The Relevance of Urban Water Metabolism to Groundwater Governance: Insights from Two South African Cities

Abstract

Groundwater is increasingly relied upon in cities, particularly during drought, yet its management often lacks coordination and systems-based decision-making. Effective governance requires inclusive participation across sectors and scales, engaging actors with diverse knowledge, experiences, and priorities. In cities, this is challenging due to the wide range of roles and responsibilities tied to groundwater. This study examines the value of urban water metabolism analysis (UWMA) for enhancing groundwater governance in Cape Town and Nelson Mandela Bay, South Africa—both recently affected by severe drought. Through a series of Learning Labs, we convened groundwater-related actors to co-develop a shared understanding of urban water systems. We brought together two methods of systems enquiry, UWMA and governance network analysis to explore physical stocks and flows of water across metropolitan boundaries with governance processes shaping groundwater management. The UWMA revealed that, prior to the 2015 drought, Nelson Mandela Bay’s water supplies were more diversified than those of Cape Town, despite Cape Town progressively pursuing managed aquifer recharge and wastewater reuse. The governance analysis surfaced the diversity of actors influencing groundwater flows across the public, private, and civil society sectors, yet highlighted the fragmented nature of the network, with geohydrology and engineering consultants often acting as intermediaries. This research found that UWMA was perceived to be most useful at larger scales (e.g., watershed/urban scales) and was considered a valuable tool for strategic discussion, though clearer language would increase accessibility. We conclude that UWMA helps identify knowledge gaps, integrate diverse perspectives, and foster stakeholder cooperation. Coupled with scenario planning, it can support participatory and inclusive decision-making.

1. Introduction

Many cities of the Global South are facing the real-time challenges associated with accessing adequate water supplies, especially in times of drought, and are operating in response mode with little budget or pro-active capacity to invest in transforming urban water systems. In South Africa, drought management remains largely reactive, with many municipalities addressing droughts as emergencies rather than through strategic planning [1]. Groundwater use typically intensifies during crises, often through the uncoordinated drilling of public and private boreholes to augment surface water supplies, contributing to user competition [2], week regulation, and fragmented management of the resource as a whole. Moreover, groundwater is seldom recognised for its wider ecological role in supporting aquatic ecosystems and the services they provide, being viewed primarily as a supplementary water source [3].

Dependence on groundwater is rising in developing cities [4], but institutional capacity to adequately govern and manage its use is often limited [5,6]. Decision-makers who are tasked with preparing cities for drought risk require a systems perspective to understand groundwater’s role in urban water cycles and the information and agency that shape water outcomes. Building such a holistic picture requires the input of diverse groundwater-related actors, which in turn requires processes that facilitate an inclusive and participatory approach to learning and knowledge co-production.

This paper explores the relevance of urban water metabolism to groundwater governance in two South African urban settings, Cape Town and Nelson Mandela Bay. Through a series of Learning Labs engaging with diverse stakeholders, we trialled two methods of systems analysis: urban water metabolism analysis (UWMA) to quantify urban water flows and actor network analysis to qualitatively identify connections between actors influencing groundwater flows. This approach provides a view of water stocks, flows, and governance arrangements, offering insights into evolving groundwater practices. Bringing together UWMA and actor network analyses within the framework of Learning Labs is a novel contribution to the small but growing body of research that integrates the material and social dimensions of urban sustainability. By applying the urban metabolism approach to processes of decision-making and strategic planning, this research centres around sustainably managing climate change impacts and building the water resilience of urban systems in the Global South.

1.1. Groundwater Governance

Groundwater governance refers to the frameworks and processes through which political, economic, and administrative authorities guide the allocation, use, and protection of groundwater [7]. It encompasses the institutions and mechanisms that allow public and private actors to express interests, resolve conflicts, and uphold their legal rights and responsibilities. In contrast, groundwater management refers more specifically to the actions taken to implement decisions on the allocation, use, and protection of groundwater, as well as connected systems such as irrigated agriculture and aquatic ecosystems, as described by Huggins et al. [8]. Closas and Villholth [9] contend that groundwater governance is not a goal or a linear process of policy decisions and the implementation of management rules, but rather an ongoing, dynamic process within an existing system that reflects the biophysical and socio-political landscape of a particular context. Existing social scientific research [5,10,11] on groundwater practices highlights that it is not only government officials who make decisions about allowable land uses, drilling permits, and (ground)water use licenses that shape the volumes and flows of groundwater beyond its natural course, but also land owners, farmers, mining companies, manufacturing companies, water engineers, water diviners or dowsers, and, of course, hydrogeologists. Verzijl [12] and Aslekar et al. [13] highlight the different and often competing logics that underpin decisions about accessing, allocating, exploiting, and protecting groundwater, whereas Thomann et al., [14] highlight the challenges and shortcomings evident in real-world attempts to adaptively manage projects and activities affecting groundwater across diverse contexts.

Water crises often stem from governance failures [15], such as poorly defined roles, the concentration of power amongst the political and economic elite, ineffective information sharing, inadequate stakeholder engagement, and insufficient capacity to anticipate or adapt to water risks [16,17]. Transforming groundwater governance requires transparent, participatory, and inclusive decision-making processes, supported by coordination across organisations, sectors, and scales [18,19]. A key component of this transformation is fostering social learning among diverse stakeholders—linking lived experiences with technical and scientific knowledge to build a deeper and wider understanding of groundwater-connected systems [20,21]. This process depends on and strengthens system-level thinking [22].

