The Geul River Basin Project

Introduction

In the Geul River catchment, important changes have taken place during the last 50 years: In agriculture, permanent pastureland was changed into cropped fields and heavy machinery was introduced. During the same period, hydraulic modifications of the river course were also introduced.

All of these factors, added to possible long-term climatic changes, contributed to an increase in the number of serious flash floods in the last 50 years. It has been shown, for instance, that the characteristic maximum water discharge based on a four-year return period was as much as 39 m3/sec during the 1980’s, up to 23 m3/sec during the 1950’s.

Recent flash floods in the Geul River and its tributaries have increasingly induced flooding of populated areas. The resulting water and mud damage cause enormous expense, costing local authorities between 5% and 40% of their annual budget to simply maintain local infrastructure. The floods of 1982 and 1987 in Valkenburg, for example, cost Dutch society millions of guilders.

At least part of this increase can be explained on the basis of recent land use changes, that may indeed account for an increased of surface water runoff and consequently for an increased risk of flooding. The (until recently) generally applied water management policy to prevent local flooding, by accelerating the evacuation of access runoff directly into the hydrographical network, has undoubtedly contributed to the problem at catchment scale.

Objectives

The overall goal in this pilot study was to define environment-friendly measures for increasing water retention in the Geul catchment, thereby reducing the risk of flash floods in the Limburg region of Belgium and the Netherlands.

Within the scope of this overall goal, the following objectives were identified:

• To inventory relevant information and develop a methodology to combine the data coming from many sources of information and under different formats;
• To perform a preliminary hydrological and hydraulic risks analysis using ISIS;
• To develop different scenarios of environment-friendly measures with emphasis on water retention on watershed or river, to reduce the risk of flash flood events;
• To analyse the effectiveness of these different scenarios to reduce the risk of flash flood events.

• The results suggested an increase of the risk of runoff on arable land. Although its acreage did not change since the 1950’s its contribution to the risk increased because of the increased proportion of weeded crops. The qualification of these risks is made in the hydrological analysis.

• In general, mean annual flow and flow coefficients are (around 30% or less, including during flood events) with respect to usual evapo-transpiration rates for this area; low annual flow coefficients for the tributaries of the Geul are observed, suggesting a “loss” of water.



• The flood rise time is short (2-4 hours in the central part of the basin; phase-leg of 4-10 hours at the outlet), related to the shape and topography of the watershed.



• Peak flows show normal values (200l/s.km2 at Hommerich), but a damping of the peak flows is observed at the outlet (100l/s.km2 at Meerssen).

• The results show that the proposed environment-friendly measures do reduce the duration ad fluxes of the water to the river network under high rainfall events, but only weakly: the reduction is maximum 10%. In general, scenario 3 (transformation of farming and pastureland with slopes higher than 10% into forest) and land-related measures of scenario 4 (transformation of all farming land into farming land with green belts) show the highest improvements as regards daily water flux reduction. These scenarios give more or less the same result as the simulation of the historic land use scenarios of the 1950’s.



• The simulated scenarios, especially the land-related measures of scenario 4 (farming land with green belts and hedgerows), are also favourable to decrease local erosion problems (gullies). This erosion problem would need some further (specific) elaboration.

• Analysis permitted to conclude where the flood-prone areas are located and which scenario appears most favourable to reduce the risks of flash flood events. Currently, major flooding occur upstream of the mills: Volmolen, Epermolen, Bovenste Molen, Wijlre, Oude Molen, Franse Molen and Groote Molen. Other flood-prone areas are the Geul River in Wallonia near the Dutch border, and the Geul River at the mouth with Eijserbeek and Gulp.



• The effectiveness of the land use measures, aimed to retain water in the uplands and to reduce the peaks of water runoff to the river network, proved insufficient to prevent flooding. Peaks in river water discharge were maximally reduced by 10% for the historic land use scenario of the 1950’s (as obtained in the hydrological study. Peaks in river water heights were maximally reduced by 22cm.



• The installation of to constriction devices in the upstream parts of the Geul River, combined with land-related measures (all farming land with green belts or hedgerows), reduces the peak river discharges by maximum 17%. This proved to be the only scenario with water heights in the Geul River generally below the heights of the riverbanks; however, it cannot guarantee the complete prevention of flooding problems.



• Combining land-related measures with civil-technical measures in the river is necessary to reduce the risk of flash flood-related inundations along the Geul River.