Navigating the definition of urban flooding: A conceptual and systematic review of the literature

Flooding is increasingly recognized as a recurring water-related hazard around the world, as sea levels rise, riverbanks overflow, and global climatic change induces greater rainfall frequency and intensity (IPCC 2021). International reports have been providing evidence of recurring and increasing flood events in many regions of the world, examples being the AR6 Synthesis Report: Climate Change of the Intergovernmental Panel on Climate Change (IPCC), the Global Awareness Report (GAR) and the Sendai Framework for Disaster Risk Reduction (DRR) from the United Nations of Disaster Risk Reduction (UNDRR 2019; IPCC 2023). Examples of flood events that have occurred in many localities are provided in Supplementary material, Table A1 in Appendix A. The flooding events analyzed were systematically selected from national disaster databases, media reports, and other reliable sources, using a combination of criteria including flood type (pluvial, fluvial, and coastal), impact severity as measured by factors such as fatalities, people affected, and economic damages, as well as a minimum water level threshold of 5 m above normal to capture the most extreme incidents. While large impactful events such as Hurricane Harvey in Houston and the Ganges River Floods in India have shown the significant risk of major flood events, flooding is pervasive (Moore et al. 2016; Yao et al. 2016; Ramos et al. 2017) and poses significant risk for many communities ( Supplementary material, Table A1). At the same time, the different flood events also demonstrate how flooding can cause significant risk to communities, even without a significant death toll.

Multiple spatial distributions are seen, with small, non-disaster declared events, such as the flooding in Seattle (Washington) and Ellicott City (Maryland) in the United States (US), or large flooding cases like the events in Germany and in Brazil in 2021. However, while the risk of flooding is perceived, the definition of what constitutes flooding, particularly in an urban context, and how it is separate from more traditional flooding terms, is still considered as a challenge in the literature. This statement is endorsed by previous publications, related to challenges in risk management, suggesting the need to rethink manners of communicating flooding specially to impacted residents (Rollason et al. 2018), and establishing appropriate terminologies for risk understanding (Monte et al. 2021).

‘Urban flooding’ (UF) is tackled differently by physical and social scientists. Traditional water management scholars define UF as ‘an inundation caused by rainfall such that the runoff caused exceeds the conveyance capacity of the urban drainage system, and exceedance flow is generated on the urban surface’ (Butler et al. 2018), focusing on the hazard (i.e., rainfall) and consequences (i.e., the exceedance flow). For social scientists and planners, UF is seen as ‘a phenomenon with the potential to harm’ (Cutter et al. 2003, 2008), where the characteristics of flooding and the society (or area) will determine the level of risk. Due to the complexity and multifaceted characteristics of disasters, for many, risk definitions are considered as imprecise leading to confusion especially when translated from other languages (Monte et al. 2021). Recent discussions among scholars for defining risk and their constituents (hazard, vulnerability, and exposure) can be seen (see more in de Ruiter & van Loon 2022 and Alves et al. 2023).

When considering UF, definitions can lack clarity as to what constitutes the disaster, especially in terms of how we plan for, mitigate, and manage flood events in the different built and rural environments (Cho & Chang 2017; Hazarika et al. 2018; Rollason et al. 2018). This ambiguity in definition is further complicated by recent advancements in urban flood modeling, which have emphasized the complexity of the urban environment and its impact on flooding dynamics. For instance, Xu et al. (2022) proposed a framework for urban flood modeling in data-poor regions, highlighting the challenges of limited drainage data and complex urban surfaces. Their approach, which uses satellite soil moisture products and measured flood data to calibrate drainage capacity, underscores the need for UF definitions to accommodate varying levels of data availability and to incorporate the unique characteristics of urban landscapes.

Using different definitions can impact the way UF is managed and mitigated. For example, if using the Sendai Framework, the risk will be considered as the intersection of hazard, vulnerability, and exposure drivers (UNDRR 2019). In this case, assessing flood risk will only be accomplished when the three disaster constituents are fully accounted for with the support of multiple tools (Shah et al. 2020). However, if the study focuses more on the physical sciences, flooding can be described as the relation of direct and indirect impacts, not necessarily focusing on the existing multifaceted vulnerability of the area. For some authors, the lack of understanding about vulnerability and its root-causes can be a barrier for developing more enhanced adaptation solutions for flood mitigation (Ashley et al. 2020; Schipper et al. 2021), and can lead to incrementing the vulnerability and compounding the effects for specific populations (Ruiter et al. 2020). In other words, the main challenges for risk management rely on properly understanding risk drivers, but also the way they are put into practice and addressed by policies.

