Assessing and screening flood exposure risks for emergency response

Assessing and screening flood exposure risks for emergency response | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Assessing and screening flood exposure risks for emergency response Daniel Adshead, Martin Jonsson, Nikolaus Mezger, Georgia Destouni, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8600985/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 4 You are reading this latest preprint version Abstract Flood occurrence and intensity changes may increasingly threaten the accessibility of emergency medical services. To investigate this threat, we develop a data-driven framework for assessing flood effects on time-critical ambulance routes, using suspected cardiac arrests Sweden as a test case. We integrate two key datasets: over 117,000 ambulance dispatch records, and seven years of high-resolution modelled hydrological risk data. First, we establish a strategic health access network through algorithmic routing of the empirical dispatch locations and intersect these routes with historical flood risk to derive their long-term exposure profile. We find that the most vulnerable emergency corridors face up to 17% annualised flood risk frequency, or approximately 62 flooded days per year. Second, we perform a risk-centric performance screening to identify specific operational degradation, isolating ambulance routes where real-time flood risk coincided with measurable response delays that exceeded local performance baselines. This screening localises critical operational delay hotspots in emergency services and quantifies delays for flood-exposed routes, with an average delay rate of 82 seconds/kilometre over the municipal median. These delayed ambulance route segments occurred most frequently within urban centres in west and central Sweden (e.g. the cities of Gothenburg and Karlstad) as well as in smaller, geographically diverse and rural areas, such as northern Sweden. Overall, the approach developed and tested here shifts risk assessment from static hazard mapping to dynamic service discontinuity, offering a general tool for prioritising operational planning measures and infrastructure investment through a public health lens to enhance emergency medical response capacity and resilience. Flood risk management critical infrastructure emergency medical services (EMS) climate-resilient health systems GIS analysis climate change and health Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 1. Introduction Hydro-climatic changes, e.g. in occurrence frequency and intensity of floods (Vieira Passos et al. 2024), are increasingly challenging access to critical services essential for achieving social and economic development (Adshead et al. 2024 ; Vieira Passos 2025a). The interconnectedness of modern infrastructure, spanning energy, transportation, water, and communication networks, as well as social infrastructure like healthcare facilities and schools, amplifies this challenge (Verschuur et al. 2024 ). Disruptions to key access infrastructure such as road and transportation networks have the potential to create cascading failures across sectors that jeopardize the delivery of vital services (Vieira Passos et al. 2025b ). Flooding, in particular, not only damages physical assets but may also slow operations and critically hamper the accessibility of emergency services, creating a situation where the societal systems intended to manage crises are themselves compromised. The potential for high-frequency, localized flash flooding to affect response times, even at moderate intensities, poses a significant, understudied operational risk to these services. Proactive adaptation measures, spanning local to national scales, are needed to reduce social risks, safeguard service continuity, and enhance resilience against climate-driven disruptions. Addressing the interplay between physical infrastructure and societal vulnerability requires a systemic approach, especially as climate adaptation policy shifts discourse towards 'riskification' (Englund et al. 2023). This means reframing climate change impacts from an immediate, exceptional 'threat' into a calculable, manageable 'risk' that can be systematically managed through established governmental tools, such as formal risk assessments, policy recommendations from expert agencies, and integration into existing planning structures, rather than necessitating extraordinary measures or entirely new governance frameworks. These approaches require localized flood risk assessment techniques to consider the potential climate threats to infrastructure assets, and to manage them accordingly. Such techniques are increasingly enabled by advances in climate and hydrological data availability, synthesis, and modelling, which have transformed our ability to predict and adapt to the impacts of extreme weather events (Panahi et al. 2023 ), including to multi-hazard exposure and at localized level with high resolution (Gudiyangada Nachappa et al. 2020). New approaches integrate high-resolution climate scenario projections with hydrological simulations to demonstrate how probabilistic models (Ziya et al. 2023) can predict flood risks and their cascading effects on infrastructure. These models, increasingly leveraging satellite data and machine learning, are capable of monitoring flood-prone regions and assessing their potential impacts with unprecedented spatial and temporal detail (Lu et al. 2025 ). Here, we address the impacts of flooding on healthcare, with specific focus on emergency service provision. Much of the research on hydrometeorological hazards to healthcare has addressed cyclone-related flood impacts to hospitals and care facilities (Abebe et al. 2025 ; Yazdani et al. 2022 ); however, the infrastructure supporting access to these facilities is equally critical (Alam et al. 2024 ). Simulation studies have been key to explore the impacts of flooding on healthcare access, offering valuable insights into potential vulnerabilities and mitigation strategies. These approaches generally focus on modelling the spatial coverage of urban emergency response services during various flood magnitudes to ensure compliance with mandatory response times set by national or municipal health authorities (Coles et al. 2017 ; Green et al. 2017 ; Yu et al. 2020 ), using models such as hydrodynamic urban waterlogging (Shi et al. 2024 ). Analysis of hazard impacts has also advanced beyond simple inundation mapping by assigning flood depths to road segments, allowing the network to be modelled as partially degraded based on flood depth and reduced vehicular speed (Tsang et al. 2020). This helps identify sites with restricted access and quantify vulnerable populations that may no longer meet mandatory response times. Scenario-based evaluations have been applied more widely to assess flood impacts on urban public services, simulating inundation across various pluvial flooding scenarios to calculate emergency response capacity and accessibility (Zhang et al. 2022a ; Zhang et al. 2022b ), with specific attention to the unique challenges of reaching vulnerable groups. Assessments have quantified the increased exposure and delay time for essential services to facilities such as kindergartens (Shi et al. 2023 ), nursing homes (Johnson & Yu 2020 ) and elderly populations (Shan et al. 2023 ), and economically-important areas like tourist sites (Shi et al. 2022 ) under severe pluvial and coastal flooding scenarios. Furthermore, these scenario approaches have been used to evaluate the cascading impacts of sea level rise and coastal flooding, simulating service areas for multiple responders under varying traffic and flood conditions (Yin et al. 2017 ). Other studies utilize Bayesian networks to examine the intricate evolution of emergency responses during specific heavy rain events, allowing decision-makers to understand key nodes and to take emergency response measures in time (Xie et al. 2023 ). Despite these advances, much of the existing research relies on simulated flood scenarios and theoretical models of traffic disruption, with limited validation against historical data. This reliance on simulation, while necessary, shows and leaves a critical gap: access to empirical, high-resolution data to validate risk models and inform localised resilience efforts. Bridging this gap requires connecting advanced environmental modelling with local operational data observed over a large scale. Empirical analysis of emergency response data has been applied in a handful of studies to understand the specific, time-sensitive demands of health emergencies like out-of-hospital cardiac arrests (OHCA) (Pell et al. 2001 ), and the factors associated with response time variation (Harries et al. 2025; Mathiesen et al. 2018 ). Prompt ambulance response and reduced travel distance is crucial for improving survival and neurological outcomes (Nicholl et al. 2007 ), while differences in survival rates between fast and slower responses underscores that minutes are critical (Holmén et al. 2020). Timely interventions not only boost immediate survival rates but also improve the chances of discharge with good neurological outcomes (Bürger 2018). Flood-induced disruptions to emergency medical services (EMS) exacerbate these risks by prolonging response times. Sweden is a country with extensive geographic and climatic diversity, making it relevant and interesting as a test case for understanding how flooding affects EMS and how critical that may be for developing resilience strategies. In this study, we develop and test a data-driven framework for assessing flood vulnerability on time-critical emergency routes, with Sweden as a test case. This case enhances our understanding of theoretical route network risks by facilitating integration of novel, high-resolution datasets, including empirical dispatch records from a centralised national emergency service and established, high resolution hydrological risk data. These datasets allow us to construct a robust, validated baseline for risk modelling and adaptation planning, with a primary goal to identify, prioritise, and spatially localise the most vulnerable segments of the emergency response network for future infrastructure and adaptation investments. 2. Methods 2.1 Data and analysis framework This study develops a two-part analytical framework designed to deliver a comprehensive assessment of flood vulnerability and performance risk for ambulance response routes and applies and tests this for the case of Sweden (Fig. 1 ). The analysis is structured to first establish a long-term strategic vulnerability profile and then provide a performance screening by examining the relationship between real-time flood conditions and response times. This analysis is empirical in its foundation but relies on a systematic approximation of real-world exposure and travel time. Specifically, for the case of Sweden, we utilize three primary datasets: Empirical emergency response data: The actual date- and time-stamped start (dispatch location) and end (patient location) points of 117,875 ambulance dispatches from 2021–2025 were used, providing a true measure of operational demand and observed response times. This cohort of dispatches containing both start and end location was filtered from a larger dataset of over 255,000 dispatches provided by the Emergency Medical Dispatch Centre (EDMC) and focuses exclusively on suspected cardiac arrests (unconscious adult without normal breathing) calls. The data covered 15 of Sweden’s 21 regions. Modelled flood risk data: We used the Daily Flood Index (DFI) as a high-resolution, empirically-derived proxy for the daily presence of localized flood conditions based on historical rainfall and soil characteristics (Deo et al. 2015 ), covering the period 2015–2021 and modelled for the whole of Sweden (Vieira Passos et al. 2024 ). This DFI dataset provides a measure of flood risk frequency on a 100-metre grid derived from historical rainfall data but does not quantify the precise flood impact (e.g. millimetres of inundation) on any given day. This index is based on effective and accumulated precipitation and therefore accounts for various hydrological factors affecting the flood risk, such as evapotranspiration and soil saturation at local scales (Muratoglu et al. 2023 ; Brocca et al. 2016 ). A DFI value greater than zero was used to indicate flood risk conditions. Algorithmic route data: The actual GPS track taken by the ambulance is not logged in the EMDC data. Instead, we used the shortest-path algorithm to calculate a standardized, reproducible baseline route for each dispatch, sourced from OpenStreetMap. Our analysis, therefore, compares observed, real-world travel times both under the daily flood risk condition (DFI > 0) and when no flood risk is assessed (DFI ≤ 0). This allows us to identify when a high-flood risk day is associated with a longer-than-expected travel time, signalling a system performance degradation that approximates the empirical operational impact of flooding. 