Predicting flooding of the river Meuse in Wallonia

Researchers at the University of Liege have published a multi-disciplinary research paper which makes it possible to estimate the damage that could be caused by variations in discharge of the river Meuse between now and 2100. The publication followed the AMICE project, an original, international, interregional and innovative collaboration project on the river and its river basin. To get a better grasp of the problem, the scientists from the University of Liege combined climatology, hydrology, hydraulics and urban planning, and focussed their research on the current considerations of the public authorities and the IPCC (Intergovernmental Panel on Climate Change). Their conclusions are clear: there needs to be better management of urbanization to limit the upward trend in considerable flood-related damage. It is estimated that, in Wallonia, the level of the river during a one hundred-year flood peak will increase by an average of 60 centimeters between now and 2050 and by 130 centimeters between now and 2100, creating a risk of flooding. Conversely, situations of prolonged low water levels are predicted for the drier months. The researchers have created an impressive chain of numerical models and relevant theories in the public interest.  

Influenced by global warming and uncontrolled urban planning, flooding and the damage it causes have increased steadily, both in terms of frequency and intensity.  Predictions about flooding by rivers do not leave any margin for optimism. The river Meuse, which flows through Wallonia over a distance of 185 km, is no exception. The river flows through 19 of our communes which, to a greater or lesser degree, are exposed to a very high risk of flooding that is likely to increase over the coming years. In fact it is estimated that in Wallonia, the level of the river during a one hundred-year flood, will increase by 60 centimeters between now and 2050, and by 130 centimeters between now and 2100, leading to a risk of flooding. Conversely, situations of prolonged low water levels are predicted for the drier months.

A multi-disciplinary research project conducted at the HECE unit (Hydraulics in Environmental and Civil Engineering) of the Faculty of Applied Sciences of ULg aims to establish the risks of flooding in our region with the aid of a new predictive model. It extends into the future as far as 2100, and by developing several possible scenarios, estimates the damage that might be caused by such disasters. In order to accurately reflect a situation that is much more than a simple question of hydrology, and which depends on many factors, the study also integrates considerations such as climate, geography, hydraulics and urban planning, because although flooding is a natural phenomenon, the material damage it causes also depend on the exposure of zones in areas prone to flooding, and therefore involves political decisions as well.

The timing of the study could not be better. At a time when natural catastrophes are making headlines, it is now, more than ever, time to rething the systems of flood-protection such as dams, dykes… which in some cases are fast becoming obsolete. However, a systematic reconstruction of such infrastructures is economically unthinkable. Such costs could be avoided however. For this to happen, a complete rethink of planning in Wallonia (which defines the exploitable urban areas) and a restriction on building in areas prone to flooding will be neccessary.

Cross-border harmonization

There is an increasing awareness of the global, or at least the inter-regional character of the risk of natural catastrophes, and the need for management of this risk to be shared by the actors concerned. The first step in sharing this risk begins with a harmonization of observations and ideas about our surrounding environment. This article is the result of research that aims to adopt the above approach to the problem, explains Benjamin Dewals, lecturer in hydraulic engineering at ULg. “This publication follows the AMICE project which focussed on the entire Meuse river basin. The originality of this approach was to enable teams of scientists and managers from each country concerned to work together (France, Germany, Wallonia, Flanders and Holland, Editor’s note.). In an overall context, the project aimed to assess all the consequences of flooding and one hundred-year flood peaks between now and 2100, by taking a coherent account of the impact of climate on a basin-wide scale. For this to be possible it was neccessary for parties involved to work together.”

Previously, each region, climate or meteorological institute relied on its own analysis, often from one border to another, or even from one team to another. In the present case, the aim of the project was to harmonize theories by taking account of a maximum number of possible factors. Predictions were therefore dealt with from a hydrological and then a hydraulic point of view before including an assessment of economic damage by including urban factors. Thus predictions made it possible to suggest a better way of managing the impact of the effects of climate change on the river Meuse. 