When decision-makers in households, businesses, and government agencies develop an holistic understanding of the biophysical processes and the social dynamics (including the experience, actions, and motivations of other actors) of the groundwater systems they influence, they can better identify constraints and opportunities for action, learn from the past, and integrate socio-ecological impacts into drought adaptation strategies. Building this understanding involves mapping the physical stocks and flows of water, identifying the actors and their capacities to influence these flows, and incorporating the lived experiences and knowledge of those impacted by poor or inequitable groundwater governance. Currently, urban metabolism offers a useful approach for conceptualising and quantifying the biophysical aspects these dynamics within urban systems [23,24], but questions remain as to how to further explore, analyse, and represent social dynamics that reflect the diverse array of influences and experiences.

1.2. Urban Water Metabolism as a Systems Analysis Tool

Urban metabolism, or a metabolic reading of cities, has taken many forms of scholarship, some more figurative and others more materialist, as expounded by Gandy [25]. Urban metabolism is often applied as a conceptual framework to quantify and analyse the flows of resources—such as materials, energy, water, greenhouse gases, and nutrients—into, within, and out of cities [26,27], with the goal of achieving high resource efficiencies as observed in natural systems. Urban water metabolism analysis (UWMA) applies this approach specifically to water, providing a mass balance of the urban water system that includes both hydrological components (precipitation, runoff, groundwater recharge, and evaporation) and anthropogenic components (water supply, consumption, wastewater discharge, etc.). Although still an emerging field, UWMA is increasingly being applied, particularly in Australian contexts [24,28,29], but also in cities such as Bengaluru in India [30], Cape Town in South Africa [31], and Adama city in Ethiopia [32].

From mass balance analyses, performance indicators can be applied to quantify whole-system metrics such as water use efficiency, water supply internalisation, hydrological performance [28], and the potential for stormwater or wastewater to replace centralised supply [30]. Such metrics can support decision-making processes by offering a benchmark against which to measure progress and by comparing the impact of certain management interventions on the water cycle. These papers have been pivotal in providing the guidelines and robust methods of quantifying urban water flows largely for the benefit of the academic community. The application of urban metabolism (not with an explicit focus on water) as a larger field has only recently been operationalised to inform strategy planning, decision-making, and policy [33,34,35]. A recent paper incorporated urban metabolism methods with an agent-based social behaviour model to address the gap in including all water-related sectors, i.e., the water–energy–food (WEF) nexus, into strategic planning [36]. Although this analysis provided detailed impacts of various water management scenarios, there was little insight offered regarding its uptake in practical terms or how well it was received and/or acted upon. Although the value of urban metabolism for strategic urban planning is recognised [37], the challenge is finding a common language in the field that enables effective two-way communication between academics and practitioners [18].

To our knowledge, there is no research that explicitly brings UWMA into a groundwater governance lens, raising the question of the value of UWMA for understanding urban groundwater in the context of social arrangements, as well as the extent to which UWMA can bridge the gap between theoretical developments and their practical applications [32]. Although urban metabolism (in its larger context) is beginning to be integrated with other analytical methods of systems enquiry [35], the question remains whether insights from analyses focusing on material flows can be combined with alternative analytical frameworks [25]. To our knowledge, bringing together UWMA and actor network analyses within the framework of Learning Labs is a novel contribution to the small but growing body of research that integrates urban metabolism into decision-making processes and strategic planning. Furthermore, our focus on groundwater governance within the UWMA and within the context of drought in the Global South positions this paper as a unique contribution to the evolving groundwater governance praxis in climate adaptation. At this juncture, we posit the value of using UWMA as a process of co-producing a systems-level understanding of groundwater as it relates to urban water at large, as well as bringing diverse stakeholders and actors into a common framework to facilitate knowledge sharing and collaborative learning.

1.3. South African Cities Context

South African cities face the triple challenges of urban growth, under-investment in infrastructure, and extreme climatic events, including severe droughts. This strains water systems and service delivery. Between 2015 and 2018, Cape Town experienced a severe drought [38,39] culminating in the ‘Day Zero’ crisis [40]. Similarly, Nelson Mandela Bay faced severe water restrictions and disruptions due to prolonged drought from 2015 to 2023 [41]. These events stressed the water systems and strained relations between water managers and users across communities, sectors, and government levels. These crises highlighted the complex challenges of sustainably governing water in growing cities with a legacy of spatial and economic inequality and a changing climate. Both cities, traditionally reliant on surface water, now face new governance challenges with the introduction of groundwater.

Historically, groundwater governance in and around Cape Town and Nelson Mandela Bay was focused on large-scale private abstractions for commercial irrigation and some industrial use, mostly concerned with water use licensing procedures and basic monitoring. As drought risk has increased, groundwater has become a pivotal resource for the City governments in their role as water service providers, insuring supply during drought years and becoming more resilient to drought by means of integrating water sensitive city principles into urban planning as a whole [42] (particularly in Cape Town). This means integrating departments and entities such as stormwater management, parks and biodiversity, interest groups, and the general public into the formal groundwater governance landscape, which has traditionally been exclusively the domain of hydrogeologists, farmers, and engineers. Realising the full potential for groundwater to support water-sensitive and drought-resilient cities will require reforms in groundwater governance [43]. This requires shifting from a techno-managerial approach to a holistic, integrative, and inclusive strategy for making and implementing water-related decisions, including groundwater as one component. This research set out to explore how UWMA and actor network mapping can contribute to facilitating this shift.

