Observed and Forecast Tracks: Northern Hemisphere 2008
All tropical cyclones in the northern hemisphere season in 2008 are listed below including details of their start and end dates and maximum sustained wind (MSW). Qualitative comments are included on the performance of the Met Office global model in predicting the track of each tropical cyclone. Click on the name of the tropical cyclone to see a chart of the observed track and forecast tracks from the Met Office global model. Verification statistics on each tropical cyclone are available upon request from tropical_cyclones@metoffice.gov.uk. At the end of the season a verification report on the performance of the Met Office global model will be published.
| Names & sID | Start & end dates | 1-/10-minute MSW | Comments on Met Office global model forecasts |
|---|---|---|---|
| 01W | 13-16 January | 35/- knots | 01W only briefly attained tropical storm status (according to JTWC) and no forecasts were verified. |
| Neoguri (02W) | 14-19 April | 95/80 knots | There was a left of track bias in early forecasts which resulted in track forecast errors above last season's average from T+48 onwards. However, skill scores against CLIPER were good. |
| Rammasun (03W) | 07-13 May | 135/105 knots | Track forecast errors were near to last season's average at short lead times, but were large at longer lead times. This was due to early forecasts failing to predict the acceleration of the typhoon to the north-east as it started extra-tropical transition. Hence, there was a slow bias at longer lead times. However, skill scores against CLIPER were high. |
| Matmo (04W) | 14-16 May | 40/45 knots | Matmo only briefly reached tropical storm status and no forecasts were verified. |
| Halong (05W) | 15-20 May | 70/60 knots | Track forecast errors were near to last season's average except at 48 hours lead time, which was higher. There was a slight slow and left of track bias. Skill scores against CLIPER were high, particularly at longer lead times. |
| Nakri (06W) | 27 May-03 June | 125/105 knots | The first few forecasts for this typhoon predicted recurvature too early, which resulted in huge errors at longer lead times. However, later forecasts were much better. This resulted in track errors near to or below last season's average at lead times of 72 hours or shorter. The model showed skill over CLIPER. |
| Fengshen (07W) | 18-26 June | 95/90 knots | Some early forecasts incorrectly predicted that Fengshen would turn northwards before reaching the Philippines. However, having crossed the eastern side of the islands the forecasts were much better. Most predicted landfall near to or just to the east of Hong Kong. Only long lead time track forecast errors were above last season's average. Skill scores were positive. |
| Kalmaegi (08W) | 14-20 July | 90/75 knots | Track forecast errors were just above and below last season's average at various lead times. Skill scores against CLIPER were positive. There was a right-of-track bias in forecasts overall. |
| Fung-wong (09W) | 24-29 July | 95/75 knots | Track forecast errors were below last season's average for this storm. Skill scores against CLIPER were positive. |
| Kammuri (10W) | 04-07 August | 50/50 knots | Kammuri was fairly short-lived, but track errors were small and skill scores high. |
| Phanfone (-) | 10-11 August | 25/40 knots | Phanfone was short-lived and no forecasts verified. |
| 11W | 13-14 August | 30/30 knots | There are no verification statistics or forecast plot for this tropical cyclone since its 1-minute averaged winds did not reach tropical storm strength. |
| Vongfong (12W) | 14-17 August | 50/40 knots | The track of this short-lived storm was well forecast. |
| Nuri (13W) | 17-22 August | 95/75 knots | Early forecasts stalled the storm and turned it towards Taiwan instead of showing landfall over China. Hence, longer lead time errors were large. Short period forecasts were good. |
| 14W | 26-28 August | 35/30 knots | No forecasts were verified for this storm. |
| Sinlaku (15W) | 08-20 September | 125/90 knots | Track forecast errors were low for this typhoon. The slow movement near Taiwan and the sharp recurvature were well predicted, although there was a slow bias in forecasts. Skill scores against CLIPER were high. |
| 16W | 10-11 September | 35/30 knots | 16W was briefly a tropical storm (according to JTWC) and just one forecast was verified. |
| 17W | 14 September | 30 knots | There are no verification statistics or forecast plot for this tropical cyclone since its 1-minute averaged winds did not reach tropical storm strength. |
| Hagupit (18W) | 18-24 September | 120/90 knots | Track errors were low for this storm and skill scores high. There was a slight slow and left-of-track bias. |
| Jangmi (19W) | 23-30 September | 145/115 knots | Early forecasts had a left-of-track bias, but later forecasts predicted recurvature well. Track forecast errors were large due to the early bias. |
| Mekkhala (20W) | 28-30 September | 55/50 knots | Mekkhala was briefly a tropical storm and no forecasts were verified. |
| Higos (21W) | 29 September-04 October | 45/40 knots | Track forecast errors were above last season's average for this storm. There was a slow bias in forecasts. |
| 22W | 14-15 October | 35/- knots | 22W was only briefly a tropical storm and no 24-hour forecasts were verified. |
| Bavi (23W) | 18-20 October | 45/45 knots | Bavi only briefly attained tropical storm status and just one forecast was verified. |