A climate model between two extremes

Projections about future flooding of the river Meuse were divided into two main climate scenarios. The “dry-climate” scenario, which is the most optimistic one and the “wet-climate” scenario which is more pessimistic and which predicts a 30% increase in the one-hundred year flood peak flow.  The reality is certainly somewhere beween the two but it is difficult to be more precise using current means of projection. 

“2100, is a long time away”, says Arnaud Beckers, a doctoral student in geography and the first author of the publication. “We are obliged to establish several scenarios to define possible future events and observe the implications of each of these”.“Another difficulty”, continues Benjamin Dewals, “is that the scenario that will occur will not only depend on physical processes linked to climate. Political, economic and social decisions will also have an influence and these are an unknown quantity. All we can do is to take stock of all possibilities and to work with that in mind. According to what we find, we will be able to develop adaptation measures for protection or management which although they will not work in an optimal way for a given scenario, will be satisfactory when weighed against the range of possibilities”.

It is also important to specify that the two climatological extremes were not just randomly selected.“For example, explains Pierre Archambeau, a research engineer at the HECE unit, “the most optimistic theory of no cimate change was formed after a series of precise calculations resulting from a chain of models taking account of greenhouse gas emissions, political decisions, climate models on a grand scale such as the evolution of the atmosphere, temperature and precipitation. These results were then integrated into a more precise hydrological model which made it possible to calculate the evolution of the water runoff in the river Meuse basin. This is not a theory that we could have anticipated at the outset. It is a result in itself.”

By estimating the likely patterns of flooding, and by concurrently developing an urban planning model, it became possible to predict the economic damage that will be caused by these future natural catastrophes. 

What is the level of vulnerability of flood-risk areas?

One of the original aspects of this course is therefore the inclusion of the urban question as well as climatological considerations. “We realized ”, explains Benjamin Dewals, “that previous research studied climate evolution, but it did not attribute equal importance to other important factors such as the development of urbanization. And yet, this factor is of the utmost importance. Urbanization in the river basin, for example, changes groundwater flow dynamics. Rainwater seeps into the ground or flows more or less quickly into the rivers and in time, changes the intensity of flood peaks”. This is a first aspect of urbanization which does not require models with a high level of precision. However, it is not this aspect that is dealt with in this publication. The aim was to observe what happened in terms of damage by restricting access to flood-risk areas. In order to do this, it was neccessary to perfect models that made it possible to work on a small scale to reproduce dwellings, roads or protection walls.

“In fact, a map of the flood-risk areas in Wallonia already exists”, explains Michel Pirotton, an associate professor of hydraulics at the HECE unit. “We extended the methodology to cover stronger occurrences. Different stages of calculations and hydraulic modeling are required in order to determine the vulnerability of a flood-risk area. For areas near the river, we used a hydraulic model with a five meters by five mesh. We simulated the flow of water on digital mockups of towns where we simulated the overbank flooding of the river Meuse. Every five meters, we obtained a value for the height of the water, the speed of the flow in each direction and then, by crossing the model with other data, the damage inflicted”. These measures were also useful for studying those undeveloped ares that the authorities consider suitable for development.

Establishing a precise mapping of the areas vulnerable to flooding was of the utmost importance in order to envisage a better management of dams, but especially, to assess the relevance of more restrictive political measures with regard to construction in flood-risk areas. These measures will have to take account of a large increase in urbanization, because by the year 2100, there will be 700,000 extra dwellings in Wallonia, and this does not include future industrial zones, new schools, businesses and road and rail networks. Given such growth, and in-keeping with current growth, all the areas proposed in the development plan, including high-risk areas, will probably be occupied by the end of the century. Given this fact, there are two possibilities to be considered.  New construction zones will have to be found, or, more realistically, population density in urban areas will have to increase. 