2. Methods

2.1. Research Design

Two South African cities facing severe drought risk, Cape Town and Gqeberha (formerly Port Elizabeth, hereafter referred to as Nelson Mandela Bay (NMB), which is the metropolitan municipality of the city and its surrounds), were used as case studies to explore how two systems analysis methods, UWMA and governance network analysis, could build a shared understanding of drought risk drivers and impacts, especially on groundwater. These methods were combined in Learning Labs (Figure 1), which brought together stakeholders to co-produce current groundwater-related knowledge of the urban systems. Cape Town and NMB differ significantly in size, climate, water demand, and governance strategies, representing a primary and secondary city, respectively (Figure 2). They were chosen for comparison because both turned to groundwater as they neared a ‘Day Zero’ scenario, in which municipal water supplies became too low to meet basic needs, providing a useful contrast in urban water metabolisms and governance structures in South Africa. The approach used and insights generated provide the basis for exploring issues of adaptive and conjunctive water management in urban contexts across the Global South, where groundwater is an increasing part of the water mix amidst high levels of social, economic, and spatial inequality and weak formal regulatory capacities.

2.2. Learning Laboratories

Learning Labs are facilitated events that bring together diverse stakeholders to engage with complex urban sustainability issues, with the aim of building a shared understanding of system dynamics, adaptation strategies, and various roles and responsibilities for enacting and implementing alternative approaches [44,45]. Rooted in principles of social learning, Learning Labs are facilitation processes that take a variety of forms with the common trait of being problem-oriented, participatory, and interactive. They are designed to be dynamic learning spaces for diverse stakeholders to experiment with ideas and share different types of knowledge to build new understandings and co-design novel solutions [46,47,48].

Between November 2021 and June 2023, five Learning Labs were held: two in-person in Cape Town, two in-person in Nelson Mandela Bay, and a fifth online event (using the Zoom platform) that brought stakeholders from both cities together. People that work on and have knowledge of diverse aspects of urban water (i.e., biophysical, social, and infrastructural) were invited, including hydrogeologists, water activists, water engineers, commercial water users, water planning and management officials from local, provincial, and national government departments, social scientists, landscape architects, and those working in non-governmental organisations brokering water-related networks and collaborations. Participants included individuals coming in their independent capacity and/or as representatives from various organisations and departments involved in groundwater-related activities and responsibilities across the state entities, the private sector, and civil society (see

Table 1. Summary of the number of Learning Lab participants in each city.

Type of Stakeholder	Cape Town	Nelson Mandela Bay
Academics/researchers	11	7
Government officials	4	7
Consultants/business representatives	8	4
Non-governmental/civil society representatives	8	0
Intermediary/network organisation representatives	2	2
TOTAL	33	20

). Although there was widespread representation, academics, hydrogeology/engineering consultants, and government officials made up the largest proportions. Social activists and businesses were least represented. Informed consent was obtained from all participants involved in the Learning Labs.

Interactive group exercises were designed and facilitated to elicit and document diverse perspectives and sources of information regarding groundwater practices and decision-making processes. Appointed notetakers and rapporteurs captured and shared qualitative data that was thematically analysed. The data generated from the Learning Labs themselves (excluding the water budget analyses that were constructed prior to the Learning labs) included: the various actors (and actor type) working in/impacting/impacted by the groundwater space in each city; the functions and capacity of each actor type to fulfil their groundwater-related functions (linked to mandate, staffing, technical expertise, budget, procurement, and partnership modalities); the various areas and levels of in-fluence associated with each actor type; and their perspectives on UWMAs and its value to their respective roles and decision-making processes.

The Learning Labs facilitated connection, knowledge exchange, and collective reflection on the physical flows and social dynamics shaping the urban water system, particularly in relation to groundwater. Through presentations and exercises, participants explored how UWMA could be better communicated, their roles within the broader system, and the information needs UWMA could or could not meet. Group activities encouraged inclusive participation through spoken, written, and visual inputs—such as using coloured sticky dots to indicate actor functions and tiles to represent levels of influence over aspects of groundwater management. The final online event shared findings from the UWMA and governance network analyses, compared the two cases, and invited reflections on strengthening groundwater sustainability in South African cities.

2.3. Urban Water Metabolism Analysis (UWMA)

In the Learning Labs, UWMA was the approach used to iteratively co-build the conceptual and quantitative understanding of urban water cycles, focusing on the direction, rate, and magnitudes of flows. Prior to the labs, a preliminary UWMA was conducted for each city, and then when presented, certain flows were amended or added as discussions took place over the course of the Learning Labs. In brief, the UWMA involves four steps:

(1)

Define the system boundary, which we limit to the metropolitan boundary (see Figure 2);

(2)

Collate data on hydrological (precipitation, runoff, evapotranspiration, and groundwater recharge) and anthropogenic flows (surface supply/dams, desalination, inter-basin transfer, springs, groundwater, and recycled water) at annual timescales. Details for how the various parameters were calculated are given in Sections S1 and S2 of the Supplementary Material;

(3)

Conduct a water mass balance where we assume a steady state and follow the equation below (adapted from Atkins et al. [31]):

Qi = Qo
(P + Bw + Ir) + R + MAR + GW_re = (W + Ru + GW_a + GW_d + Eto +Re + L) − R − MAR − GW_re

where Q_i is the sum of all inputs and Q_o is the sum of all outputs (including losses). P is precipitation, Bw is bulk water supply, which includes surface water (Bw_s), centralised groundwater abstraction (Bw_g), and centralised desalination (Bw_d); Ir is agricultural irrigation (sourced from outside the boundary); R is recycled water (R_p is potable, R_np is non potable); MAR is the managed aquifer recharge; and GW_re is the groundwater recharge from natural processes. W is wastewater, Ru is runoff, GW_a is centralised (internal) groundwater abstraction, GW_d is groundwater discharge, ETo is evapotranspiration, and L is known losses. Centralised groundwater abstraction can be sourced either from aquifers outside (Bw_g) or inside (GW_a) the municipal boundary. Recycled water (R), managed aquifer recharge (MAR), and groundwater recharge are included as both Qi and Qo but are subtracted from outputs as they are re-internalised into the system rather than counted as losses.

