What is an extreme flood and for whom?

Size: px
Start display at page:

Download "What is an extreme flood and for whom?"

Transcription

1 The Extremes of the Extremes: Extraordinary Floods (Proceedings of a symposium held al Reykjavik. Iceland. July 2000). I All's Publ. no What is an extreme flood and for whom? DAN LUNDQUIST Glommen's and Laagen's Water Management Association, PO Box 2903 Solli, N-0230 Oslo, Norway di(siglb,no Abstract Frequently the media report on extreme floods from all over the world. Many of these cause large economic damage to society and even loss of lives, while it is very seldom that they do not pose severe threats to economic assets. Some cover large areas with the gradual inundation of the land being the main issue; others influence only small areas, where intensities can be extremely high and erosion may be a major problem. The meaning of extreme is related to probability, damage potential and scale. These issues are discussed in the context of Noiwegian floods during the last 200 years. Key words floods; extremes; probability; flood intensity; economic damage; human influence; flood plains; design floods; hydropower regulation QUESTIONS What makes a flood extreme? Is it a large return period? Is it severe economic damage? Is it resulting changes to the landscape? Or does it include all of these factors? Is a flood that is reduced by human influence less extreme than it would have been otherwise? FLOODS IN AN HISTORIC PERSPECTIVE In July 1789 southern Norway experienced what would be remembered as "the flood of floods" partly because of severe weather conditions during "the Little Ice Age" and partly because of resulting landslides on the valley sides. Water levels and discharges on the River Glomma have never been so high since that event (see Fig. 1). The damage was considerable, many lives were lost, and families had to leave their farms because of the loss of the land upon which they made their livings. Some 178 years later in May and June 1967, the lower reach of the River Glomma experienced another large flood from a combination of rainfall and snowmelt. At this time the flow of the River Glomma was partly regulated for hydropower production. Nevertheless the damage was significant, especially in the more densely populated areas around Lake Oyeren (see Fig. 2) where the natural outlet acted as a bottleneck to the flood. In May and June 1995, only 28 years later, a third flood occurred on the River Glomma (see Fig. 3). It was the largest discharge flood on the main river since 1789, and without the effect of the hydropower reservoirs, it would probably have been even larger (GLB, 1996). The total damage was 230 million US dollars but could have been very much more had it not been for active flood dampening by use of reservoirs and manmade defences like dikes. Especially around the Lake Oyeren, the effect of the increased capacity of the outlet by excavations and the upgrading of a hydropower

2 368 Dan Lundquist of ^ * K is>? >S> >9> \v rs>»? >S> \" K» 3»?»v> K> & \' rs> >&> of? \" >S> of? & oc^ >&> c?? >S> of? ov.# >&> ly1 a 01 Fig. 1 Historical flood levels in the Lake 0yeren plant were very important. Without these improvements, flood stages would have been 2 m higher, which would have equalled those in Without the regulation for power production, the peak water levels would have been approximately 4 m higher. NORWAY Ml River Gaula Jostedalen Valley -fr-fw" \ll W Mount Fulufjell River Glomma - jj jt^j Lake 0yeren Fig. 2 Location map of Norway. Fig. 3 The 1995 Hood on the River Glomma.

3 What is an extreme flood and for whom? 369 Was the 1995 flood as extreme as the 1789 flood, assuming that the probability of the rainfall and meltwater contribution was of the same magnitude? Or was it less extreme because of the human intervention? The answer will depend on whether one considers the probability of natural processes or the magnitude of the economic damage. These three examples clearly show that the damage measured in economic tenus is very dependent on man's ability to protect his values, but also on the fact that industries and urban areas today are more commonly found in flood-prone areas than before when fanning was the main activity. On the other hand, insurance cover clearly weakens the incentive to reduce damage by reducing the economic risk when homes, industries, or infrastructure are established in flood-prone areas. All these floods on the River Glomma represented extremes for those involved, but on a relative scale they were small compared to the 1940 flood on the unregulated River Gaula. Even if the drainage basin of the River Gaula is only 10% of that of the River Glomma's, this flood reached the same magnitude of discharge as the last two floods on the River Glomma. This was a purely rain-fed flood with a duration of a few days. The recurrence interval of the discharge is estimated to be about 300 years. When compared to other observed floods, this was a true extreme. In 1979 another flood event on the Norwegian west coast hit the Jostedalen Valley. This was a very dramatic flood, and many smaller tributaries changed course, resulting in erosion and sedimentation from and on the land surface. Roads and bridges were washed away, and the inhabitants were isolated for many days except for transport by helicopter. In 1987 a large part of southeastern Norway was struck by the remnants of an Atlantic cyclone. Even if precipitation intensities were not very high (30-60 mm in 24 h), the effect in smaller river basins was considerable due to extremely wet weather conditions during the previous month. Precipitation with a recurrence interval of years resulted in floods with recurrence intervals of years. Flood abatement was made difficult by the large number of tree trunks that blocked the roads and by seawater levels 1-2 m above normal. Fig. 4 Remains of the flood channel from the 1997 flood at Mount Fulufjell.