2.2. Creating a historical flood risk profile The first phase of the analysis focused on historical flood risk assessment. We defined a ‘health access network’ based on over 117,000 usable empirical routes from 2021–2025. For each call, the route was computed as the shortest path by length, from the point of the ambulance dispatch to the call location, using the road network, providing a standardized, reproducible baseline against which risk can be consistently measured. Leveraging this large sample size of real-world dispatches provides a robust estimation of mobility corridors essential for critical emergency response access. While based on algorithmic efficiency rather than observed GPS tracks, this extensive dataset identifies the high-frequency pathways that must remain functional to meet the bulk of public health demand across the region. To intersect these linear routes with the gridded flood data, each computed route was densified by placing a sample point every 100 metres along the route. Drawing on previous methods of flood exposure calculation for roads (Sohn 2006 ; Papilloud et al. 2020 ; Pregnolato et al. 2017 ) and other critical infrastructure (Pant et al. 2018 ), these sampled points were then used to query the entire 7-year (2015–2021) flood risk (DFI) history, where intersection of any point with a grid cell having a DFI value greater than zero on the same day classifies the route as ‘exposed’. A route's long-term vulnerability was determined by calculating annualized flood risk days per year, which quantifies the average number of high-risk days per year for each critical route. This phase yields the strategic vulnerability map, identifying segments of the current network most frequently exposed to historical flood risk. Because the DFI is designed to measure the potential for daily flooding, flood depth is not included as a factor in this study. 2.3. Performance screening and critical hotspot localisation The second phase of the analysis focused on observed performance and risk filtering. This analysis was limited to the single year of 2021 where our usable route data and the full flood risk data overlap. This 2021 cohort enables a direct audit of response time performance under documented flood conditions. To allow for fair comparison across different route lengths, the primary response delay metric was normalised by distance, creating the ambulance response delay rate (seconds per kilometre), which measures route efficiency rather than simple duration. For each 2021 call, its real-time exposure was determined by querying DFIs for that specific route on the same day. Routes with any sample point intersecting with a DFI value greater than zero were classified as "Exposed"; the others were classified as "Non-Exposed". The classification of 'Exposed' indicates the presence of a high-frequency flash flood risk condition on that specific day, without specifying the level of inundation or flood intensity. To address risk localisation, a spatial hotspot map was generated to serve as a risk filtering tool. This map visually isolates only those routes that, on the same day in 2021, (1) were flagged as "Exposed" and (2) exhibited a delay rate greater (i.e. a normalised response time slower) than the municipal median, providing a sensitive tool for identifying anomalous delays that exceeded the baseline for that specific geographic area. This approach is designed as a risk-screening tool for infrastructure planning, not an analysis of extreme outliers. Filtering only the most extreme delays would mask the higher frequency of routes where performance deterioration begins to occur. By capturing every ambulance dispatch case where performance fell into the slower half of the distribution during flooding, we provide a more comprehensive, actionable set of locations for proactive mitigation, such as drainage improvements or alternative route planning, ensuring that minor but frequent performance drops are also addressed. We do not aim to prove that flooding always slows ambulance routes down, but rather to identify the specific locations where the combination of flood exposure and poorer than average response performance actually occurred, creating a critical vulnerability. 3. Results This section presents the spatial and operational impacts of flood risk on ambulance response routes in Sweden, first characterising long-term historical exposure across the national response network and then identifying route segments where flood exposure coincides with observed response delays. 3.1. Historical flood risk profile of emergency response routes Establishing the long-term vulnerability of the operational ambulance access network to flood exposure reveals significant spatial heterogeneity across Sweden. Figure 2(a) illustrates the overall mean DFI encountered along all sampled routes, including the majority of days with no significant flood risk, highlighting lighter-coloured areas where flood events are frequent enough to influence the long-term average exposure across the entire route network. These lighter-coloured routes, indicating more frequent flood exposure, are concentrated in several distinct regions. Notably, routes in northern Sweden (particularly Norrbotten), parts of the west coast (Västra Götaland), and the southern region of Skåne, exhibit persistently higher mean DFI values. In contrast, Fig. 2(b) focuses exclusively on the intensity of flood risk to the emergency response network, reporting the mean DFI only for the days with a flood exposure (DFI > 0). This metric isolates the severity of the hazard when it occurs, with the spatial pattern observed here highlighting areas facing the most intense flood events. The lighter-coloured routes in this panel delineate regions where the average flood hazard intensity is substantially higher when it is present, even if it is less frequent over time. This is particularly pronounced in northern municipalities, certain inland municipalities in western and southern Sweden (e.g. Västra Götaland, Skåne, and Kalmar), as well as large parts of central Sweden (Jönköping, Östergötand). This comparison reveals a crucial distinction: densely populated areas, like the Swedish capital Stockholm, face a high frequency of low-to-moderate exposure (as suggested by the overall mean DFI), while other areas face less frequent but significantly more intense flood exposure, i.e. higher mean DFI when a flood event does occur. To better contextualise the spatial flood vulnerability, Fig. 3 provides a focused view on mean DFI across five regions identified as having high vulnerability based on flood exposure intensity. Summarizing emergency response route exposure by municipality further quantifies these spatial trends. Overall exposure, measured by the proportion of flooded days, emerges as most severe in the northern Swedish municipalities of Luleå (48.75 days/year), Övertorneå (37.53 days/year), and Haparanda (36.63 days/year). Conversely, while Linköping has a moderate flood exposure frequency (11.68 days/year), its mean exposure intensity when flooded (with a mean DFI of 0.461, indicating severe hydrological stress) is the highest recorded, indicating that recorded routes within this municipality face the most severe flood magnitudes on affected days. Other municipalities with high intensity include Pajala (mean DFI of 0.434), Övertorneå (0.431), and Oskarshamn (0.449). This granular level of data is essential for emergency planning, distinguishing between areas needing high-frequency, low-level mitigation (e.g. urban drainage maintenance) versus areas requiring high-consequence preparedness (e.g. robust alternative routing or specialized rescue). Across the full set of over 117,000 individual ambulance routes, the historical exposure analysis reveals that the top exposed routes face up to 17% exposure to flood risk, corresponding to over 62 annualized flooded days per year. The majority of these top-exposed routes also share a high mean DFI intensity when flooded (approximately 0.40, denoting significant inundation levels), although others reach as high intensity as 0.62 (representing more extreme events). Clustering of nearly identical exposure metrics across consecutive entries suggests that these routes often overlap or are used for multiple ambulance calls, representing single, critical vulnerability points that service multiple points of origin and destination. Addressing these few, highly exposed emergency corridors would yield the greatest return on investment in risk reduction. 3.2. Screening risk-affected routes for ambulance response delays The second stage shifts the focus from long-term structural vulnerability to real-time operational performance screening under documented flood risk conditions. Due to limited sample sizes for flood-exposed calls in many municipalities for the one analysed year (often fewer than 10), this study adopts a targeted, risk-screening methodology that focuses on quantifying the magnitude of performance degradation where flood exposure coincides with demonstrably poorer than average service delivery, rather than formal statistical significance testing. Figure 4 presents a hotspot map identifying these critical routes. This filters the 2021 dataset to only include emergency calls where 1) any section of the route had flooding (DFI > 0) on the date of the dispatch and 2) where the recorded ambulance response delay rate (seconds per kilometre) exceeded a baseline derived from non-exposed calls. Portions of these exposed routes are indicated with red markers based on their intersection with flood grid cells and are weighted by frequency and magnitude of use during delayed responses. We use a local delay rate baseline, defined as the municipal median response time (s/km) of non-exposed routes, to identify slow response times relative to overall system performance. Using the median delay rate in each municipality allows detection of anomalies relative to local operating conditions, as the national median can mask significant variations in inherent emergency response standards between regions (e.g. rural vs. urban). Using the local non-exposed baseline equitably reveals whether the flood hazard introduces a significant local impairment to service delivery and provides a more actionable assessment of flood vulnerability tailored for regional emergency managers. The average response delay rate (seconds per kilometre, s/km) for all flood-exposed routes found to be slower than the local municipal median was 81.7 s/km. The hotspots of anomalously high delay rates do not simply show areas of high flood risk, but isolate segments where flood exposure coincides with demonstrable delay in ambulance service delivery. This shifts our assessment from identifying where flooding can happen (hazard-centric) to identifying where flooding may cause the most harm to emergency service delivery (risk-centric). Figure 5 provides a quantitative summary of the most critically affected Swedish municipalities based on the local screening criterion, highlighting areas where flood risk-exposed ambulance responses exceeded the non-exposed median for the municipality. The top 20 municipalities by average delay for routes where flood exposure coincided with poor performance are noted. This highlights certain areas where, despite low sample sizes, observed delays were severe when they did occur. Figure 6 illustrates the distribution of these high-risk areas across Sweden, clearly showing the disproportionate impact of delays in geographically diverse municipalities, such as those in northern Sweden (e.g. the municipalities of Boden, Gällivare, Kalix), the central Värmland and Dalarna regions (e.g. Leksand, Ludvika, Filipstad), alongside significant clusters in regional urban centres in central and western Sweden (e.g. Karlstad and Gothenburg). The overall findings indicate that while severe flood intensity is often observed in sparsely populated northern areas, the operational risk (ambulance response delays) is likely a complex interaction between flood characteristics and the underlying volume and criticality of the ambulance routes, with results confirming the existence of operational delay hotspots in screened routes and segments. These identified hotspot routes represent a critical set for targeted resilience interventions, as they were simultaneously exposed to flooding and incurred a measurable operational penalty. In urban areas, higher volumes of traffic and difficulty of finding quick alternative detours mean that even a minor disruption can translate into a substantial, measurable operational failure (Li et al. 2021 ). 