Media centre world map
Media centre > Special features > Billis and Saomai

Typhoons Bilis and Saomai: why the impacts were so severe

by Duan Yihong

Typhoon Bilis and the subsequent depression as it weakened, from 14 to 18 July 2006, brought heavy-to-excessive rain, strong local winds and lightning to six Chinese provinces (Fujian, Zhejiang, Jiangxi, Hunan, Guangdong and Guangxi). Some localities measured an accumulated rainfall amount of 300-500 mm. The rain triggered floods, landslides, rock- and mudflows, causing heavy loss of life and property. Bilis was unusual in terms of lifetime over land and impacts. Similarly, on 10 August 2006, supertyphoon Saomai made landfall on the coast of Cangnan County, Zhejiang Province, claiming many casualties and causing economic losses there and in parts of Fujian Province.

Why were the impacts of Bilis and Saomai so severe? Have there been other instances in history? What is the current skill in forecasting such typhoons? In order to answer these questions, some objective and scientific analyses should be made.


Bilis moved slowly after landfall and persisted longer with prolonged rainfall

After Bilis formed on 9 July, it moved north-west, made its first landfall over Taiwan, Province of China, traversed the Taiwan Straight and eventually landed in Fujian Province. Bilis’s movement was normal over the sea. What was unusual was that, despite being degraded to a tropical depression after landing, Bilis continued to move north at a very slow pace from 14 to 18 July, under the influence of the continental high to the north, with its impacts extending for up to five days.

Historical analysis indicates that, normally, typhoons, after landing, persist for about 26 hours on average. Of these typhoons, 59.3 per cent disappear within 24 hours; 28 per cent survive for no longer than two days; and only 12.7 per cent stayed for more than two days. The previous record was made by Typhoon No. 7 in 1952, whose lifetime was 129 hours (>five days) and Typhoons Nos. 4 and 3 in 1975, whose lifetimes were 119 and 101 hours, respectively. Compared with these three typhoons which moved north-west or north after landing, Bilis moved south-west and persisted for 120 hours over land, hence breaking the record in this sense.

Bilis was unusual for its unsymmetrical structure and its southern cloud band which produced heavy rain

Since Bilis formed, its intensity did not increase substantially. It differed from the strong tropical cyclones which landed in east China in 2005, being weaker by the order of several digits. Seemingly weak typhoons give the general public the wrong idea. They believe that the typhoon’s impacts go hand-in-hand with its intensity, i.e. the stronger it is, the stronger will both wind and rainfall be. In fact, this perception is rarely justified. The impacts have more to do with the surrounding weather systems at the time of landfall.

During its movement over the sea, only the southern portion of Bilis’s cloud system was seen to be active on the satellite imagery. It looked like half a typhoon, some even believed that it was one without an eye. This evident unsymmetrical distribution of cloud bands could explain why its intensity (which is normally measured by its lowest air pressure and maximum wind velocity in the vicinity of its centre) remained less strong. This feature was true for Bilis after landing, in particular its cloud system then developed instead of dissipating, indicating a large supply of moisture which brought about even more rainfall than at the time of landing.

Historically rare: unlike others whose rainfall was due to encountering cold air, Bilis’s rainfall came from the mesoscale systems triggered under the interaction with the South China Sea Monsoon after landfall.

Normally, if a typhoon moves west, north-west or north after landfall, a huge amount of moisture will be taken with it. When this moisture meets cold air coming from the north, heavy rain occurs. In China, a number of severe events induced by typhoons have occurred:

After Typhoon No. 7503 landed over Zhejiang in August 1975, it continued to move north-west, when the moisture it contained encountered the cold air from the north and the special topography channelled the wet airmass to the Zhumadian region, Henan Province, bringing three days of torrential rain to hilly areas: 1 062 mm in 24 hours, a record for the Chinese mainland.

Having made landfall over Fuqing City, Fujian Province, around midday on 1 August 1996, Typhoon No. 9608 continued to move north-west-north. As it moved, it met cold air from the north over the North China Sea, causing extensive heavy rain over central south and southern parts of north-east China. Shijiazhuang, Pingshan, Jingjing and Yuanshi in Hebei Province received 303, 324, 311 and 299 mm of rain, respectively, from 3 to 5 August, breaking local records and causing flooding along the Hai River basin. Again, the rainfall was produced by the same mechanism.

After Typhoon Talim (No. 0513) landed over Putian city, Fujian Province, it moved north-west and met cold air from the north just over northern Jiangxi Province and south-western Anhui Province, giving rise to heavy or excessive rainfall over the region. At Lu Mountain and Yuexi, 940 and 573 mm of rainfall were observed, respectively, both beating the records since observations began. Such rainfall was a combination of the remaining cloud moisture carried by the typhoon and the cold air, plus some complicated underlying topographic effects.