| Maysak (24W) | 06-10 November | 60/50 knots | Track forecast errors for this storm were a little below last season's average at 24-hours, but above the average at 48-hours. |
| Haishen (25W) | 15-16 November | 40/35 knots | Haishen was short-lived and just one 12-hour forecast was verified. |
| Noul (26W) | 16-17 November | 40/40 knots | Noul was only briefly a tropical storm and just one 12-hour forecast was verified. |
| Dolphin (27W) | 10-18 December | 90/65 knots | The recurvature of Dolphin was well predicted by the model, but there were errors in the rate of north-eastwards acceleration. Hence, track forecast errors were generally higher than last season's average, but skill scores against CLIPER were high. |
North-west Pacific observed tracks in 2008
North-west Pacific tropical cyclone names
| Name & ID | Start & end dates | 1-minute MSW | Comments on Met Office global model forecasts |
|---|---|---|---|
| Alma (01E) | 29-30 May | 55 knots | Alma only briefly reached tropical storm status and no forecasts were verified. |
| Boris (02E) | 27 June-04 July | 65 knots | Track forecast errors were below last season's average at all lead times except T+72. Skill scores were positive at T+24 and T+48. There was a left-of-track bias and the model showed an unusually low detection percentage indicating that it tended to dissipate the storm too soon. |
| Cristina (03E) | 27 June-01 July | 45 knots | Cristina was only briefly a tropical storm and just two forecasts were verified. Track errors for these forecasts were low and skill scores high. |
| Douglas (04E) | 02-04 July | 35 knots | Douglas was briefly a tropical storm and no forecasts were verified. |
| 05E | 05-07 July | 30 knots | There are no verification statistics or forecast plot for this tropical cyclone since its 1-minute averaged winds did not reach tropical storm strength. |
| Elida (06E) | 12-19 July | 90 knots | Like Boris, Elida was another storm with a poor detection percentage at longer lead times. However, track forecast errors were below last season's average and the model showed skill over CLIPER. There was a slight fast bias in forecasts. |
| Fausto (07E) | 16-22 July | 85 knots | Track forecast errors were mostly just above last season's average. There was a left-of-track bias in forecasts. Skill scores against CLIPER were positive. |
| Genevieve (08E) | 21-27 July | 65 knots | Track forecast errors were near to or below last season's average. Skill scores against CLIPER were modest and there was a left-of-track bias. |
| Hernan (09E) | 06-13 August | 105 knots | Track forecast errors were below last season's average for this storm, although skill scores against CLIPER were only positive at T+72. |
| Iselle (10E) | 13-17 August | 45 knots | Just two forecasts were verified for this storm. Track errors were above last season's average. |
| Julio (11E) | 23-26 August | 45 knots | Track forecast errors were low for this storm. |
| Karina (12E) | 02-03 September | 35 knots | Karina was only briefly a tropical storm and no 24-hour forecasts were verified. |
| Lowell (13E) | 07-12 September | 50 knots | Forecasts for Lowell were too fast and did not predict the eastward turn well. |
| Marie (14E) | 01-06 October | 70 knots | Track forecast errors were above last season's average and skill scores were negative at 48 and 72 hours lead times. This was primarily due to poor forecasts of the storms slow northwards drift part way through its life. |
| Norbert (15E) | 04-12 October | 115 knots | Early forecasts predicted a straight north-westward track, but later forecasts successfully predicted the gently curving track towards the Baja peninsula. Hence, track forecast errors were well below last season's average and skill scores against CLIPER were high. |
| Odile (16E) | 08-12 October | 55 knots | Track forecast errors were slightly above last season's average due to a slow bias. |
| 17E | 23-24 October | 30 knots | There is no forecast plot for this tropical cyclone since its 1-minute averaged winds did not reach tropical storm strength. |
| Polo (18E) | 02-05 November | 35 knots | Polo was a short-lived storm and just one forecast was verified. The track error of this forecast was small. |
| Name & ID | Start & end dates | 1-minute MSW | Comments on Met Office global model forecasts |
|---|---|---|---|
| Kika (01C) | 07-13 August | 35 knots | Track errors were a little above last season's average for this storm |
North-East and Central North Pacific observed tracks in 2008
North-East
Pacific tropical cyclone names
Central North
Pacific tropical cyclone names
| Name & ID | Start & end dates | 1-minute MSW | Comments on Met Office global model forecasts |
|---|---|---|---|
| Arthur (01L) | 31 May-02 June | 35 knots | Arthur was a short-lived storm and no forecasts were verified. |
| Bertha (02L) | 03-20 July | 105 knots | Bertha was the longest lived July Atlantic hurricane on record. Early forecasts had a left-of-track bias. The slow down and dip south-eastwards was generally well predicted, although the rate of acceleration north-eastwards was not. Hence, track errors at long lead times were large and there was a slow bias. At shorter lead times track errors were better and the model showed considerable skill over CLIPER. |
| Cristobal (03L) | 19-23 | 40 knots | The model gave good prediction of the storm's acceleration towards the north-east. Skill scores against CLIPER were high. |