4. Discussion The strategic value of this study lies in shifting governmental risk assessment from static hazard mapping to dynamic, operationally focused vulnerability analysis. Traditional climate risk assessments often focus on infrastructure exposure (e.g. which roads are in a flood plain) or damage (e.g. what it costs to repair a bridge). Here, however, we shift the focus to service discontinuity, translating environmental risk into a measurable public health consequence such as minutes lost in an emergency response. By intersecting historical flood data with a network of real-life ambulance routes, this type of analysis compels planners to view flood risk not just as an environmental or civil engineering problem, but as a direct constraint on emergency service capacity and a threat to health equity. This reframing elevates the discussion from routine infrastructure maintenance to core national healthcare priority. Furthermore, the methodology acts as a necessary inter-sectoral bridge. It translates complex hydrological science (e.g. via the Daily Flood Index) into actionable, network-based metrics that are instantly recognizable to transport planners, and finally into public health outcomes (emergency response times) that compel action from healthcare providers and policymakers. Here we used DFI as the hydrological basis for the emergency network metrics because this has been tested comparatively with other types of indicators and found to be the best-performing one for representing actual flood occurrence across Sweden over the period 1922–2021, i.e. for the geography and time of the present test case study. Alternative flood occurrence indicators (Vieira Passos et al. 2024 ) include the Standardized Precipitation Index, Standardized Precipitation and Evapotranspiration Index, and the Standardized Streamflow Index; in other applications, any of these alternative indices (also including the DFI), or other possible combinations of observed and modelled flood data, can be used for the translation from hydrology to emergency network metrics, depending on which is found or considered to best represent actual flood conditions in each specific application. In general, this translation is essential for overcoming the traditional institutional silos that separate climate, transport, and health agencies, and for introducing flexible and collaborative response mechanisms, which have been identified as lacking in previous research on health emergency preparedness to floods and other hazards (Soldà et al. 2025). Without such a unified metric, climate adaptation plans risk being isolated within environmental departments, failing to influence the budget allocations of agencies responsible for road maintenance or ambulance deployment. Robust risk and exposure maps provide a data authority that previous theoretical models lacked, establishing a transparent and empirical case for resource prioritization tied directly to time-critical public services. Isolating high-demand ambulance corridors that are both essential to health service delivery and vulnerable to flood risk enables a fundamental shift toward risk-informed investment that proactively builds system resilience, rather than merely reacting to the next extreme weather event. 4.1. Policy relevance for resilient medical emergency response The findings from this analysis point directly to two distinct categories of measures among a broader scope of adaptation options (Noble et al. 2014 ; Azhar et al. 2025 ): built environment/engineering solutions, and operational planning measures. The built solutions fall primarily within the domain of transport agencies and municipal road authorities (Mourad et al. 2022 ). For road segments identified with high historical flood exposure, engineering and design reviews are warranted. This includes traditional infrastructure measures, such as improving side-drainage channels and replacing undersized culverts to manage increased runoff from pluvial (flash) floods, which the flood risk model is well-suited to identify. The data also supports the implementation of nature-based solutions, which involves incorporating green infrastructure, such as establishing permeable pavements or roadside bioswales to vulnerable segments to slow and filter runoff. In cases of extreme vulnerability, such as routes with a high annual flood risk and no viable detours, more significant interventions like physically raising key road segments or hardening bridge approaches may be justified. The exposure data provides the evidence to target these costly and protective built options where they will have the most significant impact on maintaining health services. Operational and planning measures fall to emergency service providers and the administrative bodies responsible for healthcare delivery to residents. The maps of high-risk routes should be used to formally pre-designate and vet flood-resilient alternate routes (Oddo & Bolten 2019 ). These alternates must be integrated into dispatch systems so they can be immediately suggested, rather than relying solely on a driver's ad-hoc local knowledge during a crisis. In the longer term, flood risk data itself could be integrated as a real-time predictive layer. Future flood scenarios forecasting high risk in a given area could automatically flag a standard route as "at-risk," prompting dispatchers to use the pre-vetted alternate path and thereby reducing the risk of a critical delay. This shift towards data-driven, pre-emptive operational planning ensures continuity of service even when physical infrastructure is compromised. In Sweden, the test case of this study, national climate adaptation policy centres on the reframing of climate change as a manageable risk that must be integrated into existing tools and strategies. This study provides exactly such a tool, directly linking observed historical risk to actionable agency mandates. For the Swedish Transport Administration (Trafikverket), which manages national roads, the historical exposure of priority routes can be used as an evidence base to prioritise maintenance and adaptation budgets as suggested in its adaptation strategy update (Trafikverket 2018). Road segments with notable flood risk delays can be flagged for targeted hydraulic assessments or drainage improvements, focusing resources where the empirical risk to health services is highest. For municipalities (Kommuner) and county boards (Länsstyrelser), which are tasked with local climate adaptation planning, this analysis provides the high-resolution, localized data they need for their risk and vulnerability assessments, moving beyond generic county-level climate reports. The study also provides a concrete example of systemic risk analysis for the Swedish Civil Contingencies Agency (MSB), the current research agenda of which emphasizes research and development needs for understanding the effects of climate risks on critical response infrastructure (MSB 2023a). Notably, MSB has piloted detailed flash flood mapping methods, including projections of more extreme, higher return period floods, in Karlstad and Uddevalla – two urban areas where our risk-centric screening of observed ambulance delays detected operational degradation during more frequent, localized flood risk events (MSB 2023b). For the Swedish Regions (Regioner) and emergency medical dispatch centres (EMDCs), the findings provide a strategic tool for long-term operational planning and for integrating flood risk into national emergency planning protocols. A high-risk route that serves as the only access to a populous area, for example, may highlight a need for new dispatch stations in relation to the vulnerability, or at least a formal plan for pre-positioning resources during high-risk weather forecasts. 4.2. Future directions and global applications Limitations of this study include the temporal scope of the ambulance esponse-time analysis (a single year), which limits the statistical power to detect rare events. A second limitation is the reliance on modelled "shortest path" routing: while this is a standard and necessary methodological assumption for creating a reproducible baseline, these routes do not represent actual GPS tracks, which could deviate based on driver choice, geographical obstacles, real-time traffic, or dispatcher-led re-routing. Despite these limitations, the framework itself is highly generalizable, stemming from the fact that the core components of the problem (climate-driven hazards impacting physical road networks and the societal need for rapid emergency response) are universal. While the specific hazards, data availability, and governance structures may differ, the fundamental link between infrastructure integrity and time-critical health outcomes is a shared global challenge. The hazard layer we have used here (Daily Flood Index) could be replaced with other relevant extreme event models and indicators, such as wildfire risk perimeters, landslide susceptibility maps, or coastal inundation scenarios. This hazard layer method would be particularly valuable in regions with low infrastructure redundancy, such as areas with single-access valley roads or coastal communities, where such an analysis could clearly identify critical hotspot points of failure. In data-scarce environments, the empirical routes could be proxied using mobile phone trace data, with hazard layers derived from publicly available satellite imagery, demonstrating the framework's broad applicability for climate adaptation planning worldwide. Ultimately, this analysis offers a powerful test case study for how to synthesise hydrological science, network analysis, and public health metrics. It provides the evidence required to shift from reactive extreme weather response toward proactive, spatially-informed adaptation planning to secure emergency health access in the face of escalating threats posed by hydro-climatic changes. Declarations Acknowledgments This project was supported by Stockholm Trio for Sustainable Actions (STSA). The authors also acknowledge Fredrik Byrsell and Alessandro Berti for their efforts in securing data. Ethics approval and consent to participate The authors have obtained ethics approval from relevant authorities in order to use the anonymised emergency response data. No human participants were involved in this study. Consent for publication All authors agree with the content and consent to its publication. Competing interests The authors declare there are no conflicts of interest for this manuscript. Author contributions Daniel Adshead: Conceptualization, Methodology, Data curation, Formal analysis, Visualization, Writing – Original Draft, Funding acquisition, Project administration Martin Jonsson: Data curation, Writing – Review & Editing Nikolaus Mezger: Data curation, Writing – Review & Editing Georgia Destouni: Methodology, Writing – Review & Editing Petter Ljungman: Conceptualization, Methodology, Writing – Review & Editing Funding The authors acknowledge funding from the Stockholm Trio for Sustainable Actions (STSA). Availability of data and materials The raw data contain confidential emergency response data that are not publicly available. Access may be granted to qualified researchers upon reasonable request, subject to approval by the relevant ethics committee and the data owner(s), and the signing of a data-sharing agreement. Derived and anonymised data used in the analysis are available upon request from the corresponding author. References Abebe, Y.A., Pregnolato, M. & Jonkman, S.N. (2025) Flood impacts on healthcare facilities and disaster preparedness – A systematic review. International Journal of Disaster Risk Reduction 119, 105340. https://doi.org/10.1016/j.ijdrr.2025.105340 Adshead, D., Paszkowski, A., Gall, S.S. et al. (2024) Climate threats to coastal infrastructure and sustainable development outcomes. Nat. Clim. Chang. 