For Cape Town, the mass balance presented at the Learning Labs was taken (with permission) from Atkins et al. [31]. For Nelson Mandela Bay, the data sources and methods used to estimate certain parameters are detailed in Table S1 of the Supplementary Material. The full mass balance for both Cape Town and Nelson Mandela Bay are detailed in Table S2, for interests’ sake. The focus of this paper is less on absolute values and more on the process of working with Learning Lab participants to explore the conceptual model of the urban water mass balance. The mass balance was visualised using a Sankey diagram, which was constructed with the help of python code published by Lupton and Allwood [49]. The Sankey diagrams served as a discussion point with the participants in each Learning Lab. Participants were invited to add questions and comments to the diagram, facilitating discussions around the water cycle and to inform how the scenario planning session would be conducted.

There are inevitable data gaps and uncertainties associated with several parameters used in building the water mass balance. For the Cape Town mass balance, details of the data limitations and sensitivity analyses can be found in Atkins et al. [31]. For the Nelson Mandela Bay mass balance, error levels have been calculated and are detailed in Section S3 of the Supplementary Material.

(4)

Apply performance indicators

To assess the metabolic performance of an urban water system, performance indicators were calculated according to Renouf et al. [50] (

Table 2. Details of urban water metabolism performance indicators.

Indicator (Renouf et al. [50])	Method	Unit
Population density	Population/area	capita/km2
Intensity of water use	Total water use/area	kL/d/km2
Urban water efficiency	Centralised supply/population	L/d/capita
Supply Internalisation		%
Indicator (Paul [30])
Wastewater potential for Water supply
Centralised supply replaceability (%)	Wastewater flow/centralised water supplied	%
Total use replaceability (%)	Wastewater flow/total water use	%
Stormwater Potential for Water Supply
Centralised supply replaceability (%)	Stormwater flow/centralised water supplied	%
Total use replaceability (%)	Stormwater flow/total water supplied	%

) for the intensity of water use, urban water efficiency, and supply internalisation. Urban water efficiency is an indicator of the environmental water use of the urban system, expressed as a rate of environmental water withdrawal per inhabitant per year (kL/capita/year). Supply internalisation is the proportion of total water demand that can be met by internally harvested or recycled water. For the scenario analyses, which delved deeper into the principles of water sensitive cities, performance indicators adapted by Paul et al. [30] were also included; these measure the potential to supplement bulk water supply by both wastewater and stormwater.

2.4. Scenario Planning

Scenario analyses were a part of the Learning Lab activities, designed as a knowledge sharing process to promote discussion and dialogue amongst multi-stakeholder participants [51]. Scenarios outline plausible future pathways that aid decision-making in the face of uncertainty [52]. Scenario planning has been widely applied in studies related to climate change, land use, food security, energy use, and, to some degree, urban water management (see Sivagurunathan, Elsawah, and Khan 2022 [53] for an extensive list of references) and has been shown to enable learning across diverse stakeholders [54]. It can also play an important role in robust decision making [53], especially when ‘surprising’ or uncertain possibilities and scenarios are explored.

Together with the Learning Lab participants, we examined potential future changes in urban water cycles using scenarios reflecting changes in climate, land use, and management decisions. The focus was on using simple, plausible scenarios to promote discussion and deeper understanding of urban water cycles, land-use, and climate change, rather than on predicting specific outcomes. Scenario analyses in the Learning Labs were designed to gain a better understanding of the urban water cycles and to engage with the usefulness of the UWMA in supporting decision makers of various kinds (i.e., making water policies and bylaws, deciding on infrastructure upgrading priorities, designing specifications for wellfields, implementing water demand restrictions, assessing land use applications, etc.).

An example of the scenario planning exercise explored in the Learning Labs for NMB is provided in

Table 3. The three scenarios of various water supply options for the Nelson Mandela Bay Metropolitan. (1) The current scenario, representing the 2022 water mix (before any major interventions), (2) the future scenario, reflecting planned supply options as sourced from Zutari, and (3) the idealised water mix, reflecting the principles of a water-sensitive city.

Source	Current (Mm3/a)	Future (Mm3/a)	Idealised (Mm3/a)
Surface water	71.3	71.3	50.8
Natural spring	2.2	2.2	2.2
Water transfer	58.4	76.7	58.4
Groundwater		20.4	20.4
Desalination	11.0	27.4	11
Recycled water (Coega)	1.7	21.9	27.50
Recycled water (NMU)		1.3	1.3
Recycled water (Drinking)		3.7	13.7
Managed Aquifer Recharge (MAR)			85.71
Loss recovery		7.3	20
Total External Inputs	142.8	197.9	142.9
Total Internal flows	1.7	34.1	148.21

. Three separate water cycles were quantified in order to assess how each scenario fared in terms of being water sensitive: (1) the current scenario, representing the water mix as of 2022 before any major interventions have came online; (2) the future scenario, representing the future water mix as sourced from Zutari (an engineering consultancy working in NMB); and (3) the idealised scenario, a hypothetical water mix according to the principles of a water sensitive city. In practice, these principles manifest through the internalisation of water supply, the recovery and fit-for-purpose use of wastewater and stormwater, and the strategic integration of these flows within water-sensitive urban design to enhance aquifer recharge and support ecosystem restoration. The idealised scenario considers Managed Aquifer Recharge (MAR) as an important tool in becoming a more resilient, water-sensitive city. For the sake of clarity, we report on NMB in this manuscript. The details for scenarios explored in Cape Town are provided in Table S3 of the Supplementary Material.

