Mitigating flash flooding in the city: Drain or harvest?

Flooding fundamentally results from heavy rainfall and snowmelt. Flash flooding occurs often yearly in many cities, negatively impacting human lives through anxiety and post-traumatic stress, depression, drowning, water-borne and rodent-borne diseases, homelessness, physical injuries, and vector-borne diseases (Nkwunonwo 2016; Egbinola et al. 2017; Echendu 2020; Legese & Gum 2020). Causes of flooding include climate change (e.g., more frequent and severe rainfall events), poor land use planning due to lack of awareness of flood risk, limited efforts towards flood disaster risk reduction, rapid population growth, and rapid urbanization (Gwenzi & Nyamadzawo 2014; Nkwunonwo 2016; Valdez et al. 2016; Egbinola et al. 2017; Prakongsri & Santiboon 2020). The factors influencing the severity of flooding include: (i) the amount of precipitation that falls in the watershed and tributaries, (ii) intensity of rainfall relative to infiltration rate, (iii) land cover and land use changes (e.g., grey infrastructure, vegetation, paved surfaces), (iv) nature and structure of the soil, (v) size and shape of the catchments, and (vi) topography of the catchments (Gwenzi & Nyamadzawo 2014; Han & Nguyen 2018; Fioramonte et al. 2022; Vitale & Meijerink 2023; Waseem et al. 2023).

The diversity of causes of flooding and factors influencing its severity suggests that predicting floods and their impacts is a complex task. In fact, mitigating flood impacts implies a flood alert system managed in collaboration with a meteorological forecasting agency. In many countries of Sub-Sahara Africa (SSA), such systems are not available or lack monitoring equipment and capacity, and thus, mitigation strategies are very poor (Egbinola et al. 2017; Legese & Gum 2020). This sad situation seems impossible to address without high financial investments. Fortunately, local knowledge on water capture and conservation exists, including bunds, check dams, infiltration pits/ponds, percolation tanks, storage tanks, and terracing (Hasse 1989; Vavrus 2003; Adams et al. 2010; Kim et al. 2016; Woodhouse et al. 2017; Silayo & Pikirayi 2023). This study postulates that coordinated strategies to reduce flooding are realizable combining local rainwater harvesting (RWH) knowledge and recent scientific evidence. In this effort, the involvement of all stakeholders and communities in the project cycle from conceptualization, design, implementation, operation and performance monitoring, and evaluation is fundamental. This knowledge can be complemented by scholarly engineering tools to maximize the volume of harvested, infiltrated, and stored water.

The present study is a qualitative conceptual attempt to provide an affordable and applicable answer to questions regarding flood mitigation in developing countries in the absence of (i) reliable data for flood modelling and (ii) capitals for financing corresponding infrastructures. The study is driven by two key issues: (i) to demonstrate the key role of RWH in flood risk reduction and (ii) to promote awareness of rainfall as a resource to be harnessed. The slogan is ‘Flash flood = non-harvested rainwater’ and is related to an old Sri Lankan vision of not allowing ‘a single drop of water falling as rain flow into the sea without being used for the benefit of mankind’ (Shannon & Manawadu 2007; Nya et al. 2023). Flooding or excess rainfall occurs when rainfall intensity exceeds the infiltration rate of the soil (Umar & Gray 2023). This implies that avoiding the runoff of this excess rainfall automatically avoids flooding and related environmental consequences.

The presentation starts with a description of the hydrologic cycle (Section 2), followed by an overview of the current perception of the potential of RWH to address flash flooding (Section 3). Section 4 conceptualizes the future of water management rooted in RWH, and Section 5 prepares its realization. Section 6 presents some aspects of how the new approach will increase the living standards in the city and formulates some recommendations for its realization. A short conclusion (Section 7) closes the presentation.

The hydrologic cycle begins with precipitation or rainfall. Once rainwater reaches the ground, for example, at the top of a hill (Figure 1), a fraction is infiltrated into the soil to form soil water, while part of it forms groundwater, a second fraction is evaporated back into the atmosphere (next rain), and the last fraction runs off. The first fraction is divided into two parts: (i) ‘permanent’ groundwater or aquifer and (ii) outflowing groundwater progressively forming springs, and baseflow in streams and rivers and ending into the ocean (Taghavi-Jeloudar et al. 2013; Bensi 2020). Permanent groundwater includes ‘fossil’ groundwater characterized by long residence times of more than 12,000 years (GebreEgziabher et al. 2022). Lakes can originate from both ‘permanent’ and outflowing groundwater, while evaporation occurs at the surface of all open water bodies, including lakes, ponds, and oceans. The last source of water vapour in the atmosphere is evaporation from soil and transpiration (evapotranspiration). When moist air is lifted, it cools, and water vapour condenses to form clouds and returns to the Earth's surface as precipitation (Figure 1). It is very important to recall that, in the short term, non-‘fossil’ groundwater seeps its way into springs, streams, rivers, and oceans. In other words, springs, streams, and rivers are just ‘delayed’ runoff from rains temporarily stored in aquifers then later released as baseflow. In coastal areas and even on islands, there can be saltwater intrusion if the groundwater level is lowered too much by excessive pumping (Henriquez 1962; van Buurt 2018; Pembe-Ali et al. 2021). The present study is focused on direct runoff or excess rainfall which is regarded as the main cause of flash floods in urban areas. The ‘direct’ runoff empties into lakes, rivers, streams, and lastly back to the oceans, where evaporation occurs to form the next rain (Figure 1).