Recent studies have further emphasized the importance of considering urban infrastructure in flood risk assessment. Çirağ & Firat (2023) utilized GIS and InfoWorks ICM to analyze and calibrate the flood performance of stormwater drainage systems in Malatya, Turkey. Their study evaluated the systems' capacity under various rainfall scenarios, revealing that the current infrastructure may be inadequate to fully convey stormwater, potentially risking life and property. This research emphasizes the critical role of stormwater drainage systems in urban flood management and suggests that UF definitions should consider infrastructure capacity and performance under different rainfall intensities and durations.

Other studies simply do not explicitly define what is UF, or do not provide deep knowledge of the geographical area it takes place in, which can also negatively impact flood management and mitigation. However, similar to flooding, the definition of ‘urban areas’ is also varied across the literature and management institutions. For example, Meerow et al. (2016) considers ‘urban systems’ as a conglomeration of ecological, social, and technical components in a space, while Norris et al. (2008) classifies ‘communities’ as an entity with geographical boundaries with an intersection between built environment and the natural, social, and economic conditions (Norris et al. 2008; Meerow et al. 2016). For the US Census Bureau, ‘urban areas’ are mainly classified as (i) urbanized areas (Us) with a population of 50,000 or more and, (ii) urban places outside of UAs with at least 2,500 inhabitants. However, in other countries, this classification can be slightly different, which can change the way flooding risk mitigation is practised by policy makers of that region.

The complexity of UF and the need for a more comprehensive approach to its definition and assessment is further highlighted by recent research on urban flood resilience. Xu et al. (2023) developed a method to quantify UF resilience based on the ‘4R’ theory (robustness, redundancy, rapidity, and resourcefulness), coupling urban rainfall and flooding models. Their study in Kunming City, China, revealed a negative correlation between areas prone to waterlogging and flood resilience levels. They also identified significant local spatial clustering in flood resilience indices. This approach emphasizes the importance of incorporating resilience concepts into UF definitions and assessments, considering not only the physical aspects of flooding but also the city's capacity to cope with and recover from flood events.

In this context, as ‘urban areas,’ ‘flooding,’ and thus ‘urban flooding’ definitions are intrinsically interrelated, the main research questions of this paper are: ‘how is UF conceptualized in literature?’ (RQ1) and ‘what are the main implications of UF definitions to future research?’ (RQ2). These questions address the critical challenge of fragmented understanding across disciplines and aim to provide a more integrated perspective on UF. Our study uniquely contributes to the field by synthesizing diverse conceptualizations and proposing a comprehensive definition that bridges the physical and social sciences. To navigate through UF definitions, the main objective of this paper is to conduct a systematic literature review on the relevant studies on flooding, although not restricted to a specific field of science or to a flooding type. This paper focuses on incorporating current practices and views on UF and the role of the built environment through an analysis of the literature. For this, structured searches were conducted in the Web of Science database with articles from 2003 to 2023, from which initially 663 articles were identified, and 67 articles were selected for this review. The decision to focus the literature search on articles published between 2003 and 2023 was driven by the aim of capturing the most current and relevant research on UF studies and definitions over the past two decades. This time period was selected to provide a comprehensive yet manageable sample that would allow for identifying and analysing emerging trends, methodological approaches, and conceptual frameworks within this relatively recent body of academic literature. The main objective of this paper is not to provide an exhaustive list of UF definitions used worldwide, rather it aims to develop a systematic analysis of how studies have been conceptualizing and understanding the controlling of urban flood risk drivers, as this is the main objective of risk management (Koks et al. 2015; Cho & Chang 2017; Kreibich et al. 2022).

The paper is organized as follows. The detailed methodology is described in ‘Methods’, followed by the analysis of the literature in ‘Results’. ‘Discussion’ provides a discussion of UF definitions and their impacts on future research (RQ1 and RQ2). The conclusions drawn are presented in the final section.

The article selection process for this review involved a structured search of the Web of Science database for publications from January 2003 to December 2023. For the earlier time period of 2003 to 2019, the results were initially sorted by the number of citations, prioritizing articles with a higher number of citations (10 or more) to focus on research that was more referenced in other studies. However, recognizing that more recently published articles are likely to have fewer citations, the authors then expanded the selection to include all relevant publications from 2020 to 2023, irrespective of citation count. This approach ensured the review captured the most up-to-date research on UF definitions and studies, balancing the emphasis on impactful prior work with the inclusion of emerging perspectives. Phase 1 yielded 663 papers that examined flood events (Figure 1). As seen in Table 1, queries using ‘Urban flooding’ generally returned the most results, likely because other terms (such as ‘Interior’ or ‘Inland flooding’) are not expressively used by the academic literature, and could be better described using the specific cause of the flooding (such as ‘fluvial’, ‘alluvial’, or ‘pluvial’ flooding).