2.5. Co-Producing an Understanding of Governance Networks

The Learning Lab exercises aimed to support the organised sharing and recording of information to collectively interpret and integrate knowledge held across the network at the organisational and sub-organisational scale. Attention was directed toward mapping actors in the urban groundwater sector; their interconnections through data, financial, regulatory, and collaborative flows; and their varying capacities and degrees of interaction in exercising agency over groundwater.

Using the participatory Net-Map method [55], participants identified organisations and departments involved in groundwater use, protection, and regulation by naming actors on cards (Table S4 of the Supplementary Material). They then assessed relative influence by placing each card within concentric circles or building towers of tiles, and mapped the interactions by drawing colour-coded lines with arrows between the actors, showing the flows of data, advice, finances, authorisations, and partnering activities. The colour-coded stickers on each card indicated which of the four functions the actors fulfilled, namely: (1) understanding (aquifer delineation, yield estimation, groundwater protection zoning, and monitoring); (2) operating (borehole installation, wellfield operation, and aquifer recharge management); (3) regulating (water use licences and enforcement, borehole registration, (ground)water bylaw—preparation and enforcement, usage restrictions, and groundwater protection zone enforcement); and (4) capacitating (training, awareness raising and education, advocacy, and partnerships). Finally, actors’ capacities to fulfil their groundwater functions were scored based on the clarity and extent of their mandate, the number of staff they have working on groundwater programmes, the levels of technical expertise held, the efficiency of modalities to leverage capacities outside of the organisation (through procurement or partnering), and the budget for groundwater programmes.

Actor network analyses for each city identified key actors at organisational and sub-organisational levels and their influence over water flows. To assess the decision-relevance of information from UWMA, a questionnaire in Learning Lab 2 (Table S5 of the Supplementary Material) gauged participants’ understandings of groundwater’s role in UWMA and the value of UWMA in their decision-making.

3. Results

The results present key contrasts between Cape Town and Nelson Mandela Bay (NMB) in terms of their urban water metabolisms, groundwater governance networks, and the decision-making relevance of urban water metabolism analysis (UWMA). Together these analyses, and the data collected throughout the Learning Lab process, provide the opportunity to offer a high-level decision typology of the value of UWMA at various scales of decision-making.

3.1. Contrasting Urban Water Metabolisms

The Sankey diagrams (Figure 3) illustrate water flows for each city, highlighting significant differences in the hydrological cycle (blue) and the anthropogenic (brown) flows of the water cycle pre-drought intervention (also refer to Table S1 in the Supplementary Material for a detailed mass balance). In Cape Town, rainfall contributes to runoff (39%), groundwater recharge (46%), and evapotranspiration (14%), whereas in NMB, evaporation is much higher (86%), with minimal recharge to groundwater (5%) and runoff (13%). Anthropogenic flows vary greatly between the two cities. Cape Town relies heavily on surface supply, which is supplemented by desalination and groundwater in response to the 2018 drought. Water re-internalisation (in green) involves direct reuse and managed aquifer recharge (MAR) in Atlantis aquifers, along with non-potable reuse strategies, mainly for irrigation and industrial uses. In contrast, NMB’s anthropogenic flows include surface storage, desalination, natural springs, and inter-basin transfers, primarily for industrial purposes. Wastewater re-internalisation occurs in the Coega industrial development zone, but managed aquifer recharge is not yet a feature of the water cycle, as it is in Cape Town. NMB’s water system interacts with neighbouring municipalities and supports agricultural use within its boundaries. Agriculture was not considered in Cape Town due to data limitations. Pre-drought, NMB water supply was already more diversified than Cape Town. These Sankey diagrams were created pre-Learning Labs. The data (the overall mass balance) used to build them are detailed in Table S1 in the Supplementary Material.

The performance indicators (

Table 4. Contrasting urban water metabolisms for Cape Town and Nelson Mandela Bay (NMB) by means of a few highlighted performance indicators, as developed by Renouf et al. [28].

Indicator	Method	Unit	Cape Town	NMB
Population density	Population/area	capita/km2	1730.5	643.2
Intensity of water use	Total water use/area	kL/d/km2	418.0	208.9
Water Efficiency	Centralised supply/population	L/d/capita	210.0	310.6
Supply Internalisation	Internal supply/total supply	%	13%	5%

) show that population density is far higher in Cape Town (1730 people per km²) than in NMB (643 people per km²), and the intensity of water use is double in Cape Town (418 kL/d/km²) compared to NMB (208 kL/d/km²). However, water use efficiency is greater in Cape Town, likely reflecting the higher population served by the City as well the larger disparity in water use across the socio-economic scale (e.g., a large portion of Cape Town’s population do not have access to water in their own dwelling and thus use significantly less water). Supply internalisation is 13% in Cape Town and 5% in NMB, which reflects the managed aquifer recharge operations in Atlantis as well as water reuse in Cape Town.

3.2. Example Scenario Analysis

The performance of NMB urban water metabolism was assessed (

Table 5. Results of the scenario analysis explored in the Nelson Mandela Bay Learning Labs, using performance indicators from Renouf et al. [28] and Paul et al. [30].

