After a quick start to the season with a major hurricane in May, Flossie became the third hurricane (category 2 on the Saffir-Simpson Hurricane Scale) to develop in the northeastern Pacific Ocean. It remained well offshore the west coast of Mexico and was only a threat to marine interests.
Flossie originated from a tropical wave that moved off the west coast of Africa on 11 August. This wave initially spawned Atlantic Tropical Storm Chantal on 14 August, about 1300 n mi east of the southern Windward Islands. Chantal and the subsequent wave envelope in which it was embedded continued to track rapidly westward at 20 to 25 kt across the tropical Atlantic and into the Caribbean Sea.
By 19 August, Chantal and the associated wave had slowed down significantly as the system moved into the western Caribbean Sea. Chantal eventually stalled along the east coast of the Yucatan Peninsula late on 20 August. While Chantal remained quasi-stationary near Chetumal, Mexico, the southern portion of the wave crossed over Central America and emerged over the northeastern Pacific Ocean south of Guatemala on 21 August, as indicated by surface observations and QuikSCAT satellite-derived winds.
For the next two days, a pre-Flossie disturbance of cloudiness and showers tracked westward parallel to the southwest coast of Mexico and changed little in overall organization. On 23 August, however, a closed surface circulation appeared in QuikSCAT imagery (not shown). The associated convection was not particularly well organized. The system tracked west-northwestward close to the Mexican coast until 25 August. Later that day, outer convective bands of the broad circulation moved onshore near Manzanillo, Mexico, and produced wind gusts to tropical storm force. Satellite classifications and QuikSCAT wind data, however, indicated that the circulation was still too broad and disorganized to be classified as a tropical depression.
By 0000 UTC 26 August, the pre-Flossie circulation had moved far enough away from the negative influence of the Mexican coastal mountains. The low-level circulation became better defined and deep convection consolidated closer to the system center. Satellite intensity estimates continued to increase and, by 0600 UTC, they suggested that the system had developed into Tropical Depression Seven-E about 235 n mi south-southeast of Baja California. The "best track" chart of the tropical cyclone's path is given in Figure 1, while the best track positions and intensities are listed in Table 1. Wind and pressure plots are shown in Figure 2 and Figure 3, respectively.
The depression tracked westward and convective organization continued to improve during the day. At 1800 UTC, the cyclone strengthened into Tropical Storm Flossie about 195 n mi south of the southern tip of Baja California. Flossie continued to move westward and passed about 7 n mi north of Socorro Island. By 1800 UTC 27 August, Flossie slowed from 9 kt down to 5 kt, and rapidly intensified into a hurricane. The intensification process leveled off when Flossie began to experience southwesterly shear while interacting with an upper-level low pressure system located to the west. Over the next 24 h, Flossie turned sharply southward and made a small cyclonic loop about 200 n mi west of Socorro Island. The sharp change in track motion was probably due to binary interaction between Flossie and the upper-level low that moved cyclonically around the south and east side of the hurricane before eventually weakening and lifting out to the northeast. Once the upper-level low and its negative shear effects moved away from Flossie, the hurricane began to intensify again late on 28 August and eventually peaked at 90 kt at 1800 UTC 29 August. Tropical Rainfall Measurement Mission (TRMM) microwave satellite imagery (Figure 4 and Figure 5) indicated that Flossie had become significantly better organized during an 18 h period with the eye decreasing to nearly half its original size. Eye contraction often corresponds with rapid intensification. However, an approaching mid-latitude trough began to induce southwesterly upper-level shear on Flossie, disrupting a possible rapid intensification process. Instead, Flossie began a slow weakening trend while tracking northwestward for the next day and a half and became a tropical storm again at 0600 UTC 31 August, when it was located about 470 miles west of the southern tip of Baja California. At 1800 UTC, the cyclone began moving over sharply lower sea-surface temperatures, which helped to bring about its rapid demise. Flossie weakened to a depression at 1200 UTC 1 September and completely dissipated later that day about 200 n mi west of Punta Eugenia in western Baja California.
