Hurricane Florence - Bermuda Weather Service



Hurricane Florence

Florence began it’s life cycle as an African Easterly Wave near the Cape Verde islands. The US National Hurricane Center started issuing advisories on it as Tropical Depression 6 at 21 UTC on Sunday September 3, 2006. Even at that early stage, the steering flow was set up in such a way that forecasters at the Bermuda Weather Service begun to pay close attention to the system. As predicted by NHC, TD#6 progressed on a generally westward track and was upgraded to a tropical storm on September 5. On Wednesday the 6th, the extended track forecast began to indicate signs that Florence would begin to recurve towards the Bermuda area over the coming weekend. All eyes were on Florence, with the recent 3-year anniversary of the landfall of Category 3 Hurricane Fabian very much at the forefront of public attention.

[pic]

Figure 1: Surface analysis valid 1800 UTC Wednesday September 6, 2006. Source: Bermuda Weather Service

On the 8th, Florence, still a tropical storm, was located over 700 nautical miles to the southeast of Bermuda, but already preparations were being made. Computer model tracks were tightly clustered (Figure 2), conveying a high level of confidence, and this was born out by the consistency of the forecast track. Emergency preparedness meetings had already been arranged in light of the early lead-time the forecast provided and the high confidence that Florence would threaten the island early in the next week. Tropical Storm Florence was at that time forecast to affect Bermuda as a Category 2 hurricane late night on Sunday the 10th. A Hurricane Watch was issued by the Bermuda Weather Service at 18 UTC on Friday the 8th, in order to alert the public of the possibility of hurricane force winds affecting the local area within 36 hours.

[pic]

Figure 2: Computer model tracks of the centre of Hurricane Florence. Source: Colorado State University

Florence had a very broad wind field, with tropical storm force extending out to 250 nm from the centre of circulation. This is illustrated by the QuikSCAT imagery in Figure 3. A Tropical Storm Warning was posted at 15 UTC on Saturday the 9th (augmenting the Hurricane Watch already in place) to warn of tropical storm force winds imminent within 24 hours, despite the closest point of approach of the eye being approximately 48 hours away.

[pic]

Figure 3: QuikSCAT imagery depicting Tropical Storm Florence on September 10, prior to making landfall over Bermuda. Source: NOAA/NESDIS

In the early hours of Sunday September 10th Florence was upgraded to a Category 1 Hurricane by NHC and the Bermuda Weather Service issued a Hurricane Warning (replacing the previous watches and warnings). Tropical storm force gusts (>34 knots) began to affect the marine area and a large area of convection associated with one of the outer rainbands moved over the island, with embedded severe weather (see Figure 4 for satellite imagery). Figure 5 shows a reflectivity image of a hook echo which moved over the western parishes, causing at least one tornado. Figure 6 shows the mesoscale circulation at an altitude approximately 1000m, clearly visible in Figure 5 as a velocity diplole to the northwest of the island, downstream from the tornado’s earlier position. Property damage occurring as a result of this system supports the assertion that this was a tornado; photographs of the damage are shown in Figure 7. It is suspected that upslope enhancement of this extremely localized severe wind event occurred, as this property and the other one investigated are both facing the southeastern shore.

[pic]

Figure 4: Infrared (channel 4) GOES satellite imagery from 1145 UTC on Sunday September 10, 2006. Source: NOAA.

[pic]

Figure 5: Bermuda International Airport Doppler Radar Surface Rainfall Index product (based on Reflectivity) at 1657 UTC September 10, 2006.

[pic]

Figure 6: Bermuda International Airport Doppler Radar PPI Velocity product at 1748 UTC September 10, 2006.

[pic] [pic]

[pic] [pic]

Figure 7: Digital Photographs taken at a Southampton Parish residence, suspected to have suffered tornado damage. Interior doors are reported to have blown inwards and roof slate damage is visibly evident.

Upon clearance of this rainband, the sun was visible for most of the afternoon. Easterly winds continued to increase to tropical storm force sustained, and the automated weather observation systems located around the island reported that this threshold had been reach overnight, at around 0630 UTC on the 11th. Winds continued to steadily increase and veer southeasterly, reaching a 10-minute sustained peak wind of 71 knots with gusts to 97 knots at 1340 UTC, according to the AWOS at St. David’s (located in the eastern end of the island). The closest point of approach of Florence is estimated to have been at this time, the center being approximately 50 nautical miles to the west of the Bermuda Weather Service, also in the eastern end. Florence was never actually upgraded to greater than a Category 1 storm, and was estimated to have had sustained winds of 80 knots associated with its wind field. The total rainfall amount recorded during the passage of Florence (including the rainband on the 10th) was 33.5mm (1.32 inches).

Due to electrical problems at the tower site, the Doppler weather radar lost data between approximately 1030 UTC and 1730 UTC during the passage of the height of the storm. Nonetheless, the data provided from the radar proved valuable in fixing the position of the eye (see Figure 8). Much to the credit of the Meteorological Technicians on duty at BWS, the radiosonde ascent schedule remained uninterrupted, and balloons were successfully launched on time at 00 UTC and 12 UTC even in adverse conditions. The 12 UTC ascent only got as far aloft as 400mb, but the resulting data was nonetheless useful to both the NHC and BWS forecasters.

[pic]

[pic]

Figure 8: a) Final radar image (SRI product) prior to data outage and b) first image after restoration of radar products.

There was frequent & widespread overseas media coverage from AP, Caribbean Newswire, CNN, The Weather Channel (who sent Jim Cantore), AccuWeather and U.S. TV network affiliates as far away as Boise, Idaho in addition to usual phone calls. The increased media attention was no doubt due to the lack of Atlantic storm activity affecting any other landmasses at the time. Thanks are due to Brian McNoldy of Colorado State University for putting BWS radar imagery on his website. This assisted in reducing the number of hits on our website, for which there is limited bandwidth. We recorded over 40,000 hits on the BWS website weather.bm leading up to and during the storm, and it began to have an impact on our ability to update the tropical information on it in a timely manner. Options for mirroring the site are currently being investigated.

Local media reported 25,000 customers out of power, ½ of which were restored within 24 hours. The remainder were restored within days. Roof damage (previously mentioned) occurred in the western parishes during passage of an outer rainband. Minor damage was inflicted on the Causeway, due to erosion near the water level; 50ft of road was restricted to 1-way traffic for a few days while repairs were made. There was some damage to boats reported, but that is estimated to be less than ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download