Indicator (Renouf et al. [50])	Method	Unit	Current	Future	Idealised
Population density	Population/area	capita/km2	643.2	643.2	643.2
Intensity of water use	Total water use/area	kL/d/km2	208.8	290.9	228.0
Water Efficiency	Centralised supply/population	L/d/capita	310.6	430.3	310.6
Supply Internalisation		%	5	14	27
Indicator (Paul [30])
Wastewater potential for Water supply
Centralised supply replaceability (%)	Wastewater flow/centralised water supplied	%	36	17	11
Total use replaceability (%)	Wastewater flow/total water use	%	35	16	10
Stormwater Potential for Water Supply
Centralised supply replaceability (%)	Stormwater flow/centralised water supplied	%	96	68	26
Total use replaceability (%)	Stormwater flow/total water supplied	%	92	66	25
Wastewater and Stormwater Combined
Potential of total water use replaceability (%)	(Wastewater + stormwater)/total water use	%	126	82	36

) according to the principles of water sensitive cities, using indicators developed by Renouf et al. [28] and Paul et al. [30] (Table 2). Urban water efficiency is currently 310.6 L/d/capita, which increased substantially under the future water mix to 430.3 L/d/capita, reflecting the drought response to augment supply with additional groundwater, inter-basin transfer, and desalinisation. In our idealised scenario, urban water efficiency does not improve from the current scenario and is reflective of the choices made in the scenario to augment supply by focusing on managing internal sources (recycled water and managed aquifer recharge) rather than reducing external sources. Urban water efficiency is essentially about reducing the extraction of environmental water external to the system boundary, which could be effectively achieved through several interventions, such as increasing supply internalisation (discussed below), reducing system losses, and implementing effective demand management strategies. Mitigating the large losses reported by NMB in both apparent and real losses, as well as remedying unbilled/unauthorised consumption, would substantially improve water efficiency. Reducing demand management is another point of focus, especially in times of water crisis. The NMB New Dashboard (20 September 2022) indicated that NMB was not meeting its consumption targets of 50 L per day per person.

Supply internalisation, the proportion of total water demand that can be met by internally harvested or recycled water, is 5% under the current water mix and represents current water recycling for Coega Industrial Zone. With the Future Water Mix planned and in progress, supply internalisation would increase to 14% with the expansion of water recycling for non-potable (23.2 Mm³/a) and potable (3.7 Mm³/a) reuse. Under an idealised water-sensitive scenario, with increased recycling for both potable and non-potable water, rainwater harvesting, and the introduction of Managed Aquifer Recharge using wastewater effluent and/or stormwater, internalisation could increase to 27%. Currently, the potential for stormwater and wastewater effluent to replace supply is 126%, indicating enormous potential to integrate these sources into water management plans.

3.3. Contrasting Groundwater Governance Networks

The difference in size between the two cities is reflected in the number of actors functioning in each city’s groundwater space, with Cape Town having a larger number of actors (44 identified) compared to NMB (30 identified). More non-governmental organisations and scientific research organisations are actively working on groundwater issues in Cape Town than in NMB. More multi-stakeholder forums organized around sharing information and building partnerships to address groundwater matters are also active in Cape Town, notably the Table Mountain Strategic Water Source Partnership and the Aquifer Monitoring Committees, set up around the bulk abstraction wellfields installed by the city government. State actors in Cape Town fulfil more roles across various functions and scales than any other category, including financing, groundwater exploration and exploitation, planning, regulation, and research.

The national Department of Water and Sanitation (DWS), with regional offices in each Province, holds the legal mandate to act as the custodian of all water resources, including aquifers. Their operational focus when it comes to groundwater has largely been on estimating water available to be allocated and licensing bulk-users in the agriculture, mining, and industrial sectors, as well as municipal water service providers. The DWS also coordinates the updating of Reconciliation Strategies for each of the supply systems that the two cities are part of, i.e., the Western Cape and Algoa Supply Systems. The Reconciliation Strategies determine the water balance in terms of cumulative supplies and demand and develop future scenarios considering the sequencing of options to balance changes in demand with variable supplies.

The city governments are represented on the associated Steering Committees. In both cases, further groundwater development has been identified as priority interventions for augmenting supplies, but the resourcing to pursue these options has been slow to materialise. The ‘Day Zero’ water crises in both cities brought these options to the forefront and considerable investments were flowing into municipal groundwater developments, forcing engagements between local, provincial, and national government entities over groundwater use licensing and protection matters. Although strong ties were identified between government entities and technical consultants who undertake groundwater exploration, design, installation, and monitoring work, there was a notable gap identified during the actor mapping in government engagement with water users in the commercial and domestic spheres around the sustainable use of urban groundwater and enforcing groundwater regulations pertaining to registering of boreholes, volumes of abstraction, monitoring, and reporting. Some intermediary organisations and consultancies (e.g., Green Cape, WWF-SA, GEOSS, Umvoto, and the Business Chamber in NMB) were identified as stepping in to this brokering and information-provision role, as well as building partnerships and lobbying to strengthen governance mechanisms, such as multi-stakeholder aquifer monitoring committees in the case of Cape Town.

In both cities, local consultants and academics were identified as being involved in the research, planning, monitoring, regulation, and exploitation of local groundwater resources, with a denser network based in Cape Town than in NMB. Fairly strong capacities and ties were identified amongst the research and consulting actors; however, a lack of regular data sharing and integration was noted as a major constraint on developing an integrated understanding of groundwater-connected systems. NGOs appeared to fulfil different functions in each city. At the time of this research, environmental NGOs in NMB were not yet active in groundwater distribution and protection issues. Instead, the main NGO active in NMB was the humanitarian relief organisation, Gift of the Givers, which had been installing boreholes at schools and clinics in poor neighbourhoods badly affected by water shortages and outages, where government and private sector interventions were not forthcoming. In contrast, in Cape Town, it was identified that WWF-SA had taken an active role in convening people around groundwater concerns, establishing citizen science projects around monitoring groundwater levels and promoting awareness raising, including through faith communities and schools, actively working with the Green Anglicans movement and GreenPop, a community-based tree planting initiative. Interestingly, this work was partially funded by a private company in the beverage manufacturing industry (AB InBev) with a shared interest in protecting local groundwater sources, in addition to the Embassy of Denmark in South Africa. Similarly, the Philippi Horticultural Area (PHA) Food and Farming Campaign, a grassroots organisation, has been working with the legal support NGO Natural Justice to raise awareness around the importance of the Cape Flats Aquifer for Cape Town’s resilience in terms of water, food, and climate, advocating for more protection measures to prevent destructive land uses in the aquifer catchment. This highlights the growing variety of actors with different interests and capacities engaging in groundwater-related matters that shape the material flows of water represented in the UWMA.