The remnant low-level circulation drifted slowly westward over the northeastern Pacific Ocean as a swirl of low clouds for several more days. The mid- to upper-level circulation moved rapidly northeastward across northen Baja California before dissipating over the desert region of the southwestern United States.
Observations in Flossie (Figure 2 and Figure 3) include satellite-based Dvorak technique intensity estimates from the Tropical Analysis and Forecast Branch (TAFB), the Satellite Analysis Branch (SAB) and the U. S. Air Force Weather Agency (AFWA). Flossie's peak intensity of 90 kt at 1800 UTC on 29 August is based on a Dvorak satellite classification of 90 kt (T5.0) from the TAFB.
There were no ship or surface reports of sustained tropical storm force winds associated with Flossie. However, sustained winds of 30 kt with gusts to 40 kt were reported in squalls at Manzanillo, Mexico (MMZO) at 1645 UTC 25 August.
The remnant moisture from Flossie helped to trigger strong thunderstorms, deadly lightning, and flash floods across portions of Southern California. While there were no directly-related deaths or damage reported with Flossie when it was a tropical cyclone, there were deaths due to lightning strikes reported in southwestern California up to 36 h after Flossie had dissipated as a tropical cyclone. Four people were struck by lightning in the San Diego and San Bernardino Mountains on 2 and 3 September, and 2 of the 4 victims died. A 53-year old man was killed on 2 September when he was struck in the head by lightning while hiking and a 13-year old boy was also struck in the head and killed while standing in an open field.
In addition, more than 2 inches of rain fell in one hour and caused flash flooding in San Diego and Riverside Counties on 2-3 September. Strong downdraft winds also knocked down a tree onto a house. Total damage from Flossie's remnants was estimated at $35,000.
Average official track errors for Flossie were 30, 60, 94, 123, and 186 n mi for the 12, 24, 36, 48, and 72 h forecasts, respectively (Table 2). The number of cases ranged from 21 at 12 h to 11 at 72 h. These errors are lower than the average official track errors for the 10-year period 1991-2000 (37, 68, 99, 128, and 185 n mi, respectively), except at 72 h. While not included in the normal verification process, it is worth noting that forecast errors during all stages (including the tropical depression stage; number of cases indicated in parentheses), were slightly lower at 30 (26), 59 (24), 91 (22), 121 (20), and 182 (16) at the same time periods, respectively. Higher than average CLIPER errors, along with the best track plot (Figure 1), indicate that Flossie posed many forecast challenges. The NHC official forecast, however, correctly captured the small cyclonic loop and subsequent short-term southwesterly motion near Socorro Island on 28 August. In fact, some of the smallest official forecast (OFCL) errors occurred prior to and during this time of erratic motion, whereas all of the NHC model guidance produced large errors during this same period. The largest OFCL track errors occurred toward the end of Flossie's lifetime when the system was forecast to move slowly northwestward over colder water, while it continued on a northerly track. The best performing NHC track forecast guidance was the AVNI and GUNA models.
Average official intensity errors were 6, 8, 8, 10, and 19 kt for the 12, 24, 36, 48, and 72 h forecasts, respectively. This is slightly better than the average official intensity errors over the 10-yr period 1991-2000 (7, 11, 14, 16, and 20 kt, respectively). These errors were slightly better than the SHIPS intensity model forecasts.
