0.462026-01-09 → 2026-08-30
tmin12Znearest30d22 1 QC1.47 [1.17, 1.80]1.30-0.94 [-1.58, -0.08]1.80 [1.52, 2.03]41% [18%, 57%]59%91%-0.70-0.472026-08-09 → 2026-08-30
tmin12Znearest90d22 1 QC1.47 [1.14, 1.87]1.30-0.94 [-1.75, 0.26]1.80 [1.49, 2.07]41% [12%, 59%]59%91%-0.70-0.472026-08-09 → 2026-08-30
tmin12Znearest365d27 1 QC1.55 [1.25, 1.92]1.25-1.12 [-1.85, -0.25]1.84 [1.55, 2.09]37% [13%, 57%]59%93%+0.43+0.622026-01-09 → 2026-08-30
tmin12Znearestall27 1 QC1.55 [1.26, 1.90]1.25-1.12 [-1.85, -0.28]1.84 [1.57, 2.07]37% [14%, 54%]59%93%+0.43+0.622026-01-09 → 2026-08-30

+ model too warm · − model too cold · bias interval includes zero · ★ lowest MAE · = not distinguishable from the leader · ▼ significantly worse · n < 30 greyed and unranked · CI is a 95 % moving-block bootstrap interval.

Daily error, last 90 days

Daily forecast minus observed error in °F for GraphCast (IFS init) at KSFO, lead day 8, 00Z bilinear tmax+20+100-10-202026-08-09: -7.06 °F2026-08-10: -10.76 °F2026-08-11: -9.42 °F2026-08-12: -10.36 °F2026-08-13: -5.92 °F2026-08-14: -8.39 °F2026-08-15: -3.33 °F2026-08-16: -6.91 °F2026-08-17: -8.66 °F2026-08-18: -6.75 °F2026-08-19: -2.44 °F2026-08-20: -5.75 °F2026-08-21: -3.85 °F2026-08-22: -6.02 °F2026-08-23: -8.41 °F2026-08-24: -9.03 °F2026-08-25: -4.54 °F2026-08-26: -10.51 °F2026-08-27: -4.78 °F2026-08-28: +5.6 °F2026-08-29: -4.42 °F2026-08-30: -5.72 °F2026-08-092026-08-30°F, forecast − observed
Forecast − observed, °F, 00Z bilinear daily maximum. The dashed line is zero; the thick line is a centred 7-day mean, drawn only where the whole window is present. Gaps are days with no score, never interpolation. Hover a point for its date and value.
The same series as a table (22 days, newest first) · download CSV
Climatological dayError °F
2026-08-30-5.72
2026-08-29-4.42
2026-08-28+5.60
2026-08-27-4.78
2026-08-26-10.51
2026-08-25-4.54
2026-08-24-9.03
2026-08-23-8.41
2026-08-22-6.02
2026-08-21-3.85
2026-08-20-5.75
2026-08-19-2.44
2026-08-18-6.75
2026-08-17-8.66
2026-08-16-6.91
2026-08-15-3.33
2026-08-14-8.39
2026-08-13-5.92
2026-08-12-10.36
2026-08-11-9.42
2026-08-10-10.76
2026-08-09-7.06

Error distribution

Distribution of the signed daily error in °F for GraphCast (IFS init) at KSFO, lead day 8-20 to -18 °F: 0 days-18 to -16 °F: 0 days-16 to -14 °F: 0 days-14 to -12 °F: 0 days-12 to -10 °F: 3 days-10 to -8 °F: 5 days-8 to -6 °F: 4 days-6 to -4 °F: 7 days-4 to -2 °F: 4 days-2 to 0 °F: 3 days0 to 2 °F: 1 day2 to 4 °F: 0 days4 to 6 °F: 1 day6 to 8 °F: 0 days8 to 10 °F: 0 days10 to 12 °F: 0 days12 to 14 °F: 0 days14 to 16 °F: 0 days16 to 18 °F: 0 days18 to 20 °F: 0 daysdashed: P50 |e| 5.73 · P90 |e| 9.7 °F-200+20 °Fn = 28 days
All scored days for 00Z bilinear daily maximum, 20 equal bins. The dashed marks are the 50th and 90th percentiles of the absolute error, mirrored on both sides. Hover a bar for its bin and count.
Bin counts as a table
From °FTo °FDays
-20.00-18.000
-18.00-16.000
-16.00-14.000
-14.00-12.000
-12.00-10.003
-10.00-8.005
-8.00-6.004
-6.00-4.007
-4.00-2.004
-2.00+0.003
+0.00+2.001
+2.00+4.000
+4.00+6.001
+6.00+8.000
+8.00+10.000
+10.00+12.000
+12.00+14.000
+14.00+16.000
+16.00+18.000
+18.00+20.000

Month by month

00Z bilinear daily maximum, newest first.

Monthn MAE °FBias °F
2026-08226.76 -6.25
2026-0161.82 -1.73

Availability

ModelScored period DaysCoverage
GraphCast (IFS init)2026-01-09 → 2026-08-30 28
GraphCast (IFS init)2026-01-09 → 2026-08-30 28
GraphCast (IFS init)2026-01-09 → 2026-08-30 27
GraphCast (IFS init)2026-01-09 → 2026-08-30 27

Compared with other models

Paired bootstrap on the days both models cover (90d, 00Z, bilinear, daily maximum). Negative means GraphCast (IFS init) has the lower MAE. "Distinguishable" means the 95 % interval excludes zero.

Other modelCommon days MAE difference °FDistinguishable
ECMWF AIFS Single 22 +1.40 [-0.08, 2.30] no
NCEP GFS 22 +1.87 [0.74, 2.92] yes
ECMWF IFS HRES 17 +0.73 [0.03, 2.11] yes
Persistence (baseline) 22 +3.26 [1.79, 5.51] yes

Cite this page

ShortCastCheck, KSFO graphcast_ifs lead 8, methodology v0.2, accessed 2026-08-31, https://castcheck.zifanzhang.com/station/KSFO/model/graphcast_ifs/lead/8/

LongCastCheck (2026). Station-level verification of raw weather-model 2 m temperature forecasts: KSFO (San Francisco Intl), model graphcast_ifs, lead day 8. Methodology version 0.2, schema version 0.1, data through 2026-08-30. Accessed 2026-08-31. https://castcheck.zifanzhang.com/station/KSFO/model/graphcast_ifs/lead/8/

Accessed2026-08-31 (data through 2026-08-30; page built 2026-08-31T03:52:06+00:00)

JSON CSV all scores methodology v0.2 licence