Participatory network mapping during the Learning Labs revealed numerous weak ties among actors working on groundwater issues across the public, private, and civil society sectors. The groundwater governance network appears fragmented, with geohydrology and engineering consultants positioned as key bridging nodes. Challenges related to enforcing licensing rules, usage restrictions, and infrastructure security emerged as key concerns in both cities, fuelled by high levels of mistrust, low confidence in government capacities, and weak communication channels regarding groundwater matters.

3.4. Urban Water Metabolism Decision-Relevance Typology

The Learning Labs aimed to identify where urban water metabolism analysis could best support groundwater decision-making and governance. A decision-relevance typology emerged from this process as a way of linking biophysical and governance aspects through the lens of decision-making. From responses to a questionnaire and insights from presentations and exercises in the Learning Labs, we identified key groundwater-related decisions at scales ranging from borehole/wellfield to city-region. These decisions were categorised into activities: groundwater exploration and exploitation; the maintenance and monitoring of groundwater infrastructure; the regulation of groundwater use and aquifer protection; and the integration of groundwater into planning and management. The decision typology presented (Figure 4) is a reflection from both cities with the intention that it could be applicable to any city context.

Based on all answers provided by the Learning Lab participants, UWMA was considered least useful for finer-scale decisions, such as exploration and exploitation at borehole, wellfield, or even aquifer scale. Decisions around maintenance and monitoring were considered most supported by UWMA at the aquifer and city-region scale and only modestly at the borehole/wellfield scale. The collection and management of data or the design of monitoring programmes was considerably supported at the aquifer and city-region scale, but only modestly at the borehole/wellfield scale. The UWMA is perceived to support decision-making in terms of regulation and protection only at the aquifer catchment and city-region scale and only modestly at the aquifer scale. For the aquifer and wellfield scales, a hydrogeological model would more appropriately capture aquifer specifics.

For integration across multiple actors, UWMA was seen as beneficial at all scales. Learning Lab participants noted its usefulness to actors with more interest in the higher-level strategic picture of urban water management (Supplementary Material Table S4). UWMA can help conceptualise and visualise the system, exploring how it might change, where to intervene to create more circular flows, or how to diversify supply mix, for example. It is also perceived as very useful in terms of engagement around the UWMA; creating conversation around future scenarios, bringing people together, and also collectively identifying the unknowns in the system. However, it was considered too academic, needing clearer social components and more accessible language for broader stakeholder understanding.

4. Discussion

The UWMA and governance network analyses of Cape Town and Gqeberha (NMB) show how urban water cycles and governance dynamics are shaped by their bio-physical and socio-political landscapes. Both cities face drought risk but differ significantly in size, population, climate, water supply systems, and political and economic contexts. These differences are reflected in their water metabolisms, measured by metrics such as water use intensity or supply internalisation, as well as the extent to which groundwater is considered within the urban system and the water management objectives at large. Cape Town’s governance network is more populated, with more entities across various actor types (e.g., state, academic, NGOs, and private companies) than NMB. This likely reflects Cape Town’s advanced stage of drought-risk adaptation and larger size. A city’s size, resource flows, and socio-economic network density are key indicators of its adaptive capacity [56,57].

4.1. Reflections on the Value of Exploring Scenarios

The scenario planning exercises during the Learning Labs were an important tool facilitating active engagement with the urban water cycles and the impact of various decisions on them. The scenarios explored were quantitative, considering factors such as decreased rainfall, increased evaporation and stormwater harvesting potential, spurring engagement, and conversations across broad interests and backgrounds. These sessions highlighted the heuristic value of engaging with urban water systems management within a context of uncertainty.

As decision-making becomes more participative and inclusive, scenario planning must evolve to reflect diverse and potentially unexpected future possibilities. Claasens and colleagues [58] engaged the South African water sector stakeholders to assess the sector’s capacity to handle complexity and uncertainty. The knowledge gained through this process was considered empowering for the water sector. By engaging in participative governance, decision makers were equipped with insights into challenges that the water sector may face in the future. Quantitative approaches (e.g., Gxokwe, Xu, and Kanyerere [59]) are increasingly being integrated with qualitative narratives of plausible futures, such as the Seeds of the Good Anthropocene [60] and Climate Risk Narratives [61]. These narrative-based approaches are considered particularly effective for governance-focused engagements such as Learning Labs. Upon reflection, insufficient time was given to this aspect of the Learning Labs in this study. We recommend that future processes aiming to build adaptive governance capacities place far greater emphasis on collaborative scenario conception and planning, which requires more or longer Learning Lab sessions.

4.2. Decision-Relevance of Urban Water Metabolism

Effective groundwater governance requires the co-operation and coordination of multiple actors operating at various scales, aiming for transparent, participatory, and adaptive decision-making [19]. Modern urban water management now includes actors well beyond traditional water supply roles, integrating urban planning, waste-water management, stormwater, parks and recreation, and public health, as well as academic, private, commercial, and civil society sectors. The Learning Labs highlighted the challenges of integrating this diverse group, each with varying understandings and roles related to groundwater.

