L'hydrologie urbaine: nouvelles problématiques, nouvelles approches de solutions

La pratique de l'hydrologie, appliquée au milieu urbain, a beaucoup évolué depuis les années 1960. De nouveaux outils scientifiques, méthodologiques et technologiques ont été mis au point. En ce qui concerne les eaux pluviales urbaines, un changement de philosophie complet s'est produit. Aujourd'hui, des développements rapides continuent à se produire pour faire face de la meilleure manière possible aux graves problèmes qui se posent en zone urbaine : inondation et pollution du milieu naturel. Deux approches récentes illustrent ces développements : la gestion des risques et l'approche globale sur le bassin versant. Une évaluation de plus en plus précise de l'aléa et de la vulnérabilité s'avère nécessaire ainsi qu'une planification hydrologique bassin versant par bassin versant. Celle-ci devrait conduire à définir et à hiérarchiser les principaux objectifs que l'on se fixe dans les domaines du contrôle des inondations, du contrôle de l'érosion, de l'amélioration du milieu naturel et de la protection de l'environnement. Les actions à mener doivent utiliser les outils de modélisation des phénomènes hydrologiques sur les bassins versants. L'efficacité de ces approches est illustrée par un exemple pris sur le bassin versant de la rivière Beauport au Québec.Urban hydrology practice has evolved a great deal, keeping up with the evolution of urban problems. Thus, professional engineers have had to keep up with this evolution, in order to understand the effects of urbanization on the hydrological behavior of the systems for which they have to recommend best solutions. Engineers have also had to evolve from the traditional approach of a limited vision of the problem, of its extent and of its impacts on the system. A global vision of the entire system, grouping every watershed element, such as forestry, agriculture or urbanization, is necessary if one is to circumscribe the problem, to analyze it and to reach to the best long-lasting solution. This article represents a synthesis of the hydrological disturbance phenomenon caused by uncontrolled urbanization, not properly planned, which is often characterized by numerous short-term solutions, often ineffective in the context of a long-lasting approach.Examples of natural disasters caused by meteorological events are more and more frequent. Take the case of Vaison-la-Romaine, in September 1992, where the Groseau devastated the Vaison community. More recently, the 1996 summer floods in the Saguenay region, in Québec, demonstrate once more that water always tends to return to its original bed, now occupied by artificial structures. When these types of events are analyzed in detail, we notice that meteorological phenomena are not always the main cause of the disaster cause and that often part of the responsibility lies with the occurrence of human activity in the catchment. This observation shows the importance of using wisdom and humility towards these natural forces and of anticipating, during design, a "secure" passage for the inevitable flows that one day will exceed the structure capacity. Knowing all the interactions that exist in the heart of the stormwater management problems, it is risky to intervene in isolation without analyzing the impact of the action. The approach must then be global and coherent; the tool necessary for success is the stormwater master plan made on a watershed basis. The master plan allows us to properly describe the problematical elements, to identify the real causes and to optimize the location of the control solutions. It also allows us to manage the increase in stormwater due to urbanization, with the "flow set point" concept of each watercourse tributary.This article presents an application example where the global analysis approach is used and where the solution involves different interventions and developments, which, when combined, effectively treat backwater effects, flooding and erosion in the presented area. The master plan of the watershed containing this sector had already been conceived and the specific flows of each of its tributaries had been identified; the global approach method therefore allowed us to settle the local problems in the studied area while respecting the flow granted by the master plan. Finally, a floodable plain, containing marshes, completes the intervention plan by combining the restoration of certain watercourse ecological characteristics with the need for flood control. This example demonstrates the necessity and the advantages of approaching urban drainage problems at the watershed level. However, watershed management concepts lead planners and designers to deal with antagonisms: urbanization or renaturalization? Obviously, in urban areas, the return to natural conditions is impossible and the disturbances to watercourses generated by urbanization cannot all be compensated for. Rather, we must aspire to the establishment of a balanced environment by controlling inflows produced by the watershed and by supporting the system to receive them. The notion of feasibility must always occupy the thoughts of the master plan designers and planners. The credibility and the continuity of the master plan depend on all these considerations; any unrealistic flow order, on an implementation level, can compromise the balance of the management plan and its applicability, which in turn can lead to a confused plan, possibly having disastrous consequences. The ultimate consequence of the management plan is the protection of the watercourse, the development of which can be designed to meet different criteria, such as flood control, erosion control, ecological potential enhancement, etc. It is at this stage that management becomes most complex: flood control, for example, does not apply to the same type of events as erosion control. The former requires management of major hydrological events, whereas the second needs implies control of frequent flows, which are at the origin of bank undermining caused by differential volume excesses. The management scheme must then handle multiple events and multiple criteria; consequently, it becomes more complex but also grows in value and justifies itself more adequately. Furthermore, the whole question of watercourse quality, in the broad sense of the term, must be managed from a frequent-event point of view, as this yields the best physical, ecological and aesthetic image of the aquatic environment. To this effect, present efforts in the control of combined sewage network overflows can be seen as a predecessor of what will later be extended to the entire stormwater network. This whole aspect of management has not been treated in this paper, as many others are dedicated to doing so. The evolution of the situation is similar in France and in Québec; coherent watershed management must rest on clear political goals regarding environmental use, risk protection and urban development. The stormwater master plan must, permanently, become part of the water resource master plan for the whole watershed, with its objectives of flood control, erosion control and maintenance of acceptable physical-chemical and biochemical water quality, as well as assured resource use (drinking water--, hydroelectricity, agriculture, industrial waters, etc.). This water resource master plan will have to be integrated in the same way as the French Water Agencies do today, and as Québec is preparing to do so