Phase 2 enabled the review of the content in each article. For this, five questions were applied to each article to ensure we captured how UF was being defined in the literature. Because the main objective of this paper is to assess how the articles define ‘urban flooding’, the five questions were formulated as follows: (Q1) Does the article discuss flooding in human settlements? (Q2) Does it include a definition of UF? (Q3) Are rainfall events examined? (Q4) Is there an infrastructure element in the discussion? and (Q5) Is the definition presented original? The process of reviewing the initial 663 papers involved a systematic approach: first, we screened titles and abstracts for relevance to UF concepts. Then, we applied our five-question criteria to the full texts of the potentially relevant articles. The articles that met all the criteria were included in the final analysis (n = 67) as seen in Figure 1. This rigorous process ensured a comprehensive yet focused review of the literature.

The final selection of articles includes implicit and explicit ‘urban flooding’ definitions, as long as they had a robust discussion of UF elements. We believe that combining the two types of definitions analyses enable the discussion about the key elements for defining UF in studies. The final list of articles was carefully analyzed by the research team multiple times in separate rounds.

A full description of the questions and the number of articles in each phase of the search are presented in Figure 1. The results show the percentage of papers and number of articles (i.e., called ‘n’) within each category analyzed, including countries, continents, journals, methodology, and definitions of flooding (implicit or explicit). The final list of articles, and main categories analyzed, are presented in Supplementary material, Appendix A.

This section aims to provide an overall discussion of the reviewed literature organized for the analysis of (i) UF literature over time, (ii) uncovered terminologies, (iii) journals targeted, (iv) case studies focus, (v) methods applied, (vi) explicit and implicit definitions used, and (vii) ‘urban’ definition. While presenting those findings, our aim is to provide a general discussion of the studies present in our review (n = 67). Section four focuses on discussing those studies in relation to the broad academic literature (RQ1 and RQ2).

In general, the articles deal with flooding cases in multiple countries such as US, Canada, India, China, Ghana, Spain, Norway, Croatia, Italy, Sweden, Portugal, Estonia, Sweden, Pakistan, Nepal, Germany, Japan, and the United Kingdom (UK). The articles focus on different flooding ‘scales’ and the majority are developed for ‘catchments’ (Moore et al. 2016; Yao et al. 2016; Ramos et al. 2017), with some examples applied in ‘cities’ (Coulthard & Frostick 2010; Kaspersen et al. 2017), and ‘regions’ (Giovannettone et al. 2018; Sörensen & Emilsson 2019). Article types are divided into literature reviews (Parkinson 2003; Sörensen et al. 2016; Sandink & Robinson 2022), modeling tools (Kaspersen et al. 2017; Wu et al. 2017), including Storm Water Management Model (SWMM) (US EPA) (Meierdiercks et al. 2010; Moore et al. 2016), Geographic Information Systems – GIS tools (Maksimović et al. 2009; Smith et al. 2013; Truu et al. 2021), remote sensing (Suriya & Mudgal 2012; Ramachandran et al. 2023), machine learning (Qin et al. 2024), as well as statistical analyses (Kim & Park 2016), and public engagement (Owusu-Ansah 2016; Rözer et al. 2016; Safaei-Moghadam et al. 2023). When looking at the research focus from which each article was published, it was noted that journals tended to be more engineering-science oriented.

Our analysis reveals that 35 of the 67 papers base their discussion on the causes of UF with an implicit definition, rather than providing an explicit definition. This finding highlights the need for more precise terminologies in UF research. In some cases, UF is left undefined, especially when the study is focused on measuring a specific hydrological response. This is exemplified in Ogden et al. (2011), where the effects of urbanization such as impervious area, the density of stormwater drainage, and the addition of subsurface storm drains are examined regarding the peak flood flows of Baltimore (US). Another instance of an implicit UF definition can be observed in the study by Siljeg et al. (2024), where it states that impervious surfaces like roads, parking lots, rooftops, and sidewalks restrict the natural infiltration of rainwater into the ground, leading to increased stormwater runoff and altered hydrological dynamics in urban areas. These studies do not explicitly define UF but consider that there is a strong influence of imperviousness and drainage systems on flooding events. Conversely, an explicit definition is provided in the work of Ahmad et al. (2023), who define UF as flooding that occurs in urban areas because of the increase in impervious surface areas, which prevent water from infiltrating the soil, leading to waterlogging and flooding. They specifically mention that UF is caused by the expansion of urban impervious surface area, resulting in rainwater flowing off instead of soaking into the ground.

