National Hurricane Center 2011 Forecast Verification

National Hurricane Center 2011 Forecast Verification

John P. Cangialosi and James L. FranklinHurricane Specialist Unit

National Hurricane Center

NOAA Hurricane Conference29 November 2011

1

Verification Rules Verification rules unchanged for 2011. Results

presented here are preliminary. System must be a tropical or subtropical

cyclone at both forecast initial time and verification time. All verifications include depression stage (including GPRA goals).

Special advisories ignored (original advisory is verified.

Skill baselines are recomputed after the season from operational compute data. Decay-SHIFOR5 is the intensity skill benchmark.

2011 Atlantic Verification

Values in green exceed all-time records.

48 h error GPRA targets

Track: 87 n mi (met)

Intensity: 13 kt (missed)

So what else is new?

VT NT TRACK INT(h) (n mi) (kt)============================ 000 392 9.7 1.7012 354 29.3 5.8024 311 45.7 9.3036 273 61.2 11.6048 237 74.2 13.7072 187 110.6 16.5096 151 169.5 17.0120 127 253.1 18.1

Atlantic Track Errors vs. 5-yr Mean

Official forecasts were better than the 5-year mean, though the season’s storms were “easier” than normal.

2011 Track GuidanceOfficial forecast skill very close to consensus aids (even a little better)

EMXI and GFSI best models overall.

GFS ensemble mean not as good as deterministic GFS.

Continued poor performance of GFNI and NGPI. Bad year for EGRI.

HWRF and GHMI middle of the pack.

BAMM beat both regional models at 96 and 120 h.

Atlantic Track Errors by Storm

50.6

89.5

133.2

174.2

214.8

Bret, Don, Irene, Katia, Lee, Rina, Seansuccesses

Struggled with Maria, Ophelia

Ophelia’s Reformation

120 h forecast

Verifying position

Atlantic Track Biases

VT % of total error

24 h 20%

48 h 32%

72 h 34%

96 h 31%

120 h 34%24 h

48 h72 h

96 h

120 h

North

South

EastWest

Atlantic Intensity Errors vs. 5-yr Mean

Official forecast errors were a little better than the five-year mean but the season’s storm were easier to forecast than normal.

2011 Intensity Guidance

No operational aid was skillful at or beyond 72 h.

Dynamical models performed very poorly.

LGEM was best individual model overall.

Atlantic Intensity BiasVT % of total error

24 h 12%

48 h 16%

72 h 16%

96 h 25%

120 h 27%

2011 East Pacific VerificationVT NT TRACK INT(h) (n mi) (kt)============================000 236 7.8 1.6012 214 25.5 6.6024 195 40.5 11.2036 177 53.2 13.6048 157 69.6 14.8072 119 106.8 17.1096 87 158.8 19.2120 57 176.3 17.1

Values in green exceeded all-time lows.

E. Pacific Track Errors vs. 5-yr Mean

Even though the season’s storms were much harder than normal to forecast, the official forecast errors were still lower than the 5-year mean.

E. Pacific Intensity Errors vs. 5-yr Mean

Official forecasts were about the same as the 5-year mean, even though the season’s storms were quite a bit harder than normal.

2011 Track GuidanceOfficial forecasts near the TVCE and FSSE.

EMXI best model in this basin too.

GFS ensemble mean is quite skillful and better than the deterministic GFS.

EGRI and NGPI performed well in the EPAC.

GHMI middle of the pack, HWRF trails.

BAMS and BAMM beat the regional models at 96 and 120 h.

2011 Intensity GuidanceOfficial forecasts performed better than most of the guidance.

Good year for GFNI, much better than GHMI and HWFI.

Statistical and consensus models are pretty close.

2011 Genesis Forecast Verification

Atlantic forecasts well calibrated throughout. Much improved this year.

Some progress made in reducing the east Pacific under-forecast bias.

Atlantic East Pacific

Summary Track and intensity forecasts in both basins

were better than their long-term means. In the East Pacific, these forecasts were better than what would have been expected based on forecast difficulty.

EMXI is the best model in both basins. GFS ensemble mean was a good performer in the East Pacific.

Numerical guidance for intensity had little or no skill, particularly in the Atlantic. Regional hurricane models (GHMI and HWFI) were disappointing.

Genesis forecasts improved (especially in the Atlantic).