How has urbanisation already altered a city’s natural water balance? The case study of Hamburg to present a commonly missing step before considering green infrastructure interventions

The concept of water-sensitive cities continues to gain traction globally, as the disruptive effects of urbanisation on local hydrological processes and the potential benefits of green infrastructure become increasingly evident. Despite this, in many planning instances, consideration is only given to how the water balance will be altered and hazard risk reduced from the current urbanised state to the state after implementation of green infrastructure. Why is the understanding of the natural water balance in the pre-urbanisation state often not considered as reference point for planning? If urban green infrastructure should provide hydrological and ecosystem services, should these services be similar to those in the natural condition before urbanisation?

For our study, we recreated the daily near-natural water balance for the city of Hamburg to quantify how urbanisation has already affected the water balance, particularly in years of hydrological extremes that represent hydrological hazards. Using the fully-distributed daily water balance model mGROWA, we developed two very high resolution (25 m) models for the city of Hamburg for 1991–2020; one representing the current hydrological situation and one representing a theoretical near-natural situation. To generate the near-natural scenario, the input datasets for topography, soil and land cover were adjusted through the integration of various datasets representing non-anthropogenic conditions, while sealed surfaces and artificially drained areas were removed from the datasets. As expected, due to the lack of runoff from sealed surfaces the actual evapotranspiration is much higher (+40%) in the near-natural scenario than in the current one. Groundwater recharge was also higher in the near-natural scenario (+27%), mainly due to the lack of surface sealing. We then compared the water balance components for the two models against the SPEI meteorological drought index to assess differences in the extremely wet and extremely dry periods that represent potential hydrological hazards. This revealed an increasing divergence in some water balance components between the scenarios for the extreme conditions, quantifying the extent to which the urbanisation of the city has exacerbated hydrological hazard risks. Our study presents a transferable methodology for assessing how urbanisation has affected the natural water balance of a region, which can be used as a starting point for defining targeted solutions for green infrastructure, with the aim of achieving water-sensitive cities.