14, 344–352. https://doi.org/10.1038/s41558-024-01950-2 Alam, M.S. et al. (2024) Large-scale modeling of hurricane flooding and disrupted infrastructure impacts on accessibility to critical facilities. Journal of Transport Geography 116, 103852. https://doi.org/10.1016/j.jtrangeo.2024.103852 Azhar, M., Kane, B., Vahedifard, F. et al. (2025) Comprehensive portfolio of adaptation measures to safeguard against evolving flood risks in a changing climate. Commun Earth Environ 6, 824. https://doi.org/10.1038/s43247-025-02779-z Brocca, L. et al. (2016) Rainfall estimation by inverting SMOS soil moisture estimates: A comparison of different methods over Australia. JGR Atmospheres 121, 20: 12,062-12,079. https://doi.org/10.1002/2016JD025382 Bürger A, Wnent J, Bohn A, Jantzen T, Brenner S, Lefering R, Seewald S, Gräsner JT, Fischer M. (2018) The Effect of Ambulance Response Time on Survival Following Out-of-Hospital Cardiac Arrest. Dtsch Arztebl Int. 2018 Aug 20;115(33-34):541-548. doi: 10.3238/arztebl.2018.0541. PMID: 30189973; PMCID: PMC6156551. Coles, D., Yu, D., Wilby, R.L., Green, D. & Herring, Z. (2017) Beyond ‘flood hotspots’: Modelling emergency service accessibility during flooding in York, UK. Journal of Hydrology 546: 419-436. https://doi.org/10.1016/j.jhydrol.2016.12.013 Deo, R.C., Byun, HR., Adamowski, J.F. et al. (2015) A Real-time Flood Monitoring Index Based on Daily Effective Precipitation and its Application to Brisbane and Lockyer Valley Flood Events. Water Resour Manage 29, 4075–4093. https://doi.org/10.1007/s11269-015-1046-3 Englund, M. & Barquet, K. (2023) Threatification, riskification, or normal politics? A review of Swedish climate adaptation policy 2005–2022. Climate Risk Management 40, 100492. https://doi.org/10.1016/j.crm.2023.100492 Green, D., Yu, D., Pattison, I., Wilby, R., Bosher, L., Patel, R., Thompson, P., Trowell, K., Draycon, J., Halse, M., Yang, L., & Ryley, T. (2017) City-scale accessibility of emergency responders operating during flood events, Nat. Hazards Earth Syst. Sci., 17, 1–16. https://doi.org/10.5194/nhess-17-1-2017. Gudiyangada Nachappa, T., & Meena, S. R. (2020) A novel per pixel and object-based ensemble approach for flood susceptibility mapping. Geomatics, Natural Hazards and Risk, 11(1), 2147–2175. https://doi.org/10.1080/19475705.2020.1833990 Harries, M., Ushakova, A. (2025) Modelling emergency response times for Out-of-Hospital Cardiac Arrest (OHCA) patients in rural areas of the North of England using routinely collected data. BMC Emerg Med 25, 8. https://doi.org/10.1186/s12873-025-01170-7 Holmén, J. et al. (2020) Shortening Ambulance Response Time Increases Survival in Out‐of‐Hospital Cardiac Arrest. Journal of the American Heart Association 9, 21. https://doi.org/10.1161/JAHA.120.017048 Johnson, S. & Yu, D. (2020) From flooding to finance: NHS ambulance-assisted evacuations of care home residents in Norfolk and Suffolk, UK. Journal of Flood Risk Management 13, 1, e12592. https://doi.org/10.1111/jfr3.12592 Li, M. et al. (2021) Measuring emergency medical service (EMS) accessibility with the effect of city dynamics in a 100-year pluvial flood scenario. Cities 117, 103314. https://doi.org/10.1016/j.cities.2021.103314 Lu, Y., Huang, Y. & Liu, X. (2025) A multi-source data-driven framework for probabilistic flood risk assessment using cascade machine learning models: case study in the Sichuan Basin. Sci Rep 15, 26706. https://doi.org/10.1038/s41598-025-12391-y Mathiesen, W.T., Bjørshol, C.A., Kvaløy, J.T. et al. (2018) Effects of modifiable prehospital factors on survival after out-of-hospital cardiac arrest in rural versus urban areas. Crit Care 22, 99. https://doi.org/10.1186/s13054-018-2017-x Mourad, K.A., Nordin, L. & Andersson-Sköld, Y. (2022) Assessing flooding and possible adaptation measures using remote sensing data and hydrological modeling in Sweden. Climate Risk Management 38, 100464. https://doi.org/10.1016/j.crm.2022.100464 Muratoglu, A., Bilgen, G.K., Angin, I. & Kodal, S. (2023) Performance analyses of effective rainfall estimation methods for accurate quantification of agricultural water footprint. Water Research 238, 12001. https://doi.org/10.1016/j.watres.2023.120011 Nicholl J, West J, Goodacre S, et al. (2007) The relationship between distance to hospital and patient mortality in emergencies: an observational study. Emergency Medicine Journal 24:665-668. https://doi.org/10.1136/emj.2007.047654 Noble, I.R., S. Huq, Y.A. Anokhin, J. Carmin, D. Goudou, F.P. Lansigan, B. Osman-Elasha, and A. Villamizar (2014): Adaptation needs and options. In: Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part A: Global and Sectoral Aspects. Contribution of Working Group II to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Field, C.B., V.R. Barros, D.J. Dokken, K.J. Mach, M.D. Mastrandrea, T.E. Bilir, M. Chatterjee, K.L. Ebi, Y.O. Estrada, R.C. Genova, B. Girma, E.S. Kissel, A.N. Levy, S. MacCracken, P.R. Mastrandrea, and L.L.White (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, pp. 833-868. Oddo PC and Bolten JD (2019) The Value of Near Real-Time Earth Observations for Improved Flood Disaster Response. Front. Environ. Sci. 7:127. doi: 10.3389/fenvs.2019.00127 Panahi, M. et al. (2023) A Country Wide Evaluation of Sweden's Spatial Flood Modeling With Optimized Convolutional Neural Network Algorithms. Earth’s Future 11:11, e2023EF003749. Pant, R. et al. (2018) Critical infrastructure impact assessment due to flood exposure. Journal of Flood Risk Management 11, 1: 22-33. https://doi.org/10.1111/jfr3.12288 Papilloud, T. et al. (2020) Flood exposure analysis of road infrastructure – Comparison of different methods at national level. International Journal of Disaster Risk Reduction 47, 101548. https://doi.org/10.1016/j.ijdrr.2020.101548 Pell, J.P. et al. (2001) Effect of reducing ambulance response times on deaths from out of hospital cardiac arrest: cohort study. BMJ 322:1385. https://doi.org/10.1136/bmj.322.7299.1385 Pregnolato, M. et al. (2017) The impact of flooding on road transport: A depth-disruption function. Transportation Research Part D: Transport and Environment 55: 67-81. https://doi.org/10.1016/j.trd.2017.06.020 Shan, X., Scussolini, P., Wang, J. et al. (2023) Deficiency of Healthcare Accessibility of Elderly People Exposed to Future Extreme Coastal Floods: A Case Study of Shanghai, China. Int J Disaster Risk Sci 14, 840–857. https://doi.org/10.1007/s13753-023-00513-x Shi, Y., Qian, Y., Jiahong, W., Xi, J., Li, H. & Wang, Q. (2022) A spatial accessibility assessment of urban tourist attractions emergency response in Shanghai. International Journal of Disaster Risk Reduction 74, 102919. https://doi.org/10.1016/j.ijdrr.2022.102919 Shi, H., Zhou, M., Kong, N., Zhang, Y., & Li, X. (2023) A Study on the Accessibility of the Emergency Medical Services for Urban Kindergartens and Nursing Homes Based on Urban Pluvial Flooding Scenarios. Sustainability, 15(13), 10443. https://doi.org/10.3390/su151310443 Shi, J., Wang, H., Zhou, J., & Zhang, S. (2024) Assessment and Improvement of Emergency Rescue Service Accessibility under Urban Waterlogging Disasters. Water, 16(5), 693. https://doi.org/10.3390/w16050693 Sohn, J. (2006) Evaluating the significance of highway network links under the flood damage: An accessibility approach. Transportation Research Part A: Policy and Practice 40, 6: 491-506. https://doi.org/10.1016/j.tra.2005.08.006 Soldà, G., Molsted-Alvesson, H., Montalti, M., Reali, C., Gori, D., Von Schreeb, J. & Ljungman, P. (2025) Responder perspectives on preparedness for healthcare needs of vulnerable populations during floods and heatwaves: a qualitative study in Emilia-Romagna, Italy. BMJ Public Health. 3(2):e002459. doi: 10.1136/bmjph-2024-002459. PMID: 41099035; PMCID: PMC12519328. Swedish Civil Contingencies Agency (MSB) (2023a) MSB Research Agenda 2024–2028. Research for a safer society. Pub. no: MSB2228 – October 2023. ISBN: 978-91-7927-414-6. Swedish Civil Contingencies Agency (MSB) (2023b) Metod för skyfallskartering av tätorter. Publikationsnummer: MSB2260 – november 2023. ISBN: 978-91-7927-435-1. Swedish Transport Administration (Trafikverket) (2018) Regeringsuppdrag om Trafikverkets klimatanpassningsarbete. Publikationsnummer: 2018:195. ISBN 978-91-7725-365-5. Tsang, M. & Scott, D.M. (2020) An integrated approach to modeling the impact of floods on emergency services: A case study of Calgary, Alberta. Journal of Transport Geography 86, 102774. https://doi.org/10.1016/j.jtrangeo.2020.102774 Verschuur, J. et al. (2024) Quantifying climate risks to infrastructure systems: A comparative review of developments across infrastructure sectors. PLOS Climate 3, e0000331. https://doi.org/10.1371/journal.pclm.0000331 Vieira Passos, M., Kan, JC., Destouni, G. et al. (2024) Identifying regional hotspots of heatwaves, droughts, floods, and their co-occurrences. Stoch Environ Res Risk Assess 38, 3875–3893. https://doi.org/10.1007/s00477-024-02783-3 Vieira Passos, M., Kan, JC., Destouni, G. et al. (2025a) Estimation of cascading hydroclimatic hazard impacts on supply systems and associated economic shocks. npj Nat. Hazards 2, 81. https://doi.org/10.1038/s44304-025-00129-9 Vieira Passos, M., Barquet, K., Kan, J. C., Destouni, G., & Kalantari, Z. (2025b) Hydrometeorological resilience assessment of interconnected critical infrastructures. Sustainable and Resilient Infrastructure, 10(3), 267–283. https://doi.org/10.1080/23789689.2024.2446124 Xie, X., Huang, L., Marson, S.M. et al. (2023) Emergency response process for sudden rainstorm and flooding: scenario deduction and Bayesian network analysis using evidence theory and knowledge meta-theory. Nat Hazards 117, 3307–3329. https://doi.org/10.1007/s11069-023-05988-x Yazdani, M. et al. (2022) A modelling framework to design an evacuation support system for healthcare infrastructures in response to major flood events. Progress in Disaster Science 13, 100218. https://doi.org/10.1016/j.pdisas.2022.100218 Yin, J., Yu, D., Lin, N. & Wilby, R.L. (2017) Evaluating the cascading impacts of sea level rise and coastal flooding on emergency response spatial accessibility in Lower Manhattan, New York City. Journal of Hydrology 555: 648-658. https://doi.org/10.1016/j.jhydrol.2017.10.067 Yu, D., Yin, J., Wilby, R.L. et al. (2020) Disruption of emergency response to vulnerable populations during floods. Nat Sustain 3, 728–736. https://doi.org/10.1038/s41893-020-0516-7 Zhang, Y., Zhou, M., Kong, N., Li, X., & Zhou, X. (2022a) Evaluation of Emergency Response Capacity of Urban Pluvial Flooding Public Service Based on Scenario Simulation. International Journal of Environmental Research and Public Health, 19(24), 16542. https://doi.org/10.3390/ijerph192416542 Zhang, Y., Li, X., Kong, N., Zhou, M., & Zhou, X. (2022b) Spatial Accessibility Assessment of Emergency Response of Urban Public Services in the Context of Pluvial Flooding Scenarios: The Case of Jiaozuo Urban Area, China. Sustainability, 14(24), 16332. https://doi.org/10.3390/su142416332 Ziya, O. & Safaie, A. (2023) Probabilistic modeling framework for flood risk assessment: A case study of Poldokhtar city. Journal of Hydrology: Regional Studies 47, 101393. https://doi.org/10.1016/j.ejrh.2023.101393 Cite Share Download PDF Status: Under Review Version 1 posted Reviewers agreed at journal 11 Mar, 2026 Reviewers invited by journal 06 Mar, 2026 Editor assigned by journal 17 Jan, 2026 First submitted to journal 16 Jan, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8600985","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":601687067,"identity":"96b8d9e7-98e3-47e8-a998-d410efacdfbf","order_by":0,"name":"Daniel Adshead","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAtklEQVRIiWNgGAWjYFCCBDYgwSwnwcBDohZjCTZStSTOIFoLf3v6swcfaqzTZ87vPcD4o4IILRJn3pgbzjiWnjubjS+BmecMMdbcyGGT5mE7nDuPjceAmbGNCB3yN9KfSf/5dzhdDqiF8ec/IrQY3Egwk2ZsO5wgDdTCwNtAhBbDM2/MJHv70g1ntuUYHOY5RoQWuePpzyR+fLOWlzh8xvDhjxoitKCAA6RqGAWjYBSMglGAAwAAqOcy9C3SkOcAAAAASUVORK5CYII=","orcid":"https://orcid.org/0000-0002-0829-925X","institution":"KTH