In other cases, a catch-all definition is used. For example, the US Federal Emergency Management Agency (FEMA) eschews a specific definition of UF, instead describing flooding as, ‘A general and temporary condition of partial or complete inundation of normally dry land areas from: (i) The overflow of inland or tidal waters; (ii) The unusual and rapid accumulation or runoff of surface waters from any source; (iii) Mudslides (i.e., mudflows) which are proximately caused by flooding and are akin to a river of liquid and flowing mud on the surfaces of normally dry land areas, as when earth is carried by a current of water and deposited along the path of the current’. More recently, a National Academy report on UF deployed an all-encompassing definition, ‘Urban flooding is caused when the inflow of stormwater in urban areas exceeds the capacity of drainage systems to infiltrate stormwater into the soil or to carry it away. The inflow of stormwater results from (a) heavy rainfall, which can collect on the landscape (pluvial flooding) or cause rivers and streams to overflow their banks and inundate surrounding areas; or (b) storm surge or high tides, which push water onto coastal cities. Floodwater inundation and movement are influenced by (a) land development, which disturbs natural drainage patterns and creates hardened surfaces that inhibit infiltration of stormwater; and (b) stormwater systems that are undersized for current needs and increase exposure to drainage hazards. In older cities, sewer systems carrying both stormwater and wastewater can become surcharged during storms, causing sewer backups in homes – an often chronic and unseen form of urban flooding’ (National Academies of Sciences Engineering & Medicine 2019).

In the surveyed literature, very few authors define what is ‘urban’ compared to what is not ‘urban’. Broadly, authors have included suburban spaces and urban transition areas in the discussion of flooding. For example, for Qiao et al. (2018), ‘cities around the world are suffering from pluvial flooding because of increased urbanization and climate change, and there is a general increase in impervious surfaces, and frequency of cloudbursts in urban areas increase stormwater runoff and lead to high peak flow events’. However, the main problem seems to be that old and small-diameter combined sewers cannot cope with huge amounts of stormwater in a short time, leading to frequent pluvial flooding events. Other examples are Maksimović et al. (2009) and Kaspersen et al. (2017). Maksimović et al. (2009) base their analysis of urban pluvial flooding on the representation of the ‘urban fabric’, represented as the combination of land-use, topography, and drainage inlets. Kaspersen et al. (2017) links the high proportion of ‘imperviousness’ with urbanization and the catchment response in a flooding event. ‘Urban’ is therefore defined as a combination of impervious surfaces and human built water conveyance systems.

In this section, our objective is to systematically discuss the manner in which UF was defined by authors, highlighting their ‘key’ factors and interconnections. Instead of defining UF itself, authors implicitly present UF as a series of possible flood risks that have their own – although possibly overlapping – elements. The literature and definitions were analyzed according to ‘three general components’: the source of the water, the amount of impervious surface in the area in question, and what human or natural (possibly human augmented) drainage systems exist to capture and convey the water. These components were selected based on their consistent prominence and centrality within the established theoretical and empirical literature on urban hydrology and flood risk management. They are widely recognized as primary drivers and mediating factors that shape the occurrence, severity, and impacts of flooding in built-up environments. Therefore, they provide a robust and comprehensive framework for analysing urban flood definitions and synthesizing conceptual approaches across the body of research.

In many definitions, there is a tendency to categorize UF by the source of the water (Tingsanchali 2012; Sörensen & Emilsson 2019). ‘Water source’ refers to the different types of flooding caused by rainfall, riverbank overflow, or sea rise (i.e., pluvial, coastal, fluvial), since it will delineate how the extreme event can be discussed or simulated (if desired), and which solutions/actions can be implemented to reduce flooding risk. Because most articles are from engineering journals, this framing can be explained by the need for characterizing the nature and scale of flooding, and possible interventions and countermeasures discussed, such as levees, stormwater systems, and detention ponds.

In the surveyed literature, the most common water source discussed when it comes to UF is pluvial events (Douglas et al. 2008; Smith & Rodriguez 2017). This focus on pluvial flooding is particularly relevant given the increasing impacts of climate change and land-use changes on urban hydrology. For example, Smith & Rodriguez (2017) discuss how high-resolution radar rainfall data can be applied for mapping flash floods in New York City (US). The results show the importance of considering multiple urban ‘scales,’ because the same city can be vulnerable to multiple flooding types, especially pluvial and coastal floods. This vulnerability is often exacerbated by urban development and population growth, which convert natural areas into urban landscapes, altering local hydrological cycles (Pathirana et al. 2014). There is also a focus on describing urban pluvial flooding and stormwater management in China, and others contextualize pluvial floods in private households of Germany (Rözer et al. 2016; Jiang et al. 2018). While some papers note the danger of river and coastal flooding (Sörensen & Emilsson 2019), the focus of most papers is on water occurring outside of ‘traditional’ flood-prone areas. Most papers examine the role of human impact on flood occurrences; while settlement in flood plains raises the risk of flooding in urban areas, the question most papers seek to answer or investigate is the role of the built environment. This built environment, characterized by increased impervious surfaces, not only reduces natural infiltration capacity but also contributes to the urban heat island effect, potentially increasing rainfall intensity and frequency in urban areas (Miller & Hutchins 2017). We explain this with the fact that an examination of pluvial events and the distribution of the built environment is considered as a less complicated approach since it is possible to measure and describe what happens to the water once it hits the ground. The context changes when considering coastal and riverine flooding. These events result from a confluence of factors, making it difficult to discern to what degree human systems affected the outcome. However, the interplay between climate change, urban development, and hydrological processes significantly intensifies UF in all its forms, both in terms of frequency and severity. This complex interaction suggests that future UF definitions should explicitly consider these dynamic factors, moving beyond traditional categorizations to encompass the evolving nature of flood risks in urban environments.