The natural water cycle has been regulating life on the earth for millions of years. Five major benefits of this cycle are aquifer recharge, climate regulation, erosion control, flood control, and freshwater supply (Serrao-Neumann et al. 2017; Shahid et al. 2018; Shahid et al. 2021). For example, when water evaporates, it absorbs heat energy from the sun to cool the environment. Precipitation helps replenish the soil moisture, which is essential for plant growth and crop production. The current global urbanization movement originating from Europe during the 1830s has significantly impacted the natural water cycle in urban areas (Brown et al. 2009; Wong & Brown 2009; Angrill et al. 2012; Nkiaka 2022). This is because the construction of characteristic grey infrastructures with impervious surfaces (e.g., airports, buildings, bus stations, roads) has profoundly modified I and R values in Equation (1). The prevalence of grey infrastructures in cities has led to various social-economic and ecological implications that have in turn led to a rethink in the way to promote resilient urban systems, starting with better approaches of water resources management with RWH as a main pillar (Brown et al. 2009; Wong & Brown 2009).

Conventionally, Q is calculated based on the water demand. For example, the calculation of Q_potable, Q_toilet, and Q_washing considered the number of occupants of a residence, while the calculation of Q_irrigation is made considering the area of a garden and the crop water requirements (Sanches Fernandes et al. 2015; Vialle et al. 2015; Ferreira et al. 2023). The irrigation water requirements depend mainly on the type of crop and weather conditions. This work, on the contrary, advocates for harvesting the maximum possible volume of water at the residential level. The number of occupants and their water consumption then determine the volume of overflow or excess water to be directed to community tanks.

There is currently confusion in the scientific literature concerning the way to design the RWH infrastructure. This confusion results from the evidence that Equations (1)–(3) have been mostly considered independently (Amos et al. 2013; Vialle et al. 2015; Serrao-Neumann et al. 2017; van Ginkel et al. 2018; Huang et al. 2021a; Ferreira et al. 2023). In fact, while Equation (1) advocates for systematic RWH to mitigate the reduced infiltration caused by artificial impervious areas, the literature uses Equations (2) and (3) to discuss whether rainwater should be harvested.

An increasing body of literature exists on the design of RWH for various settings (Hasse 1989, Fonseca et al. 2017; Nguyen & Han 2017; Han & Nguyen 2018; Han 2020; Semaan et al. 2020). Thus, a detailed discussion of the design procedure is beyond the scope of the present study. Briefly, the available design approaches for RWH can be broadly grouped into two:

• (i)

  One for areas with long-term high-resolution rainfall data (daily) (e.g., developed countries). An example of such a design approach is described in earlier studies (e.g., Fonseca et al. 2017; Han & Nguyen 2018; Semaan et al. 2020).

• (ii)

  The other one is for data-scarce settings, ideal for low-income countries. An example of how to design an RWH is settings with limited or scarce rainfall data is presented in the study by Nguyen & Han (2017). This approach is appropriate for low-income regions such as Africa with sparse or no hydrometric stations (Cheo 2018; Nya et al. 2023).

Studies have also evaluated the performance of RWHs for water supply reliability (Amos et al. 2018; Bak et al. 2020; Lee et al. 2021). For evaluating water supply reliability, two variables are often used: (i) rainwater utilization ratio (RUR) and (ii) number of no-water days (NWD) (Mun & Han 2012; Mwamila et al. 2015). RUR also referred to as rainwater use efficiency is the ratio of the amount of rainwater supplied to users to the total rainwater harvested from a given catchment area (Bak et al. 2020). NWD, which measures reliability, refers to the number of days in a year when stored water is less than the 1-day demand. NWD is related to the concept of security, which is the average number of days in a given period (e.g., a year) with zero water supply (Umapathi et al. 2019). Other studies have also developed and applied various parameters to evaluate the flood mitigation potential of RWH (Freni & Liuzzo 2019; Cristiano et al. 2023).

RWH has not yet been addressed on a holistic basis. Equation (1) suggests ‘harvest whenever you can’ or harvest rainwater because it is a natural resource like coal (‘white coal’) or gold (‘blue gold’). Clearly, harvesting and even recycling rainwater should be regarded as resource conservation par excellence, while using the best available technology. Instead, reasons currently justifying RWH while using Equation (2) include: (i) economic feasibility (‘harvest if it is affordable’), (ii) water conservation (‘harvest it for non-potable uses only’ – citizen's option), (iii) as an alternative in areas lacking centralized water supply (e.g., mountain communities, remote areas, rural communities), (iv) as an alternative in areas lacking non-polluted surface water (e.g., presence of As, F, salinity, or U), and (v) to avoid erosion (environmental protection) and flooding (emergency relief). These five reasons correspond to the motivation of individual stakeholders to harvest rainwater (Hasse 1989; Han 2013a, 2013b, 2020; Fu & Butler 2021). Whenever it rains, a more or less large volume of runoff is not harvested. The fundamental question is ‘Why invest money for networks and hydraulic structures to collect and channel rainwater to the river/sea rather than to harvest and use it?’

This question is alien to a Western-educated person as recently recalled by Gwenzi et al. (2023). In fact, with the ‘sanitary revolution’ or ‘hygiene movement’ of the 19th century, rainwater was relegated to something less than waste. This is because rainwater was a disturbing factor in efforts to design reliable sewerage systems (Cook 2001; Hughes 2013; Han 2013a, 2013b; Geraldi & Ghisi 2017; Snir et al. 2022). This view completely contrasts the old wisdom that rainwater is the least polluted water source on the earth (Davis 1891; Han 2013a, 2013b; Han & Nguyen 2018; Akkerman 2020; Huang et al. 2021a; Nguyen et al. 2021).