Royal Institute of Technology: Kungliga Tekniska Hogskolan","correspondingAuthor":true,"prefix":"","firstName":"Daniel","middleName":"","lastName":"Adshead","suffix":""},{"id":601687068,"identity":"d16f2bac-9e8b-410b-84b9-bf8452f312e9","order_by":1,"name":"Martin Jonsson","email":"","orcid":"","institution":"Karolinska Institutet","correspondingAuthor":false,"prefix":"","firstName":"Martin","middleName":"","lastName":"Jonsson","suffix":""},{"id":601687069,"identity":"87cb4ce4-b0bb-43b8-a5a4-0ee93d52708a","order_by":2,"name":"Nikolaus Mezger","email":"","orcid":"","institution":"Karolinska Institutet","correspondingAuthor":false,"prefix":"","firstName":"Nikolaus","middleName":"","lastName":"Mezger","suffix":""},{"id":601687070,"identity":"95b5495e-17e8-4655-b51b-01c142b9ab93","order_by":3,"name":"Georgia Destouni","email":"","orcid":"","institution":"Stockholm University: Stockholms Universitet","correspondingAuthor":false,"prefix":"","firstName":"Georgia","middleName":"","lastName":"Destouni","suffix":""},{"id":601687071,"identity":"3bfc218a-de9e-4ece-8404-f0147b9f4a00","order_by":4,"name":"Petter Ljungman","email":"","orcid":"","institution":"Karolinska Institutet","correspondingAuthor":false,"prefix":"","firstName":"Petter","middleName":"","lastName":"Ljungman","suffix":""}],"badges":[],"createdAt":"2026-01-14 11:06:27","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8600985/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8600985/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":104424020,"identity":"e056c97c-e4a5-4fd5-b4e6-bd84567f44c6","added_by":"auto","created_at":"2026-03-11 14:23:06","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":169021,"visible":true,"origin":"","legend":"\u003cp\u003eAnalytical framework and data used in the study, including mapping of priority emergency response routes, estimation of flood risk, and response time analysis under flood conditions (DFI\u0026gt;0).\u003c/p\u003e","description":"","filename":"1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8600985/v1/3515827ec81cfacc1301eb9d.jpg"},{"id":104424018,"identity":"3755da88-db5e-493c-95a6-1a569d17697e","added_by":"auto","created_at":"2026-03-11 14:23:06","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":194532,"visible":true,"origin":"","legend":"\u003cp\u003eExposure of all calculated response routes (2021-2025) to historical flood risk (quantified by the Daily Flood Index (DFI), 2015-21), based on samples at 100m distances along the full estimated route from ambulance dispatch to response location. (a) Mean DFI for all routes including those with no flood risk assessed (DFI ≤0), with lighter-coloured routes being those more frequently exposed over the test period, (b) Mean DFI for all routes subject to flood exposure (DFI\u0026gt;0), with lighter-coloured routes indicating greater flood exposure intensity.\u003c/p\u003e","description":"","filename":"2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8600985/v1/030599e07dfb58b5f220f615.jpg"},{"id":104424017,"identity":"83c0825a-3bca-4c88-814f-d44da4bf2419","added_by":"auto","created_at":"2026-03-11 14:23:06","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":179114,"visible":true,"origin":"","legend":"\u003cp\u003eMean Daily Flood Index (DFI) for all routes with flood exposure (DFI\u0026gt;0), zooming in on key areas identified as particularly vulnerable to flood exposure across the entire country: (a) Västra Götaland (western Sweden), (b) Stockholm (eastern), (c) Skåne (southern), (d) Norbotten (northern), (e) Östergötland (central). Lighter-coloured routes indicate greater flood exposure intensity.\u003c/p\u003e","description":"","filename":"3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8600985/v1/538057f88438fa5d83102c64.jpg"},{"id":104424024,"identity":"b8a4b3f8-2935-47f7-81a4-ff16ea076b4b","added_by":"auto","created_at":"2026-03-11 14:23:06","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":89507,"visible":true,"origin":"","legend":"\u003cp\u003eFlood-exposed routes (DFI\u0026gt;0) where ambulance response time (in seconds per kilometre) exceeds the median response time for non-exposed ambulance responses in the municipality, indicating actual observed delays occurring on days with flood exposure. Darker red hotspots indicate greater observed ambulance delay above the median during flooding.\u003c/p\u003e","description":"","filename":"4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8600985/v1/116ad2b2f38bdf508253fcef.jpg"},{"id":104780227,"identity":"3ab231c3-1e67-4763-8cc1-8b8f9b9ae86f","added_by":"auto","created_at":"2026-03-17 07:51:31","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":89915,"visible":true,"origin":"","legend":"\u003cp\u003eTop 20 municipalities by number of ambulance calls coinciding with flood exposure areas and resulting in a slower response time than the municipal median where no flood risk present on the day of the call. Counts are shown on the left-hand axis, and average ambulance response delays (above the median during non-exposed periods)) in second per kilometre are shown on the right-hand axis.\u003c/p\u003e","description":"","filename":"5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8600985/v1/f3f7b23f3ed97708c151cc1d.jpg"},{"id":104424025,"identity":"91855356-bc66-4dea-a115-564976d4ea1e","added_by":"auto","created_at":"2026-03-11 14:23:06","extension":"jpg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":87974,"visible":true,"origin":"","legend":"\u003cp\u003eGeographic distribution of the top 20 Swedish municipalities with both flood exposure and higher observed response delays (s/km) in comparison with the municipal median ambulance response\u003c/p\u003e","description":"","filename":"6.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8600985/v1/e4d57cb6b1c2a23cbc226722.jpg"},{"id":104784397,"identity":"80a9efec-79c4-4a5c-946c-7f3985b00ff5","added_by":"auto","created_at":"2026-03-17 08:07:12","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1448123,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8600985/v1/32188d62-c08a-4351-bcde-0fa8cd34a175.pdf"}],"financialInterests":"","formattedTitle":"Assessing and screening flood exposure risks for emergency response","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eHydro-climatic changes, e.g. in occurrence frequency and intensity of floods (Vieira Passos et al. 2024), are increasingly challenging access to critical services essential for achieving social and economic development (Adshead et al. \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Vieira Passos 2025a). The interconnectedness of modern infrastructure, spanning energy, transportation, water, and communication networks, as well as social infrastructure like healthcare facilities and schools, amplifies this challenge (Verschuur et al. \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Disruptions to key access infrastructure such as road and transportation networks have the potential to create cascading failures across sectors that jeopardize the delivery of vital services (Vieira Passos et al. \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2025b\u003c/span\u003e). Flooding, in particular, not only damages physical assets but may also slow operations and critically hamper the accessibility of emergency services, creating a situation where the societal systems intended to manage crises are themselves compromised. The potential for high-frequency, localized flash flooding to affect response times, even at moderate intensities, poses a significant, understudied operational risk to these services.\u003c/p\u003e \u003cp\u003eProactive adaptation measures, spanning local to national scales, are needed to reduce social risks, safeguard service continuity, and enhance resilience against climate-driven disruptions. Addressing the interplay between physical infrastructure and societal vulnerability requires a systemic approach, especially as climate adaptation policy shifts discourse towards 'riskification' (Englund et al. 2023). This means reframing climate change impacts from an immediate, exceptional 'threat' into a calculable, manageable 'risk' that can be systematically managed through established governmental tools, such as formal risk assessments, policy recommendations from expert agencies, and integration into existing planning structures, rather than necessitating extraordinary measures or entirely new governance frameworks.\u003c/p\u003e \u003cp\u003eThese approaches require localized flood risk assessment techniques to consider the potential climate threats to infrastructure assets, and to manage them accordingly. Such techniques are increasingly enabled by advances in climate and hydrological data availability, synthesis, and modelling, which have transformed our ability to predict and adapt to the impacts of extreme weather events (Panahi et al. \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), including to multi-hazard exposure and at localized level with high resolution (Gudiyangada Nachappa et al. 2020). New approaches integrate high-resolution climate scenario projections with hydrological simulations to demonstrate how probabilistic models (Ziya et al. 2023) can predict flood risks and their cascading effects on infrastructure. These models, increasingly leveraging satellite data and machine learning, are capable of monitoring flood-prone regions and assessing their potential impacts with unprecedented spatial and temporal detail (Lu et al. \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eHere, we address the impacts of flooding on healthcare, with specific focus on emergency service provision. Much of the research on hydrometeorological hazards to healthcare has addressed cyclone-related flood impacts to hospitals and care facilities (Abebe et al. \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Yazdani et al. \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2022\u003c/span\u003e); however, the infrastructure supporting access to these facilities is equally critical (Alam et al. \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Simulation studies have been key to explore the impacts of flooding on healthcare access, offering valuable insights into potential vulnerabilities and mitigation strategies. These approaches generally focus on modelling the spatial coverage of urban emergency response services during various flood magnitudes to ensure compliance with mandatory response times set by national or municipal health authorities (Coles et al. \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Green et al. \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Yu et al. \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), using models such as hydrodynamic urban waterlogging (Shi et al. \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Analysis of hazard impacts has also advanced beyond simple inundation mapping by assigning flood depths to road segments, allowing the network to be modelled as partially degraded based on flood depth and reduced vehicular speed (Tsang et al. 2020). This helps identify sites with restricted access and quantify vulnerable populations that may no longer meet mandatory response times.\u003c/p\u003e \u003cp\u003eScenario-based evaluations have been applied more widely to assess flood impacts on urban public services, simulating inundation across various pluvial flooding scenarios to calculate emergency response capacity and accessibility (Zhang et al. \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2022a\u003c/span\u003e; Zhang et al. \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2022b\u003c/span\u003e), with specific attention to the unique challenges of reaching vulnerable groups. Assessments have quantified the increased exposure and delay time for essential services to facilities such as kindergartens (Shi et al. \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), nursing homes (Johnson \u0026amp; Yu \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) and elderly populations (Shan et al. \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), and economically-important areas like tourist sites (Shi et al. \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) under severe pluvial and coastal flooding scenarios. Furthermore, these scenario approaches have been used to evaluate the cascading impacts of sea level rise and coastal flooding, simulating service areas for multiple responders under varying traffic and flood conditions (Yin et al. \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Other studies utilize Bayesian networks to examine the intricate evolution of emergency responses during specific heavy rain events, allowing decision-makers to understand key nodes and to take emergency response measures in time (Xie et al. \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eDespite these advances, much of the existing research relies on simulated flood scenarios and theoretical models of traffic disruption, with limited validation against historical data. This reliance on simulation, while necessary, shows and leaves a critical gap: access to empirical, high-resolution data to validate risk models and inform localised resilience efforts. Bridging this gap requires connecting advanced environmental modelling with local operational data observed over a large scale.