Bilis, however, was extremely unusual owing to its interaction with the South China Sea (SCS) monsoon. This mechanism can be easily identified from satellite imagery. After landing, Bilis was no longer a source of moisture but dynamic source that created a rising environment for the rainfall systems. The SCS monsoon was the moisture provider. Due to the interaction of the two weather systems, the primary mesoscale rainfall subsystems generated and developed over Hunan, Guangdong, Guangxi and Yunnan Provinces. This was a completely different rain-producing mechanism than that of the three aforementioned cases.

Disaster-prone: before landfall, rainfall over the area had already been slightly more frequent and water saturation tends to reduce the water retention capacity of the soil.
From 1 to 13 July 2006, the mid-eastern parts to the south of the Yangtze River (i.e. most parts of Zhejiang, Jiangxi and Hubei Provinces) experienced precipitation of 10-50 mm, which was slightly less than normal by 2–5 per cent. Bilis brought enough rain to mitigate possible drought in these localities. On the other hand, the western part of South China (e.g. most parts of Guangxi) to the south of the Yangtze River (e.g. most of Hunan) and the eastern part of Yunnan Province had already received more than 30 per cent of precipitation than normal. The rain which persisted for 5–9 days—or heavy rain for 1–2 days in some localities—would have further reduced the water-retention capacity, thus causing serious waterlogging in farmland (see Figure 1). In this context, such regions had became highly flood-prone. This was another important factor behind the floods, landslides, mud-/rockflows and other weather-induced disasters in the southern part of Hunan Province.

Fgure 1 Figure 1 — Nationwide rainfall pattern, 1-13 July 2006



Extremely high intensity: the excessively strong winds at the time of landfall were sustained for a long period.

Saomai suddenly intensified about 36 hours before landfall: it was an unprecedented supertyphoon along the Zhejiang-Fujian coast.

Figure 2 is an image from the Chinese FY-1D satellite. The clear-cut eye of the typhoon illustrates its intensity.

Figure 2 Figure 2 — FY-1D satellite image of Saomai at 07:00 (Beijing time) on 10 August 2006

Based on the observed changes of intensities (Figure 3), it can be noted that, from 02:00 to 20:00 (Beijing time), on 9 August 2006, the central air pressure of Saomai was lowered by 50hPa, which far exceeded the standard intensity threshold (1.25 hPa per hour) of an abruptly intensifying typhoon, reaching up to 2.7 hPa per hour. A typhoon that strengthens suddenly is extremely rare over China’s coastal waters. Saomai was even much stronger than typhoon No. 5612 that had a death toll of 5 000. It had a central pressure of 925 hPa and a wind force of 55 m/s at the time of landing in Xiangshan, Zhejiang Province on 1 August 1956. From this perspective, Saomai was the strongest typhoon that ever occurred over China’s offshore region.

Figure 3

Figure 3 — Typhoon Saomai intensity changes

Saomai’s wind speed was extremely high and strong wind persisted for a long time after landfall: the strong wind was the main cause of disaster.

The velocity was extraordinarily high when supertyphoon Saomai made landfall. The observed maximum wind speeds in both Zhejiang and Fujian Provinces broke all records. In the landing process, Nanxiaguan station (Chang’nan county, Zhejiang Province) and Hezhangyan station (700 m a.s.l.) and Fuding station (Fujian Province) observed winds of 68 m/s and 75.8 m/s respectively (just below the existing typhoon record of 78.9 m/s observed at Dalaoshan, Hong Kong, on 1 September 1962).

The coastal area of south-east Zhejiang Province and part of the coastal area in north-east Fujian Province experienced strong winds varying from Beaufort Force 11 to 12, and even Force 14 to 17 in some localities. As many as 10 weather stations in Zhejjiang and Fujian provinces recorded wind exceeding Force 12. Wind gusts from 17:00 to 20:00 in Fuding city, on 10 August was over 40 m/s for three consecutive hours and caused severe damage.

The most powerful typhoon ever to make landfall over mainland China, it was a very unusual typhoon, which occurs possibly once every 100 years.

The minimum central pressure of Saomai reached 915 hPa and retained this value until it was about 80 km away from the coast. At landfall, Saomai was still a supertyphoon with a maximum wind force of Force 17 (60 m/s) and central air pressure of 920 hPa. While several typhoons have made landfall on mainland China at speeds of 45-50 m/s (see Table 1), only Saomai (No. 0608) in 2006 and Marge, which landed on Qionghai, Hainan Province, on 14 September 1973, had winds of 60 m/s at the centre. Saomai’s central air pressure at landfall, however, was 5 hPa lower than that of Marge. It may be considered, therefore, that Saomai was the most powerful typhoon ever to have made landfall over mainland China.

