Runoff and Floods in the Alps An Overview

BALDASSARE BACCHI1 AND VIGILIO VILLI2 1 Department of Civil Engineering-University of Brescia - Via Branze, 38-25123 Brescia-I, 2 CNR - Research Institute for Hydrogeological Risk Prevention - Corso Stati Uniti, 4-35127Padova-I

14.1 INTRODUCTION

In this overview, the close links between meteorological and hydrological aspects in mountain areas are briefly discussed and state-of-the-art issues related to the influence of meteorological and surface processes on flood formation are presented. Although this paper is focused on the European Alps, most considerations could be applied to mountain areas in general.

As pointed out recently by Weingarten et al. (2003), mountain floods are special since they are the result of a combination of factors unique to mountain basins such as high-intensity precipitation, steep gradients and thin soils but nevertheless ''despite much research, the production of storm runoff in responsive catchment is still poorly understand''. Consequently, when coupling meteorological and hydrological models, the effect of orography and geomorphology on floods should be considered particularly closely.

14.2 PRECIPITATION AND RUNOFF FORMATION

It is well known that heavy precipitation in the Alps is reproduced when polar and North Atlantic air masses, characterised by low temperatures, interact with warm and moist air masses of oceanic or Mediterranean origin. The resulting fronts vary in size from several hundreds to more than a 1000 km and tend to move from West to East with a mean lifespan of 2-7 days. Knowledge of particular aspects of mountain precipitation was intensified under the Mesoscale Alpine Programme (Bougeault etal. 2003) where dynamical and microphysical processes leading to the enhancement and organisation of precipitation in alpine mountain ranges were investigated by experimental and numerical models. Summarising some of the results of the project, it appears that the effect of arc-shaped alpine relief on air movement is connected to two relevant atmospheric phenomena: the deviation of precipitation fronts in almost parallel orientation with the mountain chains (Rotunno and Ferretti 2001) and the forced convection of air masses towards the water divide. These phenomena can produce persistent and spatially diffuse rainfall contributing to areas with the highest annual precipitation and frequency of thunderstorms located along the border of the Alps. In contrast, rapid decrease of atmospheric moisture with altitude combined with the rain shadow effects in the leeside valleys of mountain ridges cause east-west-oriented inner alpine valleys to gain considerably less mean annual precipitation (Alpert 1986; Buzzi and Foschini 2000; Van Delden 2001; Frei and Schar 1998). Finally, in the summer, convection produces an abrupt cooling of warm air masses, causing frequent and localised thunderstorms particularly in the headwater areas dominated by steep slopes and thin soils.

It is important to note that in alpine catchments, meteorological, hydrological and glaciological phenomena undergo complex and highly variable interactions over short spatial and temporal scales. However, effects can be

Climate and Hydrology in Mountain Areas. Edited by C. de Jong, D. Collins and R. Ranzi © 2005 John Wiley & Sons, Ltd perpetuated over long distances and time. Alpine slopes host many different environments: extreme mountain peaks with prevailing glacial and periglacial phenomena, lower pasture and forest zones, alluvial terraces marking the former flood plain and finally the valley bottoms bordering the river channels as relatively smooth strips of land. Along this ''section'', characterised by steep slopes and creeks, the dynamic equilibrium between form and process is continuously modified by the magnitude and frequency of hydrometeorological processes (rainfall, evapotranspiration, ice and snow melt, etc). The most significant evidence for these interactions is given by alluvial fans, which are generated by debris flows that originate from the steep tributaries of the main valleys. These impulsive phenomena together with landslides modify the landscape in combination with other slower phenomena. In the literature is still deeply debated whether alterations in surface flow, erosion and flooding are due to changing land use as a result of modified agricultural practices, deforestation and the increase of impermeable areas as a consequence of urbanisation (Ranzi et al. 2002, Brath et al. 2002).

By linking the hydrometeorological space-time domain with the characteristics of alpine valleys, the following spatial and temporal scales of hydrological processes can be identified in alpine regions.

At microscales of some square kilometers, local atmospheric instabilities prevail in the form of isolated storm cells triggered by orographic lifting. Here, the main processes controlling runoff formation are affected by the geomorphological and topographical characteristics of the basins, rainfall conditions, snowfall and ablation, soil moisture conditions, vegetation cover, land use, and so on.

At the mesoscale, widespread stratiform precipitation is induced by uplift of large-scale air masses with high equivalent potential temperatures, causing severe floods in basins of some thousands of square kilometres. Flood intensities reflect both meteorological and physiographic factors as well as the routing properties of the channels.

Floods in larger basins such as those that occurred in central and eastern Europe in 2002 are forced by synoptic scale precipitation from a combination of climatic events caused both by anomalies in upper-level circulations and sea-surface temperatures as well as other unusually intensive hydrometeorological interactions.

Following the discussion on runoff processes in alpine areas, a brief excursion will be taken into types of models used in reconstructing and/or predicting catchment response.