\u003c/p\u003e \u003cp\u003eEmpirical analysis of emergency response data has been applied in a handful of studies to understand the specific, time-sensitive demands of health emergencies like out-of-hospital cardiac arrests (OHCA) (Pell et al. \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2001\u003c/span\u003e), and the factors associated with response time variation (Harries et al. 2025; Mathiesen et al. \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Prompt ambulance response and reduced travel distance is crucial for improving survival and neurological outcomes (Nicholl et al. \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2007\u003c/span\u003e), while differences in survival rates between fast and slower responses underscores that minutes are critical (Holm\u0026eacute;n et al. 2020). Timely interventions not only boost immediate survival rates but also improve the chances of discharge with good neurological outcomes (B\u0026uuml;rger 2018). Flood-induced disruptions to emergency medical services (EMS) exacerbate these risks by prolonging response times. Sweden is a country with extensive geographic and climatic diversity, making it relevant and interesting as a test case for understanding how flooding affects EMS and how critical that may be for developing resilience strategies.\u003c/p\u003e \u003cp\u003eIn this study, we develop and test a data-driven framework for assessing flood vulnerability on time-critical emergency routes, with Sweden as a test case. This case enhances our understanding of theoretical route network risks by facilitating integration of novel, high-resolution datasets, including empirical dispatch records from a centralised national emergency service and established, high resolution hydrological risk data. These datasets allow us to construct a robust, validated baseline for risk modelling and adaptation planning, with a primary goal to identify, prioritise, and spatially localise the most vulnerable segments of the emergency response network for future infrastructure and adaptation investments.\u003c/p\u003e"},{"header":"2. Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Data and analysis framework\u003c/h2\u003e \u003cp\u003eThis study develops a two-part analytical framework designed to deliver a comprehensive assessment of flood vulnerability and performance risk for ambulance response routes and applies and tests this for the case of Sweden (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). The analysis is structured to first establish a long-term strategic vulnerability profile and then provide a performance screening by examining the relationship between real-time flood conditions and response times. This analysis is empirical in its foundation but relies on a systematic approximation of real-world exposure and travel time. Specifically, for the case of Sweden, we utilize three primary datasets:\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eEmpirical emergency response data: The actual date- and time-stamped start (dispatch location) and end (patient location) points of 117,875 ambulance dispatches from 2021\u0026ndash;2025 were used, providing a true measure of operational demand and observed response times. This cohort of dispatches containing both start and end location was filtered from a larger dataset of over 255,000 dispatches provided by the Emergency Medical Dispatch Centre (EDMC) and focuses exclusively on suspected cardiac arrests (unconscious adult without normal breathing) calls. The data covered 15 of Sweden\u0026rsquo;s 21 regions.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eModelled flood risk data: We used the Daily Flood Index (DFI) as a high-resolution, empirically-derived proxy for the daily presence of localized flood conditions based on historical rainfall and soil characteristics (Deo et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2015\u003c/span\u003e), covering the period 2015\u0026ndash;2021 and modelled for the whole of Sweden (Vieira Passos et al. \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). This DFI dataset provides a measure of flood risk frequency on a 100-metre grid derived from historical rainfall data but does not quantify the precise flood impact (e.g. millimetres of inundation) on any given day. This index is based on effective and accumulated precipitation and therefore accounts for various hydrological factors affecting the flood risk, such as evapotranspiration and soil saturation at local scales (Muratoglu et al. \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Brocca et al. \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). A DFI value greater than zero was used to indicate flood risk conditions.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eAlgorithmic route data: The actual GPS track taken by the ambulance is not logged in the EMDC data. Instead, we used the shortest-path algorithm to calculate a standardized, reproducible baseline route for each dispatch, sourced from OpenStreetMap. Our analysis, therefore, compares observed, real-world travel times both under the daily flood risk condition (DFI\u0026thinsp;\u0026gt;\u0026thinsp;0) and when no flood risk is assessed (DFI\u0026thinsp;\u0026le;\u0026thinsp;0). This allows us to identify when a high-flood risk day is associated with a longer-than-expected travel time, signalling a system performance degradation that approximates the empirical operational impact of flooding.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2. Creating a historical flood risk profile\u003c/h2\u003e \u003cp\u003eThe first phase of the analysis focused on historical flood risk assessment. We defined a \u0026lsquo;health access network\u0026rsquo; based on over 117,000 usable empirical routes from 2021\u0026ndash;2025. For each call, the route was computed as the shortest path by length, from the point of the ambulance dispatch to the call location, using the road network, providing a standardized, reproducible baseline against which risk can be consistently measured. Leveraging this large sample size of real-world dispatches provides a robust estimation of mobility corridors essential for critical emergency response access. While based on algorithmic efficiency rather than observed GPS tracks, this extensive dataset identifies the high-frequency pathways that must remain functional to meet the bulk of public health demand across the region.\u003c/p\u003e \u003cp\u003eTo intersect these linear routes with the gridded flood data, each computed route was densified by placing a sample point every 100 metres along the route. Drawing on previous methods of flood exposure calculation for roads (Sohn \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Papilloud et al. \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Pregnolato et al. \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) and other critical infrastructure (Pant et al. \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), these sampled points were then used to query the entire 7-year (2015\u0026ndash;2021) flood risk (DFI) history, where intersection of any point with a grid cell having a DFI value greater than zero on the same day classifies the route as \u0026lsquo;exposed\u0026rsquo;. A route's long-term vulnerability was determined by calculating annualized flood risk days per year, which quantifies the average number of high-risk days per year for each critical route. This phase yields the strategic vulnerability map, identifying segments of the current network most frequently exposed to historical flood risk. Because the DFI is designed to measure the potential for daily flooding, flood depth is not included as a factor in this study.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3. Performance screening and critical hotspot localisation\u003c/h2\u003e \u003cp\u003eThe second phase of the analysis focused on observed performance and risk filtering. This analysis was limited to the single year of 2021 where our usable route data and the full flood risk data overlap. This 2021 cohort enables a direct audit of response time performance under documented flood conditions. To allow for fair comparison across different route lengths, the primary response delay metric was normalised by distance, creating the ambulance response delay rate (seconds per kilometre), which measures route efficiency rather than simple duration. For each 2021 call, its real-time exposure was determined by querying DFIs for that specific route on the same day. Routes with any sample point intersecting with a DFI value greater than zero were classified as \"Exposed\"; the others were classified as \"Non-Exposed\". The classification of 'Exposed' indicates the presence of a high-frequency flash flood risk condition on that specific day, without specifying the level of inundation or flood intensity.\u003c/p\u003e \u003cp\u003eTo address risk localisation, a spatial hotspot map was generated to serve as a risk filtering tool. This map visually isolates only those routes that, on the same day in 2021, (1) were flagged as \"Exposed\" and (2) exhibited a delay rate greater (i.e. a normalised response time slower) than the municipal median, providing a sensitive tool for identifying anomalous delays that exceeded the baseline for that specific geographic area.\u003c/p\u003e \u003cp\u003eThis approach is designed as a risk-screening tool for infrastructure planning, not an analysis of extreme outliers. Filtering only the most extreme delays would mask the higher frequency of routes where performance deterioration begins to occur. By capturing every ambulance dispatch case where performance fell into the slower half of the distribution during flooding, we provide a more comprehensive, actionable set of locations for proactive mitigation, such as drainage improvements or alternative route planning, ensuring that minor but frequent performance drops are also addressed. We do not aim to prove that flooding always slows ambulance routes down, but rather to identify the specific locations where the combination of flood exposure and poorer than average response performance actually occurred, creating a critical vulnerability.\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Results","content":"\u003cp\u003eThis section presents the spatial and operational impacts of flood risk on ambulance response routes in Sweden, first characterising long-term historical exposure across the national response network and then identifying route segments where flood exposure coincides with observed response delays.\u003c/p\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e3.1. Historical flood risk profile of emergency response routes\u003c/h2\u003e \u003cp\u003eEstablishing the long-term vulnerability of the operational ambulance access network to flood exposure reveals significant spatial heterogeneity across Sweden. Figure\u0026nbsp;2(a) illustrates the overall mean DFI encountered along all sampled routes, including the majority of days with no significant flood risk, highlighting lighter-coloured areas where flood events are frequent enough to influence the long-term average exposure across the entire route network. These lighter-coloured routes, indicating more frequent flood exposure, are concentrated in several distinct regions. Notably, routes in northern Sweden (particularly Norrbotten), parts of the west coast (V\u0026auml;stra G\u0026ouml;taland), and the southern region of Sk\u0026aring;ne, exhibit persistently higher mean DFI values.