| Dolly (04L) | 20-24 July | 85 knots | Track forecast errors for Hurricane Dolly were low. Skill scores against CLIPER were high. There was a slight left-of-track bias in forecasts. However, the model showed skill in predicting the deceleration of the hurricane as it approached landfall. |
| Edouard (05L) | 03-05 August | 55 knots | Track forecast errors were below last season's average at all lead times except T+72. Skill scores were positive at T+24 and T+48. There was a left-of-track bias and the model showed an unusually low detection percentage indicating that it tended to dissipate the storm too soon. |
| Fay (06L) | 15-24 August | 55 knots | Early forecasts failed to predict the double crossing of Florida, but later forecasts predicted the track well. Track errors were generally low. And skill scores high. |
| Gustav (07L) | 25 August-02 September | 130 knots | Track errors were above last season's average at lead times of T+72 and longer. However, skill scores were very high. The model predicted the south-westward turn towards Jamaica very well. However, later forecasts predicted too strong a ridge over the southern USA resulting in a track for Gustav which was too slow and too far to the west. |
| Hanna (08L) | 28 August-07 September | 70 knots | Track errors were above last season's low values, but not significantly so. A southward component in the early part of the track was forecast, but not to the extent that occurred. The acceleration to the north-west and north was fairly well handled. Skill scores against CLIPER were high. |
| Ike (09L) | 01-14 September | 125 knots | In common with other models a northwards turn near the Bahamas or Florida was initially predicted. However, once the south-westwards movement commenced, track forecasts were very good right through to landfall. Hence, track errors were low and skill scores high. There was a slight slow bias. |
| Josephine (10L) | 02-06 September | 55 knots | Forecasts for Josephine were generally too fast, although the model showed skill over CLIPER. |
| Kyle (11L) | 25-29 September | 65 knots | Track forecasts for Kyle were slow, thus predicting landfall too late. Skill scores against CLIPER were high. |
| Laura (12L) | 29 September-01 October | 50 knots | Just two forecasts were verified and track forecast errors were near to average. |
| Marco (13L) | 06-08 October | 55 knots | Marco was a short-lived storm and no forecasts were verified. |
| Nana (14L) | 12-14 October | 35 knots | Nana was only briefly a tropical storm and no forecasts were verified. |
| Omar (15L) | 13-18 October | 115 knots | The acceleration north-eastwards of Omar was correctly predicted by the model although the forward speed was too slow resulting in some large errors. However, the skill scores against CLIPER were very high. |
| 16L | 14-16 October | 30 knots | There is no forecast plot for this tropical cyclone since its 1-minute averaged winds did not reach tropical storm strength. |
| Paloma (17L) | 05-10 November | 125 knots | Forecasts for Paloma were too slow in tracking the hurricane north and north-eastwards towards Cuba. As a consequence track errors were larger than last season's average. However, skill scores against CLIPER were high. |
North Atlantic observed tracks in 2008
North Atlantic tropical cyclone names
| Name & ID | Start & end dates | 1-/10-minute MSW | Comments on Met Office global model forecasts |
|---|---|---|---|
| Nargis (01B) | 27 April-03 May | 115/90 knots | Forecasts for Cyclone Nargis were erratic in their quality. The 120-hour prediction of landfall was good, whereas the 96-hour forecast was poorer. This pattern was repeated for shorter lead time forecasts. Overall, track forecast errors were near to last season's average, although were large at the 5-day lead time. Skill scores against CLIPER were high. The model had a right of track bias, which meant that many forecast tracks took Nargis south of the Irrawaddy Delta of Burma. However, 24-36 hours ahead of landfall, the forecasts were good. |
| 02B | 16 September | 35/30 knots | 02B briefly attained tropical storm status (according to JTWC). No forecasts were verified. |
| 03B | 20-23 October | 35/30 knots | 03B briefly attained tropical storm status (according to JTWC) and no 24-hour forecasts were verified. |
| Rashmi (04B) | 26-27 October | 45/35 knots | Rashmi briefly attained tropical storm status and just one forecast was verified. This forecast predicted landfall well. |
| Khai-muk (05B) | 14-16 November | 45/35 knots | Track forecast errors were low for this storm and landfall fairly well predicted, although a little too far to the south. |
| Nisha (06B) | 25-27 November | 50/45 knots | The development of Nisha was well signalled by the model. Once it had formed forecasts up to 36 hours ahead were verified. Track errors were low. |
| 07B | 04-07 December | 35/30 knots | 07B briefly reached tropical storm status according to JTWC and no forecasts were verified. |
North Indian observed tracks in 2008
North Indian tropical cyclone names
Note on estimated wind speeds
1-minute MSW: Maximum sustained wind (knots) averaged over 1 minute (JTWC Hawaii, NHC Miami, CPHC Hawaii)
10-minute MSW: Maximum sustained wind (knots) averaged over 10
minutes (JMA Japan, IMD New Delhi)
Note: Although 1-minute average winds should be stronger than 10-minute
average, this may not be the case here since estimates are obtained
from different warning centres.