Climate change adds another layer of complexity to UF in the literature. First, part of the studies note that climate change could increase the amount of flow during flooding events, ‘An increase in the intensity and frequency of extreme precipitation events may therefore result in an increase in sewer overflows and flooding of urban areas… [climate change] is expected to modify the intensity/frequency of intense rainfalls’ (Mailhot & Duchesne 2010). Secondly, we see a link between climate change and land-use land change, since the expansion of impervious surface areas due to urbanization reduces the ability of rainwater to infiltrate the soil, leading to increased surface runoff and UF (Ahmad et al. 2023). Furthermore, impervious surfaces in urban areas restrict natural infiltration and increase stormwater runoff (Šiljeg et al. 2023). While it is agreed that UF due to events exceeding system capacities is expected to occur more frequently and to be more harmful in the future, we also highlight that cities are facing the dual problem of increasing urbanization as well as an increasing number of wet-weather events. Both elements place greater demands and pressures on current drainage systems, adding to the likelihood of UF. This best explains why climate change is a part of the UF discourse.

Urbanization is consistently alluded to as one of the causes or amplifying factors of urban flood events, primarily by way of describing the increase in impervious surfaces resulting from human settlements (Sandink 2016; Jiang et al. 2018). This shorthand is likely the result of earlier studies that used the change in impervious surfaces as a way to measure urban development (Kaspersen et al. 2017; Ramos et al. 2017). More complicated definitions explicitly describe impervious surfaces and stormwater infrastructure, or lack thereof, as the human elements of flooding, see an example in Wu et al. (2017).

As a result, papers tend to focus more on the role of impervious surfaces, since natural basins and urbanization can be easily compared under that methodology. The role of infiltration and surface runoff in amplifying the flow rate of stormwater is common in definitions (Suriya & Mudgal 2012; Ramos et al. 2017). This is partially because there is a commonly accepted method of measuring the impact of surface runoff, which is the flow rate of water (Yao et al. 2016; Sörensen & Emilsson 2019). Saraswat et al. (2016) summarize this relationship well by stating that ‘in the urban environment, due to the impervious surfaces that cover the natural environment, the hydrological processes of surface water runoff become more unnatural, causing damage to infrastructure and contamination of water by pollutants’.

Most papers note that reductions in imperviousness have their greatest effect on moderate to low rainfall, with extreme rainfall events being too much for natural infiltration processes (Suriya & Mudgal 2012). What we can take from this is that UF is more frequent than natural flooding (i.e., flooding before urbanization), because it creates flood events out of low or moderate hydrological events. The impacts of urbanization on the water cycle are significant. A high proportion of impervious surfaces, such as roads, buildings, and parking lots, is primarily a result of the built environment and is directly linked to urbanization and the hydrologic response of a catchment in a flooding event (Kaspersen et al. 2017). This is particularly relevant for pluvial flooding, where the amount of water absorption is drastically reduced due to urbanization.

Another aspect to consider is where impermeable areas are located. UAs will reduce water infiltration and evaporation, altering the water cycle and increasing water runoff. Therefore, having more impervious surfaces can be one of the reasons for the increase in flood-prone areas in different communities. In various other articles on environmental science, Meierdiercks et al. (2010) note that impervious cover upstream can be more impactful on stormwater flows. Beyond the increase of floodwaters height and volume, as Saraswat et al. (2016) note, runoff can also be a carrier of pollutants, which can be a source of public health impacts (Saraswat et al. 2016). A large percentage of impervious cover around wastewater treatment and other hazardous land uses further complicates the public health risks created by UF. When dealing with flooding reduction, it is important to also point out that placing impervious surfaces in the catchment will alter the behavior and distribution of flooding. Therefore, land-use changes that reduce imperviousness can create a positive impact on the flooding downstream.