The conventional approach, known as urban drainage, to avoid the ‘direct’ runoff causing flash flood in cities has been to drain rainwater as quickly as possible to the next water course (river or stream) (Nichols 1885; Parry 1885; Hughes 2013; Snir et al. 2022; Gwenzi et al. 2023). This solution requires proper urban planning and should be dynamic as cities and poorly planned and unplanned urban settlements are growing almost randomly (urban sprawl) and more areas are made impervious (Lal 2015a, 2015b; Nkiaka 2022; Demeke & Amali 2023). While engineering tools and political will may exist to solve the arising problems, there is a lack of financial support in low-income communities in SSA and in the MENA countries (Wahab & Ojolowo 2013; Celik et al. 2017; Loudyi & Kantoush 2020; Echendu 2022; Goetz et al. 2023). Therefore, alternative, affordable, and applicable solutions are urgently needed. A good and immediately realizable idea is to incorporate indigenous and modern RWH techniques into urban flood management. Making flood management the main target (disaster prevention or emergency relief) will sustainably enable access to domestic water, promote urban agriculture, promote household food security, and combat poverty (Hasse 1989; Han 2023a, 2023b). This is because disaster prevention is clearly one of the regalian functions of the State (protection of citizens) and transcends various spatial planning scales (at communal, regional, and national levels), while access to clean water, food security, and poverty reduction can be mistakenly considered as charitable interventions, particularly in small cities and rural areas. The regalian task here entails that a sovereign State has the role to protect its population (domestic security) (Chopin & Jamet 2016).

Section 2 has suggested regarding rain as the main water source on the earth, which means rainwater is the ‘mother’ of both river and groundwater on which the conventional water supply relies. Regarding rain as the mother of all waters (Figure 1) implies a revision of conventional considerations, restricting RWH to a tool to diversify local water sources (Agarwal & Narain 1999; Farreny et al. 2011; Snir et al. 2022; Soh et al. 2023; Vitale & Meijerink 2023). The view presented herein corresponds to the Kilimanjaro concept, advocating for the creation of appropriate water infrastructures culminating in drainage networks for harvested rainwater (Marwa et al. 2018; Qi et al. 2019; Nya et al. 2023). The design of such networks can be adapted from conventional systems for collecting wastewaters (Moss 2000; Howe et al. 2012; Demeke & Amali 2023) and have the potential to mitigate water stress related to the population growth and climate change (Agarwal & Narain 1999; Cain 2014; Echendu 2020; Soh et al. 2023).

One key issue in discussing the suitability of rainwater for domestic uses has been its potability (Hasse 1989; Farreny et al. 2011; Snir et al. 2022; Nya et al. 2023). However, the following two key points are generally recognized: (i) carefully collected rainwater is of good drinking quality (Davis 1891; Kim et al. 2016; Han & Nguyen 2018; Bui et al. 2021; Akkerman 2020; Snir et al. 2022) and (ii) collecting rainwater onsite eliminates their possible contamination with pollutants carried by urban runoff (Faragò et al. 2019; Snir et al. 2022). This implies that even in cases where harvested rainwater is polluted, the pollutants are known and affordable treatment options exist (Alsulaili et al. 2020; Huang et al. 2021b; Kearns et al. 2023; Nya et al. 2023). Clearly, RWH should not only be regarded as a simple tool to support household level water demands or as a useful practice to mitigate the lack (Hasse 1989) or depletion (Snir et al. 2022) of conventional water resources. Rather, RWH (i) prevents rainwater from accumulating and reaching the streets in large quantities, (ii) reduces overloading urban drainage systems (e.g., reduced runoff volumes and peak flows), and (iii) potentially avoids flash flooding (Agarwal & Narain 1999; Wahab & Ojolowo 2013; Loudyi & Kantoush 2020; Echendu 2022).

A further advantage of RWH as derived from Figure 1 is artificial infiltration to recharge the aquifers, and thus, locally making springs and wells more productive (Adams et al. 2010; Mupangwa et al. 2012; Woodhouse et al. 2017; Chitata et al. 2021). This knowledge is indigenous in several societies and has been applied without interruption in Sri Lanka for the past 1,000 years, as an excellent example of combining rainfall and runoff harvesting system (Smakhtin & Piyankarage 2003; van Meter et al. 2014; Kirshanth & Sivakumar 2018; Ramabrahmam et al. 2023). King Parakramabahu, the inventor of this civilization in Sri Lanka once said: ‘Let no drop of water flow to the sea unused by man.’ The realization of this vision implied the intensive collection and storage of rainfall and surface runoff in human-made reservoirs (e.g., cisterns, ponds, and tanks) and provided water for domestic uses, fish farming, irrigation, and livestock rearing (Hasse 1989; Han & Nguyen 2018; Han 2023a). The reservoirs are in general located in a cascade-like manner along shallow valley courses. These reservoirs are connected by canals or furrows and are built as a complex system of floodwater harvesting, water storage, and water distribution (Jayatilaka et al. 2003; Smakhtin & Piyankarage 2003). This wisdom is found in some modern rainwater slogans like: ‘all water, for all and by all’ (Han 2023a), ‘catch it when you can’, ‘catch it where it falls’, ‘from drain city to rain city’ (Han 2013b), or ‘never let rainwater flows to the sea’. This view can be translated into the following simple equation: Flash flood = Non-harvested rainwater. Coming back to Equation (1), harvesting rainwater corresponds to reducing the R value (runoff) while increasing the I value (infiltration). As concerning the E value, its modification depends on where harvested water is stored. If rainwater (RW) is stored in an open pond, the E value increases. If RW is stored in a buried cistern, the E value is significantly reduced. The extent to which all these values are modified depends on the volume of harvested rainwater (storage capacity), which is determined by Equation (2).