Nine possible urban scenarios

Urbanization is not something that takes place without due consideration, it must follow certain trends such as ecological awareness, economic development, the cost of land for development, or the concentration of areas of major employment. Based on recent knowledge of the degree of risk associated with vulnerable areas, together with the socioeconomic questions mentioned above, the researchers have developed nine possible urbanization scenarios and have calulated a damage estimate due to overflowing of the river Meuse. 

There are three possible ways in which urbanization might develop. Firstly, the development might take the form of the current trend, by following the development plan envisaged by the government of Wallonia in the 1980s to organize urban planning in the region.“However, at a time when there were few problems such as traffic jams, public transport, and sewer networks, urban planning was organized in the form of a very extensive network in areas where current dwellings are concentrated and left a lot of space available for future urban development. Today, urbanization is very spread out and could be unfavorable in terms of the vulnerability of areas to flooding.”, explains Arnaud Beckers.

The question that now arises is what would happen if the state decided to constrain urbanization even more. The question leads to two other possible urban expansion programs and which includes a concentration and densification of urban areas in concentrated dwelling areas.“These theories were not just plucked out of thin air”, explains Arnaud Beckers. “They are in accordance with the current political thinking of the ministry responsible for urban planning. These areas of concentrated housing vary according to the density of the population, services such as schools or businesses, distance to main emploment centers and are the result of a rationale which covers considerations other than flood-prevention.” These theories aim to bring about a more rational use of territory such as limiting journeys by concentrating the population into areas that are already populated and where services are altready developed, for example. These scenarios are prospective therefore, but are connected to concrete discussions currently taking place in Wallonia. 

The first of these two scenarios of urban densification was estimated at a regional level, where there is a concentration of households near the main centres of employment, that is to say, Brussels, Charleroi and Liege. The second would involve a densification at a more local level, no longer taking account of an exodous from the countryside or from smaller towns. These possible scenarios will depend mainly on limits to future political decisions. They will become a reality following the occurrence of many factors external to public authorities, such as the availability of work, the cost of living in the big cities or the fluidity of public transport.

Each of these three urbanization scenarios (continuation of the current situation, densification at a regional level, densification at local level) was confronted by three degrees of restriction. The highest degree of restriction was a ban on building in areas with average to high flood-risk, the lowest consisted of an absence of restriction. The middle degree of restriction only banned urbanization in zones with a high flood-risk level.  By combining these results with the two climatic scenarios, the researchers were able to establish flood-related damage estimates due to flooding of the river Meuse over the next 100 years.



Ultimately, according to the optimistic “dry climate ” scenario, and following nine possible town-planning scenarios, damage caused by flooding would reach a range calculated to be somewhere between 334 and 462 million Euros, representing a growth of between 1 to 40% in relation to an estimate based on 2009 data. On the other hand, according to the less optimistic scenario, predicting an increase in flood-risk of 30% between now and the year 2100, damage would reach a range estimated to be between 2,124 and 2,408 million Euros, which represents a growth of between 540 to 630%.

The figures speak for themselves, a restriction in the areas of risk causes a marked decrease in costs related to flood-damage and this applies to all possible scenarios. Although far from being sufficient, such a restriction would allow for a reduction in catastrophic damage along with other measures such as the building of dams or dykes to avoid catastrophic damage. These figures only concern flooding of the Meuse but provide some indication of general trends. For their part, the researchers have just launched a new four-year study which will take into account new factors such as disruption to transport, economic activity, interaction with underground water, or the effects of flow into the river basin in more precise urban and climate scenarios. While waiting for their results, a first prudent measure to be taken by public authorities would be to cease building in areas prone to flooding.

A. Beckers, B. Dewals, S. Erpicum, S. Dujardin, S. Detrembleur1, J. Teller, M. Pirotton, P. Archambeau Contribution of land use changes to future flood damage along theriver Meuse in the Walloon region, Natural Hazards and Earth System Sciences, 2013 (http://www.nat-hazards-earth-syst-sci.net/13/2301/2013/nhess-13-2301-2013.html)