The urban water metabolism approach is suggested to foster institutional collaboration and cross-sectoral integration [19], yet the mechanisms by which this may happen are not clearly defined. The Learning Labs built and refined the conceptual understanding of the urban water system, identifying the who and the how of influential actors in groundwater-related activities, and assessed which decisions UWMA can support. The descriptive typology provided in Figure 4 indicates that UWMA supports decisions that require a high level of integration, here considered the decisions/actions that require/affect the activities of multiple actor types. It is most useful for more large-scale strategy, a point that was also highlighted by King and colleagues [18]. Decisions surrounding borehole/wellfield specifics are considered too particular for the UWMA to support.

Engagement with UWMA in the Learning Labs showed its value in participatory decision-making and adaptive governance through the following:

• Identifying knowledge needs and gaps. Governing groundwater involves diverse roles and responsibilities. UWMA supports urban planning in terms of understanding hydrological flows and opportunities to diversify and internalise water supply sources, as well as to quantify water use efficiency [29]. Metrics, performance indicators, and scenario planning help decision-makers identify knowledge needs for each role and responsibility and knowledge gaps, currently and what might be needed to make progress against water management aspirations. Transformative governance will create new roles requiring new knowledge and UWMA can identify these needs for drought preparedness and resilience.
• Bridging knowledge sources. Diverse types and sources of knowledge surrounding groundwater can make important contributions to its governance. Bridging knowledge sources has been defined as ‘maintaining the integrity of each knowledge source while creating settings for the two-way exchange of understanding for mutual learning’ [62]. Groundwater governance is a process of both vertical (from aquifer dynamics through to policy) and horizontal (e.g., traditional and scientific) knowledge integration. The failure to effectively integrate or weave [63] these knowledge types and sources across relevant scales may be an important reason why governance may fail to produce sustainable and adaptive outcomes. Learning Labs provide a fertile setting for social learning, where participants iteratively reflect on system perspectives and their contextual experiences. UWMA is a tool to help structure the facilitation of knowledge exchange and co-production in ‘a process of generating, sharing and/or using knowledge’ [64]. Social learning, considered a ‘process of iterative reflection that occurs when experiences are shared with others’ [65], is emerging as an important concept for collaboration, joint decision-making, and adaptive governance and co-management [66], and is at the heart of iterative rounds of problem solving. Weaving knowledge across different levels of organisation requires establishing mutual respect and trust between diverse actors [47,65].
• Promoting stakeholder cohesion and cooperation. Social cohesion, defined by trust, belonging, and cooperation, is crucial for sustainable resource management and governance [67]. The links between resource management (and, in turn, governance) and social cohesion are considered to be reciprocal [68], where social cohesion is viewed as a prerequisite for sustainable resource management and, in turn, resource management as an entry point for social cohesion. As groundwater is a distributed hidden resource, factors such as trust and willingness to cooperate become crucial in decision-making and the successful implementation of, or compliance with, those decisions. Trust and reciprocity have also been identified as enabling prerequisites for mobilising the adaptive capacities that emerge during a drought response to support transformative change [69]. A common goal, such as improving systems understanding or ultimately to make water management more sustainable and equitable, serves as a ‘collective directional movement’ [70], which is an important aspect of social learning. Designing processes and procedures that can promote cohesion across the various actors will greatly facilitate collaborative decision-making agendas and cooperative implementation. The climate change adaptation and water governance (CADWAGO) project [20] is a good example of the value of designing processes that enhance learning across various sectors. The project highlights that face-to-face and online events can be designed to support learning for water governance transformations. Participants in their multiple events found the processes helpful for thinking about how to work in a more systemic way across levels and boundaries, whether organisational, institutional, geographical, or cultural in nature.

4.3. Research Limitations

Building a complete water mass balance of an urban system requires a substantial array of data types and sources. In contexts where municipal data are not stored in data centres or are accessible only with special permissions (as is often the case in the Global South and particularly amid crises), obtaining the data can be a time-consuming process. In the case where not all the data are available, we recommend Learning Labs to focus on relative magnitudes of flows rather than absolute values. When applied in the context of informing urban management decisions, the value of the UWMA is its role in co-building a shared conceptual understanding of the urban water system rather than providing precise mass balances. A limitation of the Learning Lab approach is achieving representation of all key constituencies in the participatory process and having enough consistency of participants across events to build trust and deepen a shared understanding of the system that weaves together multiple perspectives.

5. Conclusions

A central challenge in contemporary urban water management is developing inclusive, transparent, and reflexive decision-making processes that address the system as a whole across diverse roles and responsibilities. This research combined two systems enquiry methods—UWMA and governance network mapping—through Learning Labs to assess UWMA’s potential to foster institutional collaboration and cross-sectoral integration. Although UWMA remains a technocratic and academic tool when used alone, its value for real-world decision-making increases when applied in participatory settings, such as in Learning Labs. We proposed a typology of decision types that UWMAs can support, showing that the approach is particularly useful for large-scale (e.g., municipal) and highly integrated decisions involving multiple actor types. In integrative settings, UWMA helps identify knowledge needs, connect diverse groundwater managers, and strengthen stakeholder cooperation. Scenario planning proved especially valuable, encouraging participants to engage deeply with how different climate and management futures might affect the urban water system. We recommend that future decision-making processes emphasize collaborative scenario planning. Finally, intermediary or network-based organisations are vital for sharing information and mediating interests among public, private, and civil actors. Investing in their capacity to promote aquifer health and negotiate sustainable groundwater use can help rebuild trust and reduce governance fragmentation.