Infiltration is simulated according to two basic assumptions. The first one is based on the ''Hortonian''

concept of rainfall exceeding the infiltration capacity of soils with low permeability or low depth. This identifies one of the main mechanisms by which rainfall can reach drainage networks via overland flow. Many models are based on the concept of estimation of infiltration excess rate (Green and Ampt 1911; Horton 1933, 1940). Since this conceptual scheme became inappropriate in some basins, especially in areas with gentle slopes or those characterised by greater infiltration capacities, another mechanism was proposed by Dunne et al. (1975) on the basis of the Variable Source Area concept, which was first defined by Hewlett (1961). This model assumes that catchment outflow is mainly due to overland flow that originates from saturated areas close to streams. Here, the role of topography is crucial, especially that of hollows in the development of saturation. This conceptual scheme has formed the basis of many runoff generation models: the Topmodel (Beven and Kirkby 1979) and the Probability Distributed Model (Moore and Clarke 1981).

The ice-snow melting phase is often estimated by using a simplified conceptual scheme based on air temperature. This so-called ''degree-day'' method has been used in many different ways for more than 60 years. When a complete set of meteorological data is available (radiation, wind speed, air humidity and temperature, precipitation), models based on the energy-balance equation are applied. Advantages and disadvantages of distributed snow-melt models are discussed by Kirnbauer etal. (1994). Nevertheless, the empirical degree-day method cannot be easily replaced by more physically based methods (Martinec and Rango 1995) because the required meteorological data is often missing.

Surface runoff, especially channel flow processes, is related to the hydraulic properties ofthe channel network that governs flow propagation. The kinematic formula and linear reservoir methods are probably the oldest methods available to determine the hydrologic response of drainage network. The related IUH concept was introduced by Sherman (1932) and was later applied mainly to hydrological studies. In the 1950s, a conceptual model that considers the drainage basin as a cascade of linear reservoirs was introduced (Nash 1957). This model is probably most widely adopted in hydrological practice. By successively linking hydrology with quantitative geo-morphology (Rodriguez-Iturbe and Valdes 1979; Valdes etal. 1979; Gupta etal. 1980; Rinaldo etal. 1991), attempts were made to estimate the hydrologic response ofcatchments. Relationships between geomorphological basin characteristics and parameters of the geomorpho-logical IUH function were retrieved. The Nash conceptual model with the same time lag and peak height fits the resulting GIUH very well (Rosso 1984).

14.3 FLOODS AND FREQUENCY ANALYSIS

In recent years, a number of European countries experienced damaging flood events that were caused by persistently high precipitation concentrated in mountain areas. In some countries, catastrophic floods occur almost every year, damaging villages, infrastructures, cultivated areas and causing loss of lives. The most recent floods occurred in central Europe in August 2002, in the Po basin in 1994 and 2000, in Italy (Valtellina, a combined flood and landslide event) in 1987 and in Switzerland, Austria, Germany (Rhein and Meuse) and the Netherlands in 1993, 1994, 1995 and 2000, respectively. The Oder, Morava and Danube rivers experienced floods in July 1997. In the years 1998-2002, more than 200 floods occurred in Europe, leading to the loss of 700 lives. In the 2002 flood, less than 100 lives were lost but costs cumulated to 4.1 billion Euros for insurance companies.

Scientists and citizens are challenged by the question of whether such frequent and catastrophic events are due to extraordinary rainfall episodes, to changes in our climate, or instead result from increased vulnerability of territory and insufficient maintenance of river beds.

The debate also remains open because available data is contradictory. With reference to the Italian rivers, it is important to note that whereas the gauging stations of Piacenza and Pontelagoscuro in the Po river show increasing maximum annual floods in the period 1920-2000, those in the Adige river mostly measure steady or lightly decreasing floods in the period 1923 -1997. In the Adige basin, the mean daily discharge shows a significant decreasing tendency by about 30% (Villi 2003), a fact observed for several other rivers in Europe too.

The costs and effectiveness of structural measures mitigating the effects of floods depends on the design flood estimates. The choice of the design peak-flood, QT, with a given return period, T, is of crucial importance for river training. It is based on the assumption of level of acceptable risk, a knowledge of river hydrology and the availability of hydrological data. In brief, with increasing potential damage the cost of protection increases parallel to an increasing design return period.

In the last 30 years, effective procedures have been established for flood estimation, including several techniques for regionalisation of flood data in a unique statistical model (Committee on Techniques for Estimating probabilities for extreme floods, 1988). The basic hypothesis assumes that time series of peak floods in a given region rescaled to an appropriate index value (e.g. the mean annual flood) represent the same dimensionless variable distributed homogeneously over each river section of the region. This hypothesis is acceptable only if the rescaled floods are spatially homogeneous.

In the Alps, a number of regional flood frequency methods have been applied. In Austria and Switzerland, the EU FRAMEWORK Program (flash flood risk assessment under the impact of land use changes and rivers engineering works) applied to the Region Of Influence (ROI) concept in addition to seasonality measures and the Gradex method. In Italy, a hierarchical-type region-alization model, the TCEV (Two Component Extreme Value) was applied on a national basis (see Rossi et al. 1984; Villi and Bacchi 2001).