4 370 Dan Limdqiiist In August 1997 very severe weather conditions struck the local mountain range Fulufjell on the border between Norway and Sweden. It is estimated that close to 400 mm of precipitation may have fallen within 24 h (Vedin et al, 1999). No people were directly affected, and the economic consequences were mostly related to the loss of forest by erosion and local landslides. The changes to the landscape however, were dramatic, probably larger than at any other single incident since the last Ice Age. Some trees were removed from a basin of only 22 km 2, and the flood intensities were estimated to be about s" 1 1cm" 2 (see Fig. 4). Consequently, this was a true extreme with respect to the processes forming the landscape, but not when considering the economic loss. Memories are short. We do not remember extremes for very long times. How seldom were the floods in the past that we still remember? How often can we expect floods of this magnitude to return? Can floods be even larger? FLOODS AND SCALE For the River Glomma, design floods have been calculated according to the guidelines for dam construction. The design criterion stated by the Norwegian authorities is usually the 1000-year flood and for additional safety considerations the probable maximum flood (PMF). Results from a large number of calculations for the River Glomma are shown in Fig. 5(a). We can clearly see that floods are restricted to relatively low intensities in basins larger than 5000 km", while they increase radically for basins less than km". The smallest basins represented in these data have drainage areas of only a few km" with specific 1000-year peak discharges of s" 1 1cm" 2. The point is that extreme flood intensities at one scale can be normal or even small at other scales. How do these calculated values compare with large floods observed on the River Glomma? Some observed data from the 1995 flood are plotted in Fig. 5(b). We note that all observations are smaller than the calculated 1000-year flood values, except for some sub-basins with quite small margins. What about extremes in other basins? In Fig. 5(c) some observed and estimated values from the River Gaula 1940, the Jostedalen Valley 1979 and Mount Fulufjell (3) Glomma calculated q1000 Glomma observed 1995 (c) Other basins p Fulufjell O Calc. Maxq1000i Basin (km 2 ) h Fulufjell Jostedalen V B Gaula Other -Max q1000 km 2 Fig. 5 Flood intensities and scale.

5 What is an extreme flood and for whom? , are plotted. It is obvious that other basins have had significantly larger flood intensities than experienced in the River Glomma. It should also be pointed out that PMF values for the River Glomma can reach s" 1 km" 2 in small sub-basins. What is the probability of intensive rainfall over large areas simultaneously? How large is the probable limit of flood intensities from large basins? How large can a specific flood become in a small basin? Can it be larger than the probable maximum flood? Is there a limit? Physically there is a limit. But there are indications that present methods of calculation still do not have the proper accuracy, especially when considering the influence of scale. THE INTERNATIONAL SETTING How extreme are Norwegian floods in an international perspective? The 1995 flood on the River Glomma becomes quite insignificant compared to floods like the 1998 flood on the Yangtze River in China, affecting 223 million people, killing 3000 and with an estimated economic damage of more than million US dollars. When considering loss of life, the 1995 flood on the River Glomma, with only one person drowned, becomes even less significant compared to the 1999 flood in Venezuela, where some people may have lost their lives. Not even in a Nordic context can Norwegian floods be regarded as extreme. During the jôkulhlaup in Iceland in 1996 discharge is estimated to have reached m 3 s" 1, compared to some 4000 m J s" 1 on the River Glomma in In other words, Norway does not seem to be the land of extreme floods from an international perspective. IS THERE A CONCLUSION? This is very much a question of whether the use of an absolute scale or a relative one is appropriate. A proper definition of the word extreme in this context might be "far from the normal situation in a particular area". This implies that a Norwegian flood can be extreme in a Norwegian context, but not necessarily for Chinese or Icelandic ones. The conclusion therefore could be that an extreme flood is an extraordinary flood with severe consequences for man or nature. In small Norwegian basins, flood intensities without doubt can be physically extreme and dramatic. For larger basins however, the main problem is associated with the extreme economic consequences to settlements and infrastructure along the river channel. REFERENCES GLB (1996) The 1995 Flood in the Glomma and Lâgen River Basins. Glommen's and Laagen's Water Management Association, Oslo, Norway. Vedin, H., Eklund, A. & Alexandersson, H. (1999) The rainstorm and flash flood at Mount Fulufjâllet in August 1997: the meteorological and hydrological situation. Geogr. Ann. 81A(3),