The other built environment element of UF is urban drainage and stormwater infrastructure. Most articles describe flooding occurring when drainage capacity is ‘exceeded’ (Coulthard & Frostick 2010; Meierdiercks et al. 2010). While the impact is often described, the reason why such infrastructure is inadequate is often unaddressed. Some explanations include the effect of climate change on the intensity of traditional water events (Moore et al. 2016) as well as the age and maintenance of the infrastructure (Maksimović et al. 2009). Other studies also focus on the discussion of how drainage and impervious surfaces are linked, with a higher density of drainage systems being able to offset increases in impervious surfaces. When drainage systems are discussed, the most common methods applied are hydraulic and GIS tools with the example of Drainage Density tools (Meierdiercks et al. 2010; Ogden et al. 2011), but also statistical modeling approaches (Mailhot & Duchesne 2010).

In recent years, low-impact development (LID), nature-based solution (NBS), and green infrastructure (GI) strategies have emerged as a crucial consideration in UF management and, consequently, in how we define and conceptualize UF. LID approaches, such as green roofs, permeable pavements, and bioretention systems, aim to mimic natural hydrological processes in urban environments (Eckart et al. 2017). These strategies can significantly impact UF formation and processes by reducing runoff, increasing infiltration, and improving water quality. For instance, Zhu et al. (2019) demonstrated that LID practices could reduce peak runoff by up to 40% in urban watersheds. Alves et al. (2022) modeled the benefits to be acquired with NBS strategies in Campina Grande, Brazil, and their results suggest benefits beyond flooding reduction, also having social and economic benefits for the community. Such findings suggest that comprehensive UF definitions should incorporate the potential benefits of ‘sustainable’ strategies, as they represent a shift from traditional ‘grey’ infrastructure to more sustainable and resilient urban water management approaches.

Some papers say stormwater management is dependent on infiltration, infrastructure, and other ‘upstream’ factors. Usually, those factors are mostly related to hydrology and water cycle components. For Willems et al. (2012), UF analyses require to focus on small urban catchment scales and short duration precipitation extremes. However, other studies also state that the provision and maintenance of stormwater infrastructure is also the product of political relationships between communities and state actors (Ranganathan 2015). Political decisions are well addressed in studies such as Coulthard & Frostick (2010) when the authors describe the context of the City of Hull, northeast of England. Hull is largely below sea level and relies on a pumped drainage system with no natural ways of drainage; however, the constant changes in the drainage system increased the flooding cases in the city. Papers also focus on how flooding is a result of poor maintenance, or inequitable investments in stormwater management, putting socially vulnerable communities at a higher risk of flooding from failures in drainage systems (Yu & Coulthard 2015).

The most prominently discussed issue with drainage systems is sewer flooding in combined sanitary systems, where pipes become overwhelmed or blocked and distribute water into homes, basements, or areas in which the water cannot drain (Wheater & Evans 2009; Willems et al. 2012; Rözer et al. 2016; Sörensen & Emilsson 2019). Marchi et al. (2010) and Wu et al. (2017) consider that urban inundations are subject to multiple factors including elevation, discharge, land-use, soil types, rainfall, and stormwater and sewage drainage networks (Marchi et al. 2010; Wu et al. 2017). A survey of UF conducted by the National Academies of Sciences, Engineering, and Medicine noted that this flooding type is often hidden, as the flooding is highly localized to a few sites. Another perspective seen is that UF is not necessarily an engineering problem, but rather an issue with the distribution of resources (Sandink 2016).

In this section, we describe a multifaceted and comprehensive definition of UF. In the surveyed literature, UF is described with elements that represent a combination of origin (i.e., water source) and built environment (i.e., urbanization, imperviousness, and stormwater networks). However, we also recognize the smaller focus of the literature linking flooding with the local aspects of communities. We highlight that flooding events will depend on physical elements of the spatial context, but its frequency and intensity are also dependent on the social context lived by communities. Defining UF therefore asks for an interdisciplinary perspective that considers not only a reflection about human-built infrastructure, but also the understanding about settlements where communities live and about the communities themselves.

The Urban Water Transect shows UF as a continuum process that reflects the changes between ground infiltration, stormwater infrastructure, and built-up increase (from the T1 ‘Natural Zone’, with no infrastructure, to the SD ‘Special District’ zone, with mainly infrastructure and no nature) as depicted in Figure 3. Next, we discuss three main challenges that should be considered for incorporating interdisciplinary perspective, exemplified by our comprehensive definition and the Urban Water Transect, for UF definitions in future research. These challenges are ongoing, and because of that they are not only discussed according to the articles analyzed in this review but are also put into context with other literature in the field.