Summarized, flooding arises from excess rainfall, when rainfall intensity exceeds the infiltration rate of the soil. Excess water is intensified by the abundance of human-made impervious surfaces within the city. In the absence of modern infrastructures, a considerable fraction of falling rain is retained (i) in soil, lakes, and ponds or (ii) by vegetation. In natural or undisturbed ecosystems, the non-retained water typically flows as surface runoff without or with limited destructive potential (Wahab & Ojolowo 2013; Loudyi & Kantoush 2020). The recognition that the removal of these natural water retaining features is the roots cause of urban flooding has led to mitigation tools like low impact development, sustainable drainage systems, and sponge cities (Eckart et al. 2018; Lashford et al. 2019; Dijk et al. 2020). However, these approaches are rooted in intensively draining rainwater away from residential areas (homes) with a better control of issues like: (i) blocked hydraulic structures and (ii) siltation of water channels. In contrast, rainwater management as introduced herein (Kilimanjaro concept) creates a network to enable multiple uses of any drop of rainwater on a home-to-home, farm-to-farm, and hill-to-hill basis (Marwa et al. 2018; Qi et al. 2019) as it has been done in Sri Lanka for centuries (Jayatilaka et al. 2003; Pandey et al. 2003; Smakhtin & Piyankarage 2003; van Meter et al. 2014).

The Kilimanjaro concept roots integrated water resources management (IWRM) on RWH while considering the specific needs of all stakeholders (Ndé-Tchoupé et al. 2019; Pembe-Ali et al. 2021; Nandi & Gonela 2022; Nya et al. 2023). Herein, it is about avoiding flash floods for the whole urban population. The presentation until here has demonstrated the suitability of RWH to reduce the volume of runoff resulting from impervious urban surface. The focus of this study is to showcase the potential of RWH in significantly reducing surface runoff to the extent that the risks of urban flood are practically eliminated. This implies the construction of community cisterns, retention ponds, and tanks which can be emptied in a controlled manner depending on the weather forecast. Even in the absence of any reliable system for weather forecast, intensively harvesting rainwater certainly reduces the intensity of floods. Thus, the concept described herein is immediately applicable using available or adaptable methods for water conservation (Hasse 1989; Dallman et al. 2016; Dallman et al. 2021; Han 2023a).

It is very important to insist on the fact that the concept presented herein will work everywhere despite rapid population growth, limited financial resources, past design flaws (poor city planning, poor waste management), and current design weaknesses (poor operational management, weak institutions). In other words, rainwater storage systems should be designed and installed in cities and slums of all sizes to mitigate the effects of grey infrastructures on the water cycle. The further use of stored water for multiple purposes is obvious, including safe drinking water supply (Han & Mun 2011; Han & Nguyen 2018; Mun & Han 2012; Nandi & Gonela 2022; Kearns et al. 2023). Clearly, the Kilimanjaro concept enables flood mitigation even in poorly planned cities (slums) as well. A similar concept for sanitation in slums (e.g., SOIL = Sustainable Organic Integrated Livelihoods) was recently presented in Haiti (Preneta et al. 2013; Tilmans et al. 2015; Moya et al. 2019a, 2019b; Mallory et al. 2020; Carrard et al. 2021; Musaazi et al. 2023). Considering that SOIL systems solved the sanitation problem in cities and slums, the KC is regarded as the missing puzzle to achieve Goal 6 of UN Sustainable Development Goals (SDGs) during the remaining 7 years to 2030 (Hering et al. 2016; Krounbi et al. 2019; Xiao et al. 2023).

The concept of IWRM as promoted today results from a segmented perspective in which available water is managed in a coordinated manner to satisfy all stakeholders and solve arising issues (Biswas 2004, 2008; Han & Nguyen 2018; Islam et al. 2023). As stated earlier, available water is mainly piped from aquifers or rivers. Rooting IWRM on RWH implies further equitably maximizing economic and social welfare inherent to the conventional approach while intensifying the environmental benefits (Han 2013a, 2013b; Han & Nguyen 2018; Marwa et al. 2018). The intensification of environmental benefits is derived from the following: (i) no or less piping network to carry water from distant rivers, (ii) less soil erosion by torrents from residential and industrial areas, and (iii) more local water infiltration and vegetation development. Moreover, public participation is a prerequisite for this approach as excess water (overflow) from each individual residence is expected in community cisterns, ponds, and tanks (Han & Nguyen 2018; Nandi & Gonela 2022; Nya et al. 2023). Clearly, wherever possible (e.g., rainfall data, soil structure, topography), RWH-based IWRM should secure water for drinking, ecosystem development, irrigated agriculture for food production, industrial production, livestock rearing, and all other uses (Ghisi et al. 2007; Kim et al. 2016; Kirs et al. 2017; Han & Nguyen 2018; Nya et al. 2023). The main question is how to create or protect vital ecosystems while dealing with the spatio-temporal variability of rainwater (Hasse 1989; Han 2007; Kim et al. 2016). Two things are certain: (i) RWH-based IWRM creates popular awareness of water management, and it makes the whole population a community of ‘water managers’, and (ii) RWH-based IWRM uses a local infrastructure for collection and distribution, thus avoiding water wastage due to leakage in distant supply systems (Han & Nguyen 2018; Loen 2023).

Rainwater as the main source of water for whole communities is known only on some islands where groundwater and surface water sources are polluted, limited, or simply not available (Liaw & Chiang 2014; Unsal et al. 2014; Kirs et al. 2017; Loen 2023; Parker et al. 2023). The common situation is that rainwater (e.g., stored in community tanks) is introduced as an alternative to conventional freshwater sources including municipal tap water, commercial bottled water, spring water, stream water, and well water (Brooks 2014; Kirs et al. 2017; Parker et al. 2023). Within this paradigm, pilot-scale studies are used to evaluate the suitability of tank rainwater, and eventually establish the conditions of its use without the risk of water-borne illnesses for the community (Han 2007; Kim et al. 2016; Kirs et al. 2017; Bak et al. 2020). Following this strategy, there is a broad agreement on the view that stakeholder education on (i) RWH system design options, and (ii) the importance of regular maintenance activities will secure the health of populations (Cheo 2016; Kim et al. 2016; Kirs et al. 2017; Han & Nguyen 2018; Kearns et al. 2023). The presentation herein takes advantage of this paradigm and acknowledges that where rainwater is polluted, methods for its treatment are available (Ghisi et al. 2007; Aumeier 2020; Huang et al. 2021b; Kearns et al. 2023). The remaining challenge is to convince people to use rainwater as a priority rather than a supplementary source of water (Morote-Seguido et al. 2020; Pembe-Ali et al. 2021; Nandi & Gonela 2022). It is operationally considered that this effort is financially supported as part of disaster prevention (Section 2).