There were no watches or warnings associated with Flossie.
|Lat. (°N)||Lon. (°W)|
|26 / 0600||19.1||108.5||1004||30||tropical depression|
|26 / 1200||19.7||109.7||1004||30||"|
|26 / 1800||20.0||110.6||1003||35||tropical storm|
|27 / 0000||20.1||111.6||1002||40||"|
|27 / 0600||20.2||112.5||1000||45||"|
|27 / 1200||20.1||113.3||994||55||"|
|27 / 1800||19.9||113.8||989||65||hurricane|
|28 / 0000||19.5||114.1||989||65||"|
|28 / 0600||19.2||114.3||989||65||"|
|28 / 1200||19.1||114.3||985||70||"|
|28 / 1800||19.4||114.7||985||70||"|
|29 / 0000||19.8||115.1||983||75||"|
|29 / 0600||20.1||115.4||980||80||"|
|29 / 1200||20.6||116.2||977||85||"|
|29 / 1800||20.9||116.6||972||90||"|
|30 / 0000||21.5||117.1||975||85||"|
|30 / 0600||21.9||117.7||979||80||"|
|30 / 1200||22.4||118.0||983||70||"|
|30 / 1800||22.9||118.1||987||65||"|
|31 / 0000||23.4||118.4||989||65||"|
|31 / 0600||23.6||118.6||994||60||tropical storm|
|31 / 1200||23.9||118.9||997||55||"|
|31 / 1800||24.3||119.2||999||50||"|
|01 / 0000||24.8||119.3||1002||40||"|
|01 / 0600||25.2||119.1||1005||35||"|
|01 / 1200||25.6||118.9||1005||30||tropical depression|
|01 / 1800||26.0||118.6||1006||30||"|
|02 / 0000||26.6||118.4||1007||25||"|
|02 / 0600||dissipated|
|29 / 1800||20.9||116.6||972||90||minimum pressure|
|Forecast Technique||Period (hours)|
|CLIP||34 (21)||78 (19)||127 (17)||175 (15)||262 (11)|
|GFDI||31 (21)||67 (19)||110 (17)||165 (15)||288 (11)|
|GFNI||43 (20)||84 (18)||128 (16)||172 (14)||292 (10)|
|LBAR||36 (21)||79 (19)||124 (17)||167 (15)||276 (11)|
|AVNI||33 (20)||68 (19)||102 (17)||128 (15)||141 (9)|
|BAMD||33 (21)||67 (19)||97 (17)||125 (15)||193 (11)|
|BAMM||35 (21)||78 (19)||134 (17)||197 (15)||305 (11)|
|BAMS||37 (21)||75 (19)||129 (17)||194 (15)||320 (11)|
|NGPI||48 (21)||95 (19)||132 (17)||173 (15)||268 (11)|
|UKMI||33 (20)||63 (19)||103 (17)||138 (15)||218 (11)|
|GUNS#||33 (20)||64 (19)||96 (17)||127 (15)||210 (11)|
|GUNA#||31 (20)||63 (19)||91 (17)||118 (15)||166 (9)|
|NHC Official||30 (21)||60 (19)||94 (17)||123 (15)||186 (11)|
|NHC Official (1991-2000 mean)||37 (2273)||68 (1835)||99 (1646)||128 (1475)||185 (1187)|
|*Output from these models was unavailable at time of forecast issuance.
#GUNS is the GFDI-UKMI-NGPI model ensemble average and GUNA is the GFDI-UKMI-NGPI-AVNI model ensemble average.
Figure 1: Best track for Hurricane Flossie, 26 August - 2 September 2001.
Figure 2: Best track maximum sustained surface wind speed curve for Hurricane Flossie, 26 August - 2 September 2001, and the observations on which the best track curve is based.
Figure 3: Best track minimum central pressure curve for Hurricane Flossie, 26 August - 2 September 2001, and the observations on which the best track curve is based.
Figure 4: Flossie near 75 kt intensity (0200 UTC 29 Aug 2001) with a partial eyewall indicated in the TRMM microwave channels (lower-left and lower-right panels), whereas no eye feature is indicated in conventional infrared imagery (upper-left and upper-right panels). Image is courtesy of the Naval Research Laboratory.
Figure 5: Hurricane Flossie at peak intensity of 90 kt (1831 UTC 29 Aug 2001). Note the small (less than 10 n mi) diameter eye in the 85 GHz composite data (lower-right panel). The eye is more distnict and is almost half the size it was 18 h earlier as seen in Figure 4. Image is courtesy of the Naval Research Laboratory.