Comparing the effectiveness of upstream nature-based solutions with building-level adaptation measures: a case study for the Geul river

In July 2021 large flooding took place in North-Western Europe. The Geul river, which is shared between the Netherlands, Belgium and Germany, was one of the flooded catchments, with total damages estimated to be €250 million. Since then, there has been a call for additional flood risk reduction measures in the area, including transboundary nature-based solutions in upstream parts of Belgium and local scale flood-proofing of buildings in The Netherlands.

The main novelty of our study is to make an economic trade-off between upstream nature-based solutions (NBS) and downstream building-level measures. For this, we further develop GEB, a coupled agent-based hydrological model and integrate the hydrodynamic model SFINCS into GEB. Furthermore, to calculate high-resolution risk estimates for buildings, we use object-based exposure data from OpenStreetMap and empirically derived vulnerability curves using survey data at the building level. The model allows us to 1) understand current flood risk in the Geul catchment at the object-level and 2) evaluate the effect of several flood risk reduction measures. The model validation shows good performance against observations of flood extent (CSI=0.66), flood depth, and damage of the July 2021 flood.

We then quantify the risk reduction of several nature-based solutions (wetland restoration, reforestation, retention ponds and the conversion of agricultural land to natural grassland) and building-level adaptation measures (wet-proofing and dry-proofing). Moreover, we examine the effect of upstream nature-based solutions on downstream communities. Finally, we perform a cost-benefit analysis (CBA) to gain insight into which combinations of measures are most desirable. Our results show that NBS are especially effective for less extreme floods with high return periods (<1/25). For extreme floods (>1/25), benefit-cost ratios (BCR) may drop to 0.25 or lower. However, these numbers do not account for co-benefits (e.g. tourism). The results can be used by policymakers to design effective flood risk management strategies.