As a final remark, this brief overview demonstrates both progress and uncertainty in our understanding of the hydrometeorological aspects of mountain floods and we hope that future research will improve our knowledge on these complex and relevant phenomena.

14.4 ACKNOWLEDGEMENTS

The manuscript benefited from comments by Roberto Ranzi and English proofreading by Carmen de Jong and it is a pleasure to acknowledge their assistance and suggestions.

REFERENCES

AlpertP., Mesoscale indexing of distribution of orographic precipitation over high mountains. J. Appl. Meteorol., 25, 532-545, 1986. Beven K.J., Kirkby M.J., A physically-based variable contributing area model of basin hydrology, Hydrol. Sci. Bull., 24, 43-69, 1979.

Brath A., Castellarin A., Montanari A., Assessing the effects of land use changes on annual average gross erosion. Hydrol. Earth Syst. Sci., 6(2), 255-265, 2002. BougeaultP., Houze R.A. jr., Rotunno R., Volkert H., Guest Ed., Mesoscale Alpine Programme (MAP). Q. J. Meteor. Soc., 129, 588, Part B, 2003. Buzzi A., Foschini L., Mesoscale meteorological features associated with heavy precipitation in southern Alps. Atmos. Phys., 72, 131-146, 2000. Committee on Techniques for Estimating probabilities for extreme floods, Techniques for Estimating Probabilities of Extreme Floods: Methods and Recommended. National Academy Press, Washington, DC, 1988. Dunne T., Moore R.T., Taylor C.H., Recognition and prediction of runoff-producing zones in humid regions. Hydrol. Sci. Bull., 20(3), 305-327, 1975. Frei C., Schar C., A precipitation climatology of the Alps from high-resolution rain-gauge observations. Int. J. Climatol., 18, 873-900, 1998.

Green W.H., Ampt G.A., Studies on physics, part I, the flow of air and water through soils. J. Agric. Sci., 4(1), 1-24, 1911.

Gupta W.K., Waymire E., Wang C.T., A representation of an instantaneous unit hydrograph from geomorphology. Water Resour. Res., 16(5), 855-862, 1980.

Hewlett J.D., Soil Moisture as a Source Baseflow from Steep Mountain Watersheds, (Station Paper 132). U.S. Forest Service, Southeastern Forest Experiment Station, Asheville, NC, 1961.

Horton R.E., The role of infiltration in the hydrologic cycle. Trans. Am. Geophys. Union, 14, 446-460, 1933.

Horton R.E., An approach towards a physical interpretation of infiltration capacity. Proc. Soil Sci. Soc. Am., 5, 399-417, 1940.

KirnbauerR., Bloschl G., GutnechtD., Entering era of distributed snow model. Nord. Hydrol., 25, 1-24, 1994.

Martinec J., Rango A., Revisiting the degree-day method for snowmelt computation. Water Resour. Bull., 31(4), 657-669, 1995.

Moore R.J., Clarke R.T., A distribution function approach to rainfall-runoff modelling. Water Resour. Res., 17(5), 1367-1382, 1981.

NashJ.E., The form of Instantaneous Unit Hydrograph. IAHS Publication No. 42, 1379-1394, 1957.

Ranzi R., Bochicchio M., Bacchi B., Effects on floods of recent afforestation and urbanisation in the Mella River (Italian Alps). Hydrol. Earth Syst. Sci., 6(2), 239-265, 2002.

Rinaldo A., Marani R., Rigon R., Geomorphological dispersion. Water Resour. Res., 27(4), 513-525, 1991.

Rodriguez-Iturbe I., Valdes J., The geomorphological structure of hydrologic response. Water Resour. Res., 15(5), 1409-1420, 1979.

Rossi F., Fiorentino M., Versace P., Two-component extreme value distribution for flood frequency estimation. Water Resour. Res., 20(7), 847-856, 1984.

Rosso R., Nash model relation to Horton order ratios, Water Resour. Res., 20(7), 914-920, 1984.

Rotunno R., Ferretti R., Mechanism of Intense Alpine Rainfall. J. Atmos. Sci., 58, 1732-1749, 2001.

Sherman L.K., Streamflow from rainfall by the unit-graph method. Eng. NewsRec., 18, 501-505, 1932.

Valdes B.J., Fiallo J., Rodriguez-Iturbe I., A rainfall-runoff analysis of the geomorphologic IUH. Water Resour. Res., 15(6), 1421-1444, 1979.

VanDelden A., The synoptic setting of thunderstorms in western Europe. Atmos. Res., 56, 89-110. 2001.

Villi V., The reduction of the daily discharges in the upper Adige River and possible causes (in Italian). Boll. Geofisico, XXVI(3-4), 73-87, 2003.

Villi V., Bacchi B., Evaluation of Peak Flood in North-Eastern Italy. CNR-GNDCI, Publication No. 2511 (in Italian), 2001.

Weingarten R., BarbenM., Spreafico M., Floods in mountain areas-an overview based on examples from Switzerland. J. Hydrol., 282, 10-24, 2003.

Was this article helpful?

0 0

Post a comment