During the past 30 years or more, RWH has gained popularity in both the developed and the developing countries as a sustainable water management approach (Cheo 2016; Kim et al. 2016; Celik et al. 2017; Chirhakarhula et al. 2018; Lashford et al. 2019). Huge numbers of case studies have analysed the feasibility; water quality; and economic, social, and environmental implications of RWH in several parts of the world. However, despite existing reports on ‘community-based rainwater harvesting’ (Kim et al. 2016; Bashar et al. 2018; Amos et al. 2020; Ross et al. 2022; Puppala et al. 2023), no design concept comparable to the one described herein was found. As a rule, community tanks are tanks constructed to harvest rainwater from public buildings like churches (Brooks 2014), city halls (Ghisi 2010), commercial buildings (Sousa et al. 2018), hospitals (Bak et al. 2020; Lee et al. 2021), mosques (Hurayra & Rahman 2022; Kapli et al. 2023), schools (Kim et al. 2016), and universities (Almeida et al. 2021). This is water from relatively larger catchment areas as compared to small roofs of individual houses (Morel 1988; Laborde & Morel 1991; Brooks 2014; Cheo 2018). Residential RWH as a decentralized water supply system is considered to have the following drawbacks, making it unreliable: ‘water quantities collected during the dry season are insufficient, the quality of water collected could be hazardous, and structures suffer due to a lack of expertise’ (Kim et al. 2016). Retaining the idea of a community-based approach as a core principle, and considering the drawbacks of individual residential RWH systems, this study introduces an innovative approach in which community tanks are filled by overflows from residential tanks. This approach is regarded as an update of the cascade system of Sri Lanka (van Meter et al. 2014; Bhattacharya 2015), adapted to any local rural or urban architecture. The premise is that, once the community-based approach achieves a self-regulatory and transparent operating system, it will lead to predictable and sustainable water quantity and quality, achievable with more regularity and self-determination (Bhattacharya 2015; Kim et al. 2016).

A key feature of this approach is that harvested rainwater becomes a source of income within each community. Water-meters can measure the overflow from each residence and the corresponding monetary value, in case water is sold. Harvested rainwater can be locally used for livelihood and food security activities such as brick-making, livestock watering, or gardening. Alternatively, water from community tanks can be sold to water truckers (water trucking) for supply to communities without water or for industrial uses. On the other hand, households with water needs exceeding their harvesting capacities can buy water in the neighbourhood. Finally, in case water is not available in the neighbourhood, distant clean water can be purchased to fill the domestic tanks, for instance towards the end of the dry season when shortages can be expected (emergency water trucking). Emergency water trucking is well established as a short-term intervention that is used to cover interruptions in water service (Bradford et al. 2018; Sikder et al. 2020; van der Heijden et al. 2022; Lane & Kumpel 2023).

In the context of RWH to mitigate flash flood, rainfall data are needed to assess the maximum volume of water that can be harvested. Population size, number of dwellings to be supplied, and water demand in each community are not the primary goal. The intention is to minimize surface runoff while maximizing the productive uses of harvested rainwater (Han 2007; Lal 2015a, 2015b; Nya et al. 2023; Puppala et al. 2023). In other words, groundwater recharge, irrigated agriculture, and potable water supply are positive side effects of combatting flash flooding by reducing surface runoff (and not the drivers of RWH). Although the presentation is focused on residential areas, agricultural areas should also create structures (e.g., lakes, pits, ponds) to quantitatively store and infiltrate rainwater, for example, on a farm-by-farm or hill-by-hill basis.

The presentation until here has insisted on the evidence that RWH is an ancient practice that has just gained renewed attention in recent years due to its potential to address global water scarcity challenges and promote sustainable water management (Angrill et al. 2012; Amos et al. 2016; Mwamila et al. 2016; Alim et al. 2020; Amos et al. 2020). Figure 1 recalls that global renewable freshwater resources are finite (Lal 2015a, 2015b). These resources are also prone to eutrophication, over-exploitation, and pollution. Thus, water security is intricately and intrinsically linked to water quality and food security, and thus, food security cannot be achieved without achieving the water security and the vice versa (Lal 2015a, 2015b; Puppala et al. 2023). This section presents the concept of universal RWH at residential and community levels, as suggested by the Kilimanjaro concept (Marwa et al. 2018; Qi et al. 2019; Huang et al. 2021a). More so, the approach leads to the increased water self-sufficiency ratio (WSSR), which is the ratio of the amount of self-supplied water to total water use (Han & Kim 2007; Rygaard et al. 2011). The WSSR encourages a reduction in reliance on external water sources, such as rivers and streams, boreholes, water utilities, and vendors. The notion of WSSR suggests that RWH should be universally adopted, in cases rainwater can cover all water needs and in cases rainwater can only cover a fraction of the needs. The precipitation data (e.g., 1,400 mm/a in Bangangté, Cameroon) and the length of the dry season (e.g., 5 months in Bangangté/Cameroon) determine the storage capacity to store enough water for all-year use (Nya et al. 2023).