Table 1 — Typhoons making landfall on mainland China with wind speed > 45 m/s

Typhoon no./Name

Place of landing

Time of landing

Maximum central wind speed at landing (m/s)

Central air pressure at landing (hPa)

0608 Saomai

Changnan, Zhejiang




0518 Damrey

Wanning, Hainan




0515 Khanun

Taizhou, Zhejiang




0509 Matsa

Yuhuan, Zhejiang




0414 Rananim

Wenling, Zhejiang




9615 Sally

Wuchuan-Zhanjiang, Guangdong




9111 Fred

Xuwen, Guangdong




8510 Mamie

Changle, Fujiang




8105 Kelly

Yaxian (Sanya), Hainan




8015 Percy

Zhangpu, Fujian




7314 Marge

Qionghai, Hainan




6903 Viola

Huilai, Guangdong




6614 Alice

Luoyuan, Fujian




6608 Ora

Haikang-Xuwen, Guangdong




5822 Grace

Fuding, Fujian




5612 Wanda

Xiangshan, Zhejiang




5413 Ida

Zhanjiang-Haikang, Guangdong




5310 Nina

Leqing, Zhejiang





According to the statistics, a total of 686 tropical cyclones landed on China (including Taiwan, Province of China) from 1949 to 2005 (including those which landed on Taiwan first and then on mainland China, and those which made landfall on the mainland more than once). Of these, only Marge had a wind speed at landing of 60 m/s (an estimated value afterwards; the observed value was 42 m/s) and a central air pressure of 925 hPa. According to the cumulative frequency distribution of wind speeds, therefore, the frequency being at 56 m/s is only 99.85 per cent. Taking all this into consideration, Saomai has been registered as a type of typhoon that tends to occur once every 100 years.

Saomai’s intensity at landfall was stronger than that of hurricane Katrina in the USA in 2005.

Hurricane Katrina killed more than 1 833 people and caused more than US$ 81 billion of property losses: the heaviest weather-related losses in US history. Katrina landed on Louisiana on the morning of 29 August 2005 and seriously affected the southern regions of the USA. Maximum central wind speed at landing was 57 m/s (60 m/s for Saomai). The air pressure in the centre was 920 hPa for both.

Figure 4

Figure 4 — FY-1D Satellite image of typhoon Saomai at 07: 00 on 10 August 2006

Figure 5

Figure 5 — NOAA-16 satellite image of hurricane Katrina at 04: 00 on 29 August 2005

Based on a comparative analysis of satellite and radar images before the landings of both Saomai and Katrina, a clear-cut typhoon eye could be seen from the FY-1D satellite image of Saomai (see Figure 4). A clear hurricane eye was also seen from the US NOAA-16 satellite image of Katrina (see Figure 5), but the boundary of its eye zone became slightly blurred, indicating that Katrina was not as intense as Saomai before landfall. From radar echo images (see Figures 6 and 7), it an be noted that the intensity of Katrina at landfall was less than that of Saomai. There was a small but clearer eye when Saomai landed and the strong echoes around it were distributed symmetrically. When Katrina was landing, its eye became somewhat distorted due to topographical effects and the eye zone was less clear. Moreover, no strong echoes were detected, partially because of the southern cloud bands. This demonstrates that Saomai was stronger than Katrina before and after landing.
Saomai’s sphere was not large, but its winds were destructive. In Xiaguan town, Cangnan County, Zhejiang Province, winds reached 68 m/s; while the maximum wind speed observed at Hezhangyan, Fuding city, Fujian Province, at 17:10 (Beijing time) on 10 August 2006 was 75.8 m/s, breaking all records in both provinces. Moreover, the Force 17 wind covered a radius of 45 km before and after landing. The observed extreme velocity when hurricane Katrina hit southern regions of the USA was 60 m/s.

Figure 6

Figure 6 — Radar echo image at 17:00, 10 August 2006, in Wenzhou, Zhejiang

Figure 7

Figure 7 — US radar image of Hurricane Katrina at 20:00, 29 August 2005


Concluding remarks

The Central Meteorological Observ­atory (CMO) at CMA attaches great importance to typhoon monitoring, forecasting, early warning and related services. CMO made qualitative analyses and prognoses during tropical storm Bilis concerning heavy rainfall and duration, which proved to be fairly accurate. Nevertheless, the following questions will always need answers:

  • What will be the specific amount of precipitation?
  • At what specific time and place will heavy rainfall take place?
  • Where will the disasters occur? Where will the flood and water­logging occur

Based on current science and technology, it is still quite impossible, in the short-term (1–3 day) weather forecast, to give highly detailed and precise rainfall quantities at any specific time and place, which are often urgently required by the national disaster prevention and mitigation authorities. However, through satellite, radar and other modern monitoring technology, we are in a position to track closely and understand better the behaviour of moisture-carrying cloud bands of a typhoon, to analyse the evolving features of short-term heavy rainfall and to make improved nowcasts 0-3 hours and even 0-6 hours ahead. It is now possible to seize opportunities in making forecasts and early warnings, so as to give longer lead-times and identify the locations which might be affected to the decision-makers in disaster prevention and mitigation.



© World Meteorological Organization, 7bis, avenue de la Paix, Case postale No. 2300, CH-1211 Geneva 2, Switzerland - Tel.: +41(0)22 730 81 11 - Fax: +41(0)22 730 81 81 Contact us Copyright | Privacy | Scams | Disclaimer |  Guidelines | Procurement | UN System | Accessibility