The first challenge is that the literature is more focused on describing impacts of the extreme events (outcome) rather than targeting flooding causes or origins. This barrier is seen in great part of the surveyed literature, in which UF is described as the damage caused by and for stormwater infrastructure and physical environments, with minor reflection about the conditions causing the flooding. While critical infrastructure and physical components are directly linked to the ‘causations’, they are not necessarily the root-causes behind flooding events. For example, flooding can take place in different configurations of ‘urban areas’ (i.e., with or without stormwater infrastructure), but it can be generated combined with other reasons that form their vulnerability and exposure (i.e., rainfall characteristics, the location of residents in flood-prone areas, or changes in the land-use cover) (Shah et al. 2020; Schipper et al. 2021). Another example is seen when human settlements greatly change the environment leading to UF. This alteration can generate more unpredictable flood risks as natural processes are replaced by human processes. At the same time, increasing settlements in urban areas can coincide with increasing social and economic inequalities (Pallathadka et al. 2022), which adds more complexity to the analysis especially because changes are dynamic. While infrastructure plays a key role as a mechanism that can improve runoff infiltration with engineered and natural techniques, we also highlight that it can diffuse environmental injustice and further social vulnerability both every day and during disaster events. In the studies, probable causes of UF are implicitly considered, but there is a bigger focus on describing impacts than flooding root-causes, from both the physical and social perspectives. Flooding root-causes should be seen as an intersection of factors that generate flooding risk (i.e., hazard, vulnerability, and exposure), beyond only considering imperviousness or presence of drainage systems.

However, UF studies should not consider that all factors are always causing flooding. Considering multiple ‘urban’ zones can assist the understanding of ‘what’ factors generate flooding, but not all factors are present in each zone of the Urban Water Transect (Figure 3). For example, research can look to urban sprawl and its effects on flooding risk, but at the same time should also question ‘is this flooding caused by a result of sprawl in relation to a region that was previously natural (T1) or it was already considered urban (T6)?’ or ‘how was the process of urbanization in this area and how it contributes to the distribution of flood-prone areas?’ or even ‘is this flooding impacting more a community because of the inequities present in the area?’ Another applicability is that while infrastructure like dikes and dams do exist in some urban areas, what qualifies as ‘urban’ is not always clear in studies, as shown in the Results section above. Answering these questions is complex because the landscape is changing at different temporal perspectives, in different zones, which creates more unpredictable water flows, but the risk of floods could already be present before urban growth. In this sense, we consider that sectorizing geographical areas into ‘different zones’ can be a tool for reducing simplistic and generalist approaches, since UF will be discussed according to the ‘specific conditions’ of the geographical area (or transect zones) as depicted in Figure 3.

Current studies provide a small reflection about the cascading impacts of flooding with other hydrological hazards in place. Allied with vulnerabilities existing in each zone, urban flood risk can pose real health risks to communities as a source of pollution (Parkinson 2003) and disease transmission (Douglas et al. 2008), creating or worsening additional environmental hazards (Sandink 2016). Figure 3 exemplifies this concept by showing the many flooding pathways and how floods are impacted by and can trigger other hazards. Urban floods are simultaneously affected by physical aspects, such as the distribution of current infrastructure or types of network systems such as combined or separate sewer systems or the rainfall event itself. However, while the literature acknowledges that flooding creates impacts on health and infrastructure, they can also occur on similar spatial–temporal scales of other disasters. This refers to ‘compound’ and ‘cascading’ disasters defined as ‘situations when the same area is simultaneously or consecutively vulnerable to flooding and other disasters for a period of time’ (i.e., see more in de Ruiter et al. (2020) and Alves et al. 2023).

The compounded and cascading disasters will amplify existing vulnerabilities with direct or indirect effects on communities at risk, which will be asked to cope with more than one disaster at a time (Leonard et al. 2014; Ruiter et al. 2020; Ward et al. 2020). This appears to be even more important to urban planning and management studies, since there are examples of flooding cases allied with the occurrence of landslides, urban heat islands, hurricanes, droughts (Coseo & Larsen 2019; Alves et al. 2020; Fereshtehnejad et al. 2021), or even the COVID-19 pandemic (Di Baldassarre et al. 2021). UF should therefore be considered as an extreme event not only driven by urban infrastructure that intensifies and spreads damage as a result of impervious surfaces and poor drainage and stormwater management systems, but it is also impacted by other conditions, and disasters, in place as depicted in Figure 3 and in the comprehensive definition. Flooding impacts will have a direct relationship with vulnerability and exposure of communities, creating damage to buildings and personal property and increased public health risks from the transmission of pollutants.