Conventional sustainable water management involves technologies to (i) increase the green water storage in soil, (ii) purify the grey water, (iii) reduce the export of virtual water, (iv) increase water use efficiency, and (v) desalinize brackish water (Angrill et al. 2012; Liaw & Chiang 2014; Lal 2015b; Campisano et al. 2017). RWH is not explicitly named in this agenda although it does contribute to all five pillars, including avoiding desalination (Pembe-Ali et al. 2021; Nya et al. 2024). As stated already, the Kilimanjaro concept advocates for universal RWH, starting at the residential level.

In addition, at a residential scale, harvesting impact can be evaluated collectively. If all households can embark on harvesting volumes of rainwater equivalent to what is occupied by the respective building infrastructures, that will control surface runoff as well as address dry season water supply challenges. For instance, for a household with a 100 m² metal roof (C = 0.8 in Equation (2)), in a day receiving 100 mm of rainfall, a runoff volume of approximately 8,000 L is produced, which prior to the building construction would have spatially infiltrated into the soil, a fraction of it ponded, some of it evaporated and the rest of it is runoff. If it were only one household, then there is not much to worry about as the impact would be insignificant. However, there are several houses simultaneously discharging similar quantities of water towards areas which do not have proper stormwater management plan, subsequently the discharge of large volumes of surface runoff from the said houses tends to flood and destroy the surrounding environment in that particular time. Receiving water bodies such as River Tegeta, River Msimbazi in Dar es Salaam Tanzania, whose performance is already hampered by the introduction of household waste continuously dumped on it, is expected to not only receive but also contain this excess runoff water. With such rivers overwhelmed and failing to handle the excess rainwater, the outcome is the floods experienced in most parts of our African cities. There is a need to be proactive, change mindset (paradigm shift), and exploit available resources towards sustainable development. For example, the said runoff generated volume of 8,000 L if harvested could serve the daily demand of 80 people at 100 L per person per day. Thus, dry season water shortage problems are actually solvable. ‘Harvest water during the rainy season, safely store it, and use it during the dry season’ should be the motto.

Water supply primarily using RWH requires well-established operation and maintenance (O & M) systems for their sustainability at both household and community levels (Słyś & Stec 2020; MacAllister et al. 2022). This means that each community owns and manages (or has affordable access to) all water management services. This requires that the management, operation, and maintenance of facilities are done by the users themselves, while public health authorities control water quality at all levels. In this model, users cover costs related to the installation, operation, and maintenance and after-life replacement of all components of their RWH systems. Operation and maintenance can be managed by a well-trained team (team of caretakers) according to norms fixed by the administration, for example, each residential tank should be cleaned yearly by the team or under its supervision. This approach could create jobs for some members of the community while making sure that older members are not left alone with maintenance operations. Therefore, there is a need to set affordable tariffs for all O & M operations. It is recalled that subsidies must be made available, at least for community tanks because it is about disaster prevention (Section 2).

A fundamental question to answer while installing the RWHI is: What is the appropriate size of the tank? Usually, calculations are made to satisfy the immediate needs of a stakeholder. For example, a person using harvested rainwater for garden maintenance and necessitating 250 L/day for 100 dry days needs to store 25,000 L or 25 m³ of water per year. For this person, it suffices to build a tank with a capacity of 25 m³ or buy five rainwater barrels at 5 m³. (The capacity of each barrel is 5 m³). In a second example, a family of five members, using each 25 L water per day needs 5 × 25 × 365 = 54,750 L water per year. That is, some 55 m³ of water per year. This question in this work is: Can this family cover his whole year water supply while using harvested rainwater alone? In case there is excess how to use it? In case of deficit, where from to have the needed water volumes?

The calculations in Section 5.1.1 were for 100 mm of rain falling on a roof area of 100 m² and giving a runoff volume of 8,000 L (8 m³). The only assumption was a runoff coefficient of 0.8 (80% of rainfall is effectively collected). This means that 8 m³ of water can potentially be harvested from 100 mm of rainfall in Dar es salaam (Tanzania), Göttingen (Germany), Maroua (Cameroon), Postdam (Germany), and Tlemcen (Algeria). The chosen cities correspond to the affiliations of the authorship of this article. The named cities correspond to four different climatic zones: (i) Mediterranean (Tlemcen), (ii) semi-arid (Maroua), (iii) temperate (Göttingen and Postdam), and (iv) tropical (Dar es salaam). Each climate zone is characterized by its annual average temperature (and precipitation). The average temperature is the key parameter controlling the evaporation, E values in Equation (1) (Table 1). Normally extreme temperatures should be considered because it is about evaporation. For example, the driest days in Maroua correspond to temperatures averaging 40 °C, while 30 °C is scarce in Dar es salaam. This is the main reason why harvested rainwater should be stored in a (buried and) closed tank.

Table 1 shows clearly that, if the harvested water should satisfy the needs of the five-member family (55 m³ per year), a 100 m² roof is sufficient only in Dar es Salaam (67% excess) and Maroua (6% excess). The three other cities have to increase the catchment area (roof area > 100 m²) or find alternative water sources (e.g., piped water, springs, wells). The premise of this work is that harvesting rainwater lowers the competition for alternative freshwater sources while protecting the environment and creating new opportunities (Henriquez 1962; van Buurt 2018; Nya et al. 2023). In this perspective, recycling domestic wastewater is regarded as the first option to compensate for the deficit in Göttingen, Postdam, and Tlemcen.

