Improving suitability of urban canals and canalized rivers for transportation, thermal energy extraction and recreation in two European delta cities

Canals and canalized rivers form a major part of surface water systems in European delta cities and societal ambitions to use these waters increase. This is the first assessment of how suitability of these waters can improve for three important uses: transportation, thermal energy extraction (TEE) and recreation. We assess suitability with Suitability Indices (SIs) and identify which alterations in the water system are needed to improve SI scores in Amsterdam, The Netherlands, and Ghent, Belgium. The results show spatial variability in suitability scores. Current suitability for transportation is low (SI score = 1) to excellent (SI score = 4), for TEE fair (SI score = 2) to excellent (SI score = 4), and suitability for recreation is low (SI score = 1). Suitability could improve by enlarging specific waterway dimensions, increasing discharge and clarity, and by enhancing microbiological water quality. The same methodology can be applied to optimize designs for new water bodies and for more water uses. Supplementary Information The online version contains supplementary material available at 10.1007/s13280-022-01759-3.


2/10
Appendix S1: Suitability Indices Suitability Index for Thermal Energy Extraction (TEE) The SI score is related to indicative heat extraction capacity, assuming heat extraction during the three warmest months (Van der Meulen et al., 2022). During summer months, when water temperatures are the highest, heat is extracted from surface water with a heat exchanger. Warm water is stored for later use, for example in an Aquifer Thermal Energy Storage in the subsurface. When the heat is needed for heating of buildings during colder months, warm water is pumped up from the storage. Heat is extracted from the water through a heat exchanger and a heat pump further heats the water to the desired temperature for the building's hot water network. Data from the three warmest months are used for calculating the SI score. If the precondition for water depth is met, the SI score is determined by integrating the sub-scores for width, discharge and temperature (Table S1). Because a low value for one parameter can be counteracted by a high value for another parameter to some extent the integration method is the geometric mean of the sub-scores. ≥ 0.5 a n.a. n.a. n.a. Integration of sub-scores into SI score: geometric mean b

SI Recreation
The SI score is related to the level of risk for adult swimmers (Van der Meulen et al, 2022). If the precondition for water depth is met, the SI score is determined by integrating the sub-scores for E.coli bacteria (indicator of faecal pollution), Cyanobacteria, pH, clarity and depth (Table  S3). Data from the period April-September, the swimming season, are used for calculating the SI score. Since unsuitable conditions for one parameter cannot be counteracted by another, the minimum operator approach is used for integration of the sub-scores. Appendix S2: Used data The dataset for analyse with all data used for calculating the original sub-scores, SI scores and the impact of raising sub-scores on the SI values per HU or point location is provided in the separate file vandermeulenetal_datasetS1.xls. Table S4 contains an overview of all used data sources to create the dataset for calculation of SI scores for TEE, transport and recreation. Where applicable, additional remarks or motivations are provided below.

TEE Amsterdam
 Minimum depth in the fairway: This is an underestimation of the minimum depth in the waterway, but all values are above the precondition of 0.5 m.  Temperature: data from point locations, at least one measurement per month. At all locations, average temperature and the 95th percentile of decreasing values is above 15°C during the three warmest months, even when taking into account that night-time temperatures are ≤1 °C lower (see Van der Meulen et al., 2022) than daytime values. Therefore, a sub-score of 4 is applied to all HUs.  Discharge: values generated for line segments in the centre axis of waterways are assigned to the HUs that they cross. The 75 th instead of the 95 th percentile of decreasing values is used because flow direction variations otherwise result in (near-) zero discharge.  Width: maximum width is the navigation database is the best available value but will give some overestimation.
TEE Ghent  Depth: value is derived from allowed ship draft (see section on Transportation). This is an underestimation of maximum depth, but this is no problem because the precondition of 0.5 m is met everywhere.  Temperature: The dataset includes temperature figures for 42 locations covering many waterways around the city of Gent. Values during the three warmest months are all ≥ 17.3 °C and therefore a sub-score of 4 is applied to all HUs.  Width: Average width per polygon is determined in ArcGIS by measuring width at several points, the number of points depending on the length and shape of the polygon, and calculating the average.

Transport Amsterdam
 Minimum depth in the fairway: This is an underestimation of the minimum depth in the waterway but all values are above the precondition of 0.35 m.
Transport Ghent  Depth: value is derived by multiplying allowed ship draft by 1.2 as best estimate of depth. This factor is based on applied keel clearance targets in several local sources (Arcadis, 2007;De Rijck, 2011;Rijkswaterstaat, 2017; https://emis.vito.be/en/node/12231).  Minimum width: measured by means of Google maps, taking into account jetties and other large permanent structures visible on the satellite image. Like in Amsterdam, some bridges (may) have piers. This seems less frequent pthan in Amsterdam. Their influence on width is not taken into account unless guidance jetties are dividing the fairway.

5/10
Recreation Amsterdam  Water quality data at point locations based on field and laboratory measurements in samples from 0.3 m depth. Except for cyanobacteria, samples are taken at least monthly during summer months.  We include all locations with at least 3 years of data.

Recreation Ghent
 We include locations with at least 3 days of E.coli-and pH data. At three locations (Houtdok and two locations in Watersportbaan), dozens of datapoints are available for E.coli and pH; for the other locations a maximum of 4 (pH) or 5 (E.coli) days of data are available. Apart from the three frequently monitored locations, there are less than 3 days of clarity data available.
(   1 Dataset with data from field measurements between 2004 and 2016, data are still valid. Depth and air draft are expressed in a unit that requires correction for water level, which is done with water level data from AGV. 2 By means of a lead line. For Amsterdam field campaign in January 2021, for Ghent August 2021.

8/10
Appendix S3: extra Figures   Figures S1-3 illustrate notable geographic differences with respect to the parameters that need to be altered for improving SI scores.
Figure S1 Improving SI Transport scores in Amsterdam requires higher sub-scores for depth, air draft and/or width. In most HUs, two of these parameters need to be altered. HUs where all three parameters need to be improved are mostly located in the city center. Figure S3 Improving SI Recreation scores in Ghent requires higher sub-scores for clarity (C) alone, or in combination with Cyanobacteria (Y) and E.coli bacteria (E).