\u003c/p\u003e \u003cp\u003eIn contrast, Fig.\u0026nbsp;2(b) focuses exclusively on the intensity of flood risk to the emergency response network, reporting the mean DFI only for the days with a flood exposure (DFI\u0026thinsp;\u0026gt;\u0026thinsp;0). This metric isolates the severity of the hazard when it occurs, with the spatial pattern observed here highlighting areas facing the most intense flood events. The lighter-coloured routes in this panel delineate regions where the average flood hazard intensity is substantially higher when it is present, even if it is less frequent over time. This is particularly pronounced in northern municipalities, certain inland municipalities in western and southern Sweden (e.g. V\u0026auml;stra G\u0026ouml;taland, Sk\u0026aring;ne, and Kalmar), as well as large parts of central Sweden (J\u0026ouml;nk\u0026ouml;ping, \u0026Ouml;sterg\u0026ouml;tand). This comparison reveals a crucial distinction: densely populated areas, like the Swedish capital Stockholm, face a high frequency of low-to-moderate exposure (as suggested by the overall mean DFI), while other areas face less frequent but significantly more intense flood exposure, i.e. higher mean DFI when a flood event does occur.\u003c/p\u003e \u003cp\u003eTo better contextualise the spatial flood vulnerability, Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e3\u003c/span\u003e provides a focused view on mean DFI across five regions identified as having high vulnerability based on flood exposure intensity. Summarizing emergency response route exposure by municipality further quantifies these spatial trends. Overall exposure, measured by the proportion of flooded days, emerges as most severe in the northern Swedish municipalities of Lule\u0026aring; (48.75 days/year), \u0026Ouml;vertorne\u0026aring; (37.53 days/year), and Haparanda (36.63 days/year). Conversely, while Link\u0026ouml;ping has a moderate flood exposure frequency (11.68 days/year), its mean exposure intensity when flooded (with a mean DFI of 0.461, indicating severe hydrological stress) is the highest recorded, indicating that recorded routes within this municipality face the most severe flood magnitudes on affected days. Other municipalities with high intensity include Pajala (mean DFI of 0.434), \u0026Ouml;vertorne\u0026aring; (0.431), and Oskarshamn (0.449). This granular level of data is essential for emergency planning, distinguishing between areas needing high-frequency, low-level mitigation (e.g. urban drainage maintenance) versus areas requiring high-consequence preparedness (e.g. robust alternative routing or specialized rescue).\u003c/p\u003e \u003cp\u003eAcross the full set of over 117,000 individual ambulance routes, the historical exposure analysis reveals that the top exposed routes face up to 17% exposure to flood risk, corresponding to over 62 annualized flooded days per year. The majority of these top-exposed routes also share a high mean DFI intensity when flooded (approximately 0.40, denoting significant inundation levels), although others reach as high intensity as 0.62 (representing more extreme events). Clustering of nearly identical exposure metrics across consecutive entries suggests that these routes often overlap or are used for multiple ambulance calls, representing single, critical vulnerability points that service multiple points of origin and destination. Addressing these few, highly exposed emergency corridors would yield the greatest return on investment in risk reduction.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e3.2. Screening risk-affected routes for ambulance response delays\u003c/h2\u003e \u003cp\u003eThe second stage shifts the focus from long-term structural vulnerability to real-time operational performance screening under documented flood risk conditions. Due to limited sample sizes for flood-exposed calls in many municipalities for the one analysed year (often fewer than 10), this study adopts a targeted, risk-screening methodology that focuses on quantifying the magnitude of performance degradation where flood exposure coincides with demonstrably poorer than average service delivery, rather than formal statistical significance testing.\u003c/p\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e4\u003c/span\u003e presents a hotspot map identifying these critical routes. This filters the 2021 dataset to only include emergency calls where 1) any section of the route had flooding (DFI\u0026thinsp;\u0026gt;\u0026thinsp;0) on the date of the dispatch and 2) where the recorded ambulance response delay rate (seconds per kilometre) exceeded a baseline derived from non-exposed calls. Portions of these exposed routes are indicated with red markers based on their intersection with flood grid cells and are weighted by frequency and magnitude of use during delayed responses.\u003c/p\u003e \u003cp\u003eWe use a local delay rate baseline, defined as the municipal median response time (s/km) of non-exposed routes, to identify slow response times relative to overall system performance. Using the median delay rate in each municipality allows detection of anomalies relative to local operating conditions, as the national median can mask significant variations in inherent emergency response standards between regions (e.g. rural vs. urban). Using the local non-exposed baseline equitably reveals whether the flood hazard introduces a significant local impairment to service delivery and provides a more actionable assessment of flood vulnerability tailored for regional emergency managers. The average response delay rate (seconds per kilometre, s/km) for all flood-exposed routes found to be slower than the local municipal median was 81.7 s/km.\u003c/p\u003e \u003cp\u003eThe hotspots of anomalously high delay rates do not simply show areas of high flood risk, but isolate segments where flood exposure coincides with demonstrable delay in ambulance service delivery. This shifts our assessment from identifying where flooding can happen (hazard-centric) to identifying where flooding may cause the most harm to emergency service delivery (risk-centric).\u003c/p\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e5\u003c/span\u003e provides a quantitative summary of the most critically affected Swedish municipalities based on the local screening criterion, highlighting areas where flood risk-exposed ambulance responses exceeded the non-exposed median for the municipality. The top 20 municipalities by average delay for routes where flood exposure coincided with poor performance are noted. This highlights certain areas where, despite low sample sizes, observed delays were severe when they did occur. Figure\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e6\u003c/span\u003e illustrates the distribution of these high-risk areas across Sweden, clearly showing the disproportionate impact of delays in geographically diverse municipalities, such as those in northern Sweden (e.g. the municipalities of Boden, G\u0026auml;llivare, Kalix), the central V\u0026auml;rmland and Dalarna regions (e.g. Leksand, Ludvika, Filipstad), alongside significant clusters in regional urban centres in central and western Sweden (e.g. Karlstad and Gothenburg).\u003c/p\u003e \u003cp\u003eThe overall findings indicate that while severe flood intensity is often observed in sparsely populated northern areas, the operational risk (ambulance response delays) is likely a complex interaction between flood characteristics and the underlying volume and criticality of the ambulance routes, with results confirming the existence of operational delay hotspots in screened routes and segments. These identified hotspot routes represent a critical set for targeted resilience interventions, as they were simultaneously exposed to flooding and incurred a measurable operational penalty. In urban areas, higher volumes of traffic and difficulty of finding quick alternative detours mean that even a minor disruption can translate into a substantial, measurable operational failure (Li et al. \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2021\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eThe strategic value of this study lies in shifting governmental risk assessment from static hazard mapping to dynamic, operationally focused vulnerability analysis. Traditional climate risk assessments often focus on infrastructure exposure (e.g. which roads are in a flood plain) or damage (e.g. what it costs to repair a bridge). Here, however, we shift the focus to service discontinuity, translating environmental risk into a measurable public health consequence such as minutes lost in an emergency response. By intersecting historical flood data with a network of real-life ambulance routes, this type of analysis compels planners to view flood risk not just as an environmental or civil engineering problem, but as a direct constraint on emergency service capacity and a threat to health equity. This reframing elevates the discussion from routine infrastructure maintenance to core national healthcare priority.\u003c/p\u003e \u003cp\u003eFurthermore, the methodology acts as a necessary inter-sectoral bridge. It translates complex hydrological science (e.g. via the Daily Flood Index) into actionable, network-based metrics that are instantly recognizable to transport planners, and finally into public health outcomes (emergency response times) that compel action from healthcare providers and policymakers. Here we used DFI as the hydrological basis for the emergency network metrics because this has been tested comparatively with other types of indicators and found to be the best-performing one for representing actual flood occurrence across Sweden over the period 1922\u0026ndash;2021, i.e. for the geography and time of the present test case study. Alternative flood occurrence indicators (Vieira Passos et al. \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) include the Standardized Precipitation Index, Standardized Precipitation and Evapotranspiration Index, and the Standardized Streamflow Index; in other applications, any of these alternative indices (also including the DFI), or other possible combinations of observed and modelled flood data, can be used for the translation from hydrology to emergency network metrics, depending on which is found or considered to best represent actual flood conditions in each specific application.\u003c/p\u003e \u003cp\u003eIn general, this translation is essential for overcoming the traditional institutional silos that separate climate, transport, and health agencies, and for introducing flexible and collaborative response mechanisms, which have been identified as lacking in previous research on health emergency preparedness to floods and other hazards (Sold\u0026agrave; et al. 2025). Without such a unified metric, climate adaptation plans risk being isolated within environmental departments, failing to influence the budget allocations of agencies responsible for road maintenance or ambulance deployment. Robust risk and exposure maps provide a data authority that previous theoretical models lacked, establishing a transparent and empirical case for resource prioritization tied directly to time-critical public services. Isolating high-demand ambulance corridors that are both essential to health service delivery and vulnerable to flood risk enables a fundamental shift toward risk-informed investment that proactively builds system resilience, rather than merely reacting to the next extreme weather event.