Conventionally, there are two approaches for rainwater management: centralized and decentralized systems (Sharma et al. 2010; Domènech 2011; Domènech & Saurí 2011; Younos 2011; Vallés-Casas et al. 2016; Słyś & Stec 2020; Richards et al. 2022; Kalbar & Lokhande 2023; Puppala et al. 2023; Soh et al. 2023). In centralized management systems, rainwater from several roofs, green areas, or traffic areas is collected/drained and stored or infiltrated into community reservoirs (e.g., cisterns, tanks) or infiltration plants (e.g., basins, ditches) (Angrill et al. 2012). In decentralized rainwater management, rainwater from individual, small-scale roofs, traffic areas, and green areas is managed or seeped away at the point of origin (e.g., farm, household) (Angrill et al. 2012). The innovative approach suggested herein strives to drain only excess water from small-scale reservoirs and store/infiltrate it at a community scale. In essence, it corresponds to a centralized system, with the subtle but key difference that each point of origin (e.g. farm, residence) is a water production unit for all uses (including drinking water). Clearly, RWH is made ‘everyone's issue’. Making RWH everyone's issue corresponds to several known slogans including: (i) ‘cash it when you can’ (Durodola et al. 2020) and (ii) ‘use rainwater instead of dispose it’ (Gueldenberg 1998).

The Kilimanjaro concept advocates for universal decentralized drinking water treatment at the household level, whether centralized systems are likely to be accessible and reliable. Fortunately, versatile household-based water treatment processes to cope with eventually polluted rainwater are available (Alsulaili et al. 2020; Huang et al. 2021b; Kearns et al. 2023). Such processes (i) depict a minimum dependence on supply of spare parts, (ii) are typically easy to operate and maintain, and (iii) use locally available materials (e.g., biochar, metallic iron) and skills (Banerji & Chaudhari 2017; Aumeier 2020; Huang et al. 2021b; Kearns et al. 2021; Kearns et al. 2023). Several water treatment processes fulfil the aforementioned requirements and have demonstrated their technical feasibility in self-sustaining operation at household level (Aumeier 2020; Tepong-Tsindé 2021; Richards et al. 2022; Kearns et al. 2023). In other words, the Kilimanjaro concept potentially safeguards access to universal safe drinking water (fundamental human right) and protects millions of people from flash flooding worldwide.

‘The tragedy of the commons’ is a relatively old metaphor that has motivated the discussion on the rationale use of natural resources over the past five decades (Hardin 1968; Ostrom 1996, 2000; Araral 2014; Ponte et al. 2021; Cashore & Bernstein 2023). The tragedy of the commons considers that a common pool of not privately owned natural resources is prone to failure, misuse, and ruin. Within the water supply community for low-income settings, the named metaphor is mostly used to sustain the discussion on the willingness to pay for water services (Whittington et al. 2012; Akkerman 2020; Nya et al. 2023).

The tragedy of the commons can be considered for community tanks as discussed herein (Section 5.1.2). However, water flowing into such tanks comes from individual residential tanks (overflows) (Figure 3). The driving force of the KC is that the rational man will prefer to have his own water at home before releasing the excess into the community tanks (Akkerman 2020; Nya et al. 2023). Assuming that this is true for everyone, the remaining challenge is to find the best way to equip individual residences with well-sized and operating RWHIs. Once RWHIs are made available, people can even commercialize water they have harvested or use it for productive activities (Han 2023a, 2023b; Nya et al. 2023).

The other evidence is that rainwater is perhaps the sole natural resource that is automatically wasted when it is not harvested. As discussed in Section 2, RWH has several economic, environmental, and social benefits. This is because it is rather the non-harvesting of rainwater that is linked to damages (e.g., erosion, flood, water contamination via leaching of wastes). In other words, encouraging RWH has no potential harm but only has benefits for human and ecology.

The KC is an attempt to create centralized RWHIs at a neighbourhood level. This means that no outsiders to the neighbourhood have access to the harvested water. However, among community members, rules are needed for the systems to be effective. Such rules shall be elaborated, known, understood, and revised to cope with changing conditions. It is certain that each community can find effective and sustainable management rules for the common locally harvested rainwater. As pointed out by Nya et al. (2023) the state through the municipality should supervise quality control of harvested rainwater and makes sure that appropriate treatment technologies are applied where the water is used as potable water. To ease this task, equipped and accredited laboratory should be made available and citizens train to ask for controlled safe drinking water (Ogisma 2022).

Households without access to piped water or relying on a private water source are considered water-poor. Such households are mainly in areas of high elevation or scattered settlements (Domènech et al. 2012; Heijnen 2013; Biswas & Mandal 2014; Cassivi et al. 2023). In the developed world, there are efforts to support their access to piped water or at least to make sure that residents receive some form of safe drinking water (Bradford et al. 2018; Anderson et al. 2023; Balasooriya et al. 2023; Cassivi et al. 2023). Following the conventional approach, the problem is that connecting to mains water is typically expensive and the costs fall on individuals. An alternative approach has been to let people have their private water supply systems while ensuring the responsibility for the water quality (Domènech 2011; Vallés-Casas et al. 2016; Alsulaili et al. 2020; Kalbar & Lokhande 2023). Although privately ensuring the water quality can be regarded as an additional burden or an injustice, it has been a viable approach in many cases (Murphy et al. 2010; Scalley 2012; Kalbar & Lokhande 2023; Loen 2023). Within the framework of the Kilimanjaro concept, the responsibility for making the water potable falls clearly on individual households, however, under strong public control (Nya et al. 2023). For this reason, municipalities are invited to create or promote the creation of accredited water laboratories (Reijtenbagh 2010; Ogisma 2022; Loen 2023; Nya et al. 2023). In other words, RWH is an optimal solution for water supply everywhere there is rainfall: (i) in the valley and at higher elevations, (ii) in humid and semi-arid regions, and (ii) in rural and urban areas. Depending on the abundance of rain, RWH can be either (i) a stand-alone technology for water supply or (ii) a supplementary amount of water supplied by other sources, which have been considered to be conventional (e.g. rivers, springs, wells). On a very local scale, RWH has two main advantages: (i) wells are more productive (more water available via artificial recharge) and (ii) the demand for well water has decreased (water conservation).