Finally, the literature is usually focused on either physical or social aspects of ‘flooding’, but not both. Findings show that even though the surveyed literature in this review is published in interdisciplinary journals, most studies are still from the ‘engineering’ perspective (Supplementary material, Appendix A). In general, studies focus more on analysing the physical implications of UF. This result corroborates that of the previous studies of Vanelli et al. (2022) and Fasihi et al. (2021) that discuss the fragmentation in the field, with the prevalence of ‘monodisciplinary’ studies focusing on the physical elements (Fasihi et al. 2021; Vanelli et al. 2022). In this review, findings also point that although great effort is seen in both engineering and social sciences, studies are developed mainly with a lack of participatory tools for understanding the social context. Participatory approaches are only developed in a few studies (Suriya & Mudgal 2012; Owusu-Ansah 2016; Smith & Rodriguez 2017; Safaei-Moghadam et al. 2023).

Combining social and physical perspectives is complex, and this may be the reason why it is still limited in the literature. For that, understanding the relationship between the social and physical systems, in different spatial–temporal scales, are key for their integration (Figure 3). Social–physical integration is seen in some examples of the literature. Findings from Rosenheim et al. (2019) point out that while damage to residential structures and associated infrastructure are critical determinants of impacts, the characteristics of the social systems also play critical roles in shaping the nature of the impacts and subsequent response and recovery patterns but are rarely considered (Rosenheim et al. 2019). In a similar perspective, Brody et al. (2014) highlight how assessing the effects of different urban areas’ configurations is critical for flooding reduction since it affects the structural and social conditions of residents, which can also be classified as an environmental justice topic. For Park et al. (2024), environmental justice (EJ), vulnerability, and segregation are linked in stormwater events, especially because there are disparities in the distribution of sewer pipelines and GI among communities with differing social vulnerabilities. Growing inequalities play a key role in this context; however, the hazard and disaster literature still tend to be disconnected from EJ literature (i.e., see more in Rufat et al. 2015; Howell & Elliott 2018; Hendricks & Van Zandt 2021).

UF is a product of human planning or the lack thereof, such as the placement of impervious surfaces, investment in and maintenance of drainage infrastructure, as well as the willingness of residents and authorities to adopt adaptive and coping strategies. In this sense, addressing UF requires a multifaceted and interdisciplinary approach that considers not only physical infrastructure but also social vulnerabilities, community engagement, and equitable access to resources for disaster response and recovery.

Across the literature, it is possible to see that a clear and more inclusive definition of UF is needed. Unlike more traditional forms of flooding, UF has additional public health and socioeconomic dimensions that make smaller scale flood events even more impactful. In this article, we aimed to understand how flooding studies are being developed in multiple fields of science (RQ1), covering different UF epistemologies and terminologies, and their impacts on future research (RQ2).

The results show most articles do not explicitly define UF, considering mainly the relationships among the three main elements: origin of flooding, urbanization, and stormwater and drainage infrastructure. While some studies address the social conditions of flood-affected areas, most articles prioritize examining rainfall and hydraulic parameters for modeling UF, especially because they are from an engineering perspective. The main emphasis of the studies in the literature review is on water origin (pluvial, coastal, and fluvial flooding) and on the built environment aspects (stormwater infrastructure and imperviousness).

In this sense, we discuss the articles by (i) delineating an interdisciplinary and comprehensive UF definition, and (ii) exemplifying flooding pathways and interconnections with the Urban Water Transect. We discuss the main takeaways and challenges to improve the manner in which UF is defined in the literature. Further studies need to clarify and investigate what the origins and causations of flooding are, without targeting only the consequences of flooding. Likewise, while urbanization (i.e., imperviousness) and drainage structure are key for determining flooding impacts, we also consider that the role of other natural, built, and social conditions should not be neglected in flooding analysis. Similarly, studies should also consider the other extreme events that may be occurring in the same urban area. Identifying the factors behind risk, multiple hazards, as well as vulnerability and exposure offers an opportunity to deeply understand the specificities of each place to identify flooding origins and urban dynamics. Finally, we point to the need that both physical and social aspects of communities should be incorporated into UF studies and analysis.

In summary, this paper shows that considering UF definitions with only water source and built environment and overlooking the social aspects and interconnectedness with other systems will impact the way these events are managed and mitigated, including roles and responsibilities, jurisdictional boundaries, mitigation and coping practices, response measures, vulnerability, disaster risk assessments, and beyond.

The authors of this paper are thankful for the support given by researchers from the Stormwater Infrastructure Resilience and Justice (SIRJ) Lab at the School of Architecture, Planning and Preservation at the University of Maryland. We also thank Chad Smith for developing the awesome graphics of this paper.

This work was supported by the Gulf Research Program of the National Academies of Sciences, Engineering, and Medicine under the Grant Agreement number 2000009661. The content is solely the responsibility of the authors and does not necessarily represent the official views of the Gulf Research Program or the National Academies of Sciences, Engineering, and Medicine.

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

The authors declare there is no conflict.