The soundest foundation of human health and prosperity is universal access to a safe drinking water supply (UN SDG 6.1) (Hering et al. 2016). Historical evidence from the Dutch Caribbean (e.g., Bonaire, Curaçao, Saba, St. Eustatius) and from South-East Asia (India, Korea, Sri Lanka) demonstrates that large-scale, self-reliance water management rooted in RWH is applicable and very efficient (Henriquez 1962; Dharmasena 1994; Han 2013a; van Meter et al. 2014; Loen 2023). The affordability of such decentralized systems has been discussed (Musz-Pomorska et al. 2020; Nya et al. 2023). However, it is certain that their design can be regarded as low-tech compared to piping and treating water from distant rivers. The merit of such systems is that they correspond to the paradigm ‘small is beautiful’ (Schumacher 1973) and are a powerful tool to achieve a ‘circular economy’ (Bauwens et al. 2020; Friant et al. 2020; Giampietro & Funtowicz 2020). Large-scale RWH adoption can have significant economic implications for both the supply and demand sides of water resources management (Cook 2004; Heijnen 2013).

• On the supply side, RWH reduces the dependence on conventional sources of water, such as surface water and groundwater, which are often scarce, expensive, and vulnerable to climate change. RWH can reduce the need for costly water infrastructure, such as dams, pipelines, treatment plants, and pumping stations, which require high capital investment, operation and maintenance costs, and energy consumption. RWH also reduces the environmental impacts of water extraction and distribution, such as land degradation, biodiversity loss, water pollution, and greenhouse gas emissions.

• On the demand side, RWH can increase the availability and accessibility of water for various uses, especially in rural and peri-urban areas where water services are often inadequate or unreliable. RWH improves the quality and quantity of water for agricultural production, which can enhance food security, income generation, and livelihood diversification. RWH can also improve the quality and quantity of water for domestic use, which can improve health, hygiene, sanitation, and gender equity. RWH also provides a buffer against droughts and floods, which can reduce the vulnerability and resilience of communities to climate shocks.

Large-scale RWH adoption can generate multiple benefits to local economies in terms of value creation, cost savings, income generation, employment creation, and social welfare (Henriquez 1962; Domènech et al. 2012; Heijnen 2013; Naval 2016). It can also increase the availability and accessibility of water for various uses and provide a buffer against climate shocks. Some of benefits of large-scale RWH are quantifiable and measurable, while others are qualitative and intangible. The present article presented a qualitative conceptual framework for mitigating floods in cities via RWH. The quantities of harvestable rainwater, the demand met via RWH, and the extent of flood mitigation will invariably differ among cities. Therefore, further work is required to pilot-test and to validate the conceptual framework in typical urban settings.

Table 2 summarizes some of the potential benefits to local economies from RWH adoption.

In light of the foregoing, the following recommendations are made:

• Embrace universal rainwater harvesting (RWH) implementation: To address global water scarcity challenges and promote sustainable water management, governments should proactively encourage universal RWH. Financial incentives should be offered to individuals and communities to invest in RWH infrastructure. Mandatory laws can be passed, requiring RWH in new construction sites to make RWH a norm rather than an exception. By fostering collaboration with public and private sectors, RWH initiatives can be further accelerated, making it an integral part of national and regional water management policies.

• Raise awareness through public education campaigns: Effective public education campaigns are crucial to raise awareness about the benefits of RWH. Governments, NGOs, and communities must join forces to educate the public on how RWH can mitigate flash floods, provide an alternative water source during droughts, and contribute to overall water conservation. Integrating RWH education into curricula can instil a culture of water conservation from an early age, ensuring a generation that values water as a precious resource.

• Strengthen public–private partnerships: Public–private partnerships play a pivotal role in driving innovation and making RWH technology more cost-effective. Governments must collaborate with the private sector to advance RWH technology while ensuring policy integration of national and regional water management policies. By establishing favourable provisions and regulations supporting RWH adoption, the private sector can work hand in hand with governments to make RWH initiatives seamless.

• Capacity building and technical advancements: NGOs can play a crucial role in promoting RWH in communities through capacity building and technical approaches. Local communities should be empowered through training programs and workshops, ensuring they have the necessary skills, knowledge, and financial support to successfully implement and sustain RWH initiatives. Investing in research and advancements, such as sensor-based control systems and hybrid water supply management, can enhance the efficiency and reliability of RWH systems, making them even more effective in mitigating flash floods and conserving water.

Adoption of RWH on a global scale would pave the way for sustainable water management and offer a solid remedy for issues like flash floods and water scarcity. Governments, NGOs, and communities can work together to make RWH the norm, ensuring a better, more secure future for water management. Societies can adopt RWH and benefit from its many advantages by providing financial incentives, enacting legislation that is required, and promoting awareness through public education initiatives.

A major economic impact on both the supply and demand sides of water resource management would result from the widespread use of RWH. RWH lessens environmental effects, lowers capital investment in water infrastructure, and lowers operational costs by lowering reliance on conventional water sources. On the demand side, RWH expands the amount of water that is available for domestic and agricultural usage, improving food security, income production, and general wellbeing. RWH adoption becomes both a necessity for the environment and a competitive advantage for local economies as a result of the value creation, cost savings, revenue production, and employment opportunities it generates. Therefore, RWH can continue to develop as a potent weapon to address water scarcity and enhance water management practices with a dedication to research and technological developments. Societies may build sustainable water systems that protect against climate shocks and enable a more prosperous future by establishing water management on RWH and incorporating it into wider water management frameworks.

The authors are grateful to Coretta Tchouandem Nzali (PAUWES, Tlemcen/Algeria) for proofreading the revised manuscript. The authors would like to acknowledge anonymous reviewers of Water Supply for their helpful comments.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.