\u003c/p\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e4.1. Policy relevance for resilient medical emergency response\u003c/h2\u003e \u003cp\u003eThe findings from this analysis point directly to two distinct categories of measures among a broader scope of adaptation options (Noble et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Azhar et al. \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2025\u003c/span\u003e): built environment/engineering solutions, and operational planning measures. The built solutions fall primarily within the domain of transport agencies and municipal road authorities (Mourad et al. \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). For road segments identified with high historical flood exposure, engineering and design reviews are warranted. This includes traditional infrastructure measures, such as improving side-drainage channels and replacing undersized culverts to manage increased runoff from pluvial (flash) floods, which the flood risk model is well-suited to identify. The data also supports the implementation of nature-based solutions, which involves incorporating green infrastructure, such as establishing permeable pavements or roadside bioswales to vulnerable segments to slow and filter runoff. In cases of extreme vulnerability, such as routes with a high annual flood risk and no viable detours, more significant interventions like physically raising key road segments or hardening bridge approaches may be justified. The exposure data provides the evidence to target these costly and protective built options where they will have the most significant impact on maintaining health services.\u003c/p\u003e \u003cp\u003eOperational and planning measures fall to emergency service providers and the administrative bodies responsible for healthcare delivery to residents. The maps of high-risk routes should be used to formally pre-designate and vet flood-resilient alternate routes (Oddo \u0026amp; Bolten \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). These alternates must be integrated into dispatch systems so they can be immediately suggested, rather than relying solely on a driver's ad-hoc local knowledge during a crisis. In the longer term, flood risk data itself could be integrated as a real-time predictive layer. Future flood scenarios forecasting high risk in a given area could automatically flag a standard route as \"at-risk,\" prompting dispatchers to use the pre-vetted alternate path and thereby reducing the risk of a critical delay. This shift towards data-driven, pre-emptive operational planning ensures continuity of service even when physical infrastructure is compromised.\u003c/p\u003e \u003cp\u003eIn Sweden, the test case of this study, national climate adaptation policy centres on the reframing of climate change as a manageable risk that must be integrated into existing tools and strategies. This study provides exactly such a tool, directly linking observed historical risk to actionable agency mandates. For the Swedish Transport Administration (Trafikverket), which manages national roads, the historical exposure of priority routes can be used as an evidence base to prioritise maintenance and adaptation budgets as suggested in its adaptation strategy update (Trafikverket 2018). Road segments with notable flood risk delays can be flagged for targeted hydraulic assessments or drainage improvements, focusing resources where the empirical risk to health services is highest. For municipalities (Kommuner) and county boards (L\u0026auml;nsstyrelser), which are tasked with local climate adaptation planning, this analysis provides the high-resolution, localized data they need for their risk and vulnerability assessments, moving beyond generic county-level climate reports.\u003c/p\u003e \u003cp\u003eThe study also provides a concrete example of systemic risk analysis for the Swedish Civil Contingencies Agency (MSB), the current research agenda of which emphasizes research and development needs for understanding the effects of climate risks on critical response infrastructure (MSB 2023a). Notably, MSB has piloted detailed flash flood mapping methods, including projections of more extreme, higher return period floods, in Karlstad and Uddevalla \u0026ndash; two urban areas where our risk-centric screening of observed ambulance delays detected operational degradation during more frequent, localized flood risk events (MSB 2023b). For the Swedish Regions (Regioner) and emergency medical dispatch centres (EMDCs), the findings provide a strategic tool for long-term operational planning and for integrating flood risk into national emergency planning protocols. A high-risk route that serves as the only access to a populous area, for example, may highlight a need for new dispatch stations in relation to the vulnerability, or at least a formal plan for pre-positioning resources during high-risk weather forecasts.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e4.2. Future directions and global applications\u003c/h2\u003e \u003cp\u003eLimitations of this study include the temporal scope of the ambulance esponse-time analysis (a single year), which limits the statistical power to detect rare events. A second limitation is the reliance on modelled \"shortest path\" routing: while this is a standard and necessary methodological assumption for creating a reproducible baseline, these routes do not represent actual GPS tracks, which could deviate based on driver choice, geographical obstacles, real-time traffic, or dispatcher-led re-routing. Despite these limitations, the framework itself is highly generalizable, stemming from the fact that the core components of the problem (climate-driven hazards impacting physical road networks and the societal need for rapid emergency response) are universal.\u003c/p\u003e \u003cp\u003eWhile the specific hazards, data availability, and governance structures may differ, the fundamental link between infrastructure integrity and time-critical health outcomes is a shared global challenge. The hazard layer we have used here (Daily Flood Index) could be replaced with other relevant extreme event models and indicators, such as wildfire risk perimeters, landslide susceptibility maps, or coastal inundation scenarios. This hazard layer method would be particularly valuable in regions with low infrastructure redundancy, such as areas with single-access valley roads or coastal communities, where such an analysis could clearly identify critical hotspot points of failure. In data-scarce environments, the empirical routes could be proxied using mobile phone trace data, with hazard layers derived from publicly available satellite imagery, demonstrating the framework's broad applicability for climate adaptation planning worldwide.\u003c/p\u003e \u003cp\u003eUltimately, this analysis offers a powerful test case study for how to synthesise hydrological science, network analysis, and public health metrics. It provides the evidence required to shift from reactive extreme weather response toward proactive, spatially-informed adaptation planning to secure emergency health access in the face of escalating threats posed by hydro-climatic changes.\u003c/p\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003ch3\u003eAcknowledgments\u003c/h3\u003e\n\u003cp\u003eThis project was supported by Stockholm Trio for Sustainable Actions (STSA). The authors also acknowledge Fredrik Byrsell and Alessandro Berti for their efforts in securing data.\u003c/p\u003e\n\u003ch3\u003eEthics approval and consent to participate\u003c/h3\u003e\n\u003cp\u003eThe authors have obtained ethics approval from relevant authorities in order to use the anonymised emergency response data. No human participants were involved in this study.\u003c/p\u003e\n\u003ch3\u003eConsent for publication\u003c/h3\u003e\n\u003cp\u003eAll authors agree with the content and consent to its publication.\u003c/p\u003e\n\u003ch3\u003eCompeting interests\u003c/h3\u003e\n\u003cp\u003eThe authors declare there are no conflicts of interest for this manuscript.\u003c/p\u003e\n\u003ch3\u003eAuthor contributions\u003c/h3\u003e\n\u003cp\u003e\u003cstrong\u003eDaniel Adshead:\u003c/strong\u003e Conceptualization, Methodology, Data curation, Formal analysis, Visualization, Writing – Original Draft, Funding acquisition, Project administration\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMartin Jonsson:\u003c/strong\u003e Data curation, Writing – Review \u0026amp; Editing\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eNikolaus Mezger:\u003c/strong\u003e Data curation, Writing – Review \u0026amp; Editing\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eGeorgia Destouni:\u003c/strong\u003e Methodology, Writing – Review \u0026amp; Editing\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePetter Ljungman:\u003c/strong\u003e Conceptualization, Methodology, Writing – Review \u0026amp; Editing\u003c/p\u003e\n\u003ch3\u003eFunding\u003c/h3\u003e\n\u003cp\u003eThe authors acknowledge funding from the Stockholm Trio for Sustainable Actions (STSA).\u003c/p\u003e\n\u003ch3\u003eAvailability of data and materials\u003c/h3\u003e\n\u003cp\u003eThe raw data contain confidential emergency response data that are not publicly available. Access may be granted to qualified researchers upon reasonable request, subject to approval by the relevant ethics committee and the data owner(s), and the signing of a data-sharing agreement. Derived and anonymised data used in the analysis are available upon request from the corresponding author.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAbebe, Y.A., Pregnolato, M. \u0026amp; Jonkman, S.N. (2025) Flood impacts on healthcare facilities and disaster preparedness \u0026ndash; A systematic review. International Journal of Disaster Risk Reduction 119, 105340. https://doi.org/10.1016/j.ijdrr.2025.105340\u003c/li\u003e\n\u003cli\u003eAdshead, D., Paszkowski, A., Gall, S.S. et al. (2024) Climate threats to coastal infrastructure and sustainable development outcomes. Nat. Clim. Chang. 14, 344\u0026ndash;352. https://doi.org/10.1038/s41558-024-01950-2\u003c/li\u003e\n\u003cli\u003eAlam, M.S. et al. (2024) Large-scale modeling of hurricane flooding and disrupted infrastructure impacts on accessibility to critical facilities. Journal of Transport Geography 116, 103852. https://doi.org/10.1016/j.jtrangeo.2024.103852\u003c/li\u003e\n\u003cli\u003eAzhar, M., Kane, B., Vahedifard, F. et al. (2025) Comprehensive portfolio of adaptation measures to safeguard against evolving flood risks in a changing climate. Commun Earth Environ 6, 824. https://doi.org/10.1038/s43247-025-02779-z\u003c/li\u003e\n\u003cli\u003eBrocca, L. et al. (2016) Rainfall estimation by inverting SMOS soil moisture estimates: A comparison of different methods over Australia. JGR Atmospheres 121, 20: 12,062-12,079. https://doi.org/10.1002/2016JD025382\u003c/li\u003e\n\u003cli\u003eB\u0026uuml;rger A, Wnent J, Bohn A, Jantzen T, Brenner S, Lefering R, Seewald S, Gr\u0026auml;sner JT, Fischer M. (2018) The Effect of Ambulance Response Time on Survival Following Out-of-Hospital Cardiac Arrest. Dtsch Arztebl Int. 2018 Aug 20;115(33-34):541-548. doi: 10.3238/arztebl.2018.0541. PMID: 30189973; PMCID: PMC6156551.\u003c/li\u003e\n\u003cli\u003eColes, D., Yu, D., Wilby, R.L., Green, D. \u0026amp; Herring, Z. (2017) Beyond \u0026lsquo;flood hotspots\u0026rsquo;: Modelling emergency service accessibility during flooding in York, UK. Journal of Hydrology 546: 419-436. https://doi.org/10.1016/j.jhydrol.2016.12.013\u003c/li\u003e\n\u003cli\u003eDeo, R.C., Byun, HR., Adamowski, J.F. et al. (2015) A Real-time Flood Monitoring Index Based on Daily Effective Precipitation and its Application to Brisbane and Lockyer Valley Flood Events. Water Resour Manage 29, 4075\u0026ndash;4093. https://doi.org/10.1007/s11269-015-1046-3\u003c/li\u003e\n\u003cli\u003eEnglund, M. \u0026amp; Barquet, K. (2023) Threatification, riskification, or normal politics? A review of Swedish climate adaptation policy 2005\u0026ndash;2022. Climate Risk Management 40, 100492. https://doi.org/10.1016/j.crm.2023.100492\u003c/li\u003e\n\u003cli\u003eGreen, D., Yu, D., Pattison, I., Wilby, R., Bosher, L., Patel, R., Thompson, P., Trowell, K., Draycon, J., Halse, M., Yang, L., \u0026amp; Ryley, T. (2017) City-scale accessibility of emergency responders operating during flood events, Nat. Hazards Earth Syst. Sci., 17, 1\u0026ndash;16. https://doi.org/10.5194/nhess-17-1-2017.\u003c/li\u003e\n\u003cli\u003eGudiyangada Nachappa, T., \u0026amp; Meena, S. R. (2020) A novel per pixel and object-based ensemble approach for flood susceptibility mapping. Geomatics, Natural Hazards and Risk, 11(1), 2147\u0026ndash;2175. https://doi.org/10.1080/19475705.2020.1833990\u003c/li\u003e\n\u003cli\u003eHarries, M., Ushakova, A. (2025) Modelling emergency response times for Out-of-Hospital Cardiac Arrest (OHCA) patients in rural areas of the North of England using routinely collected data. BMC Emerg Med 25, 8. https://doi.org/10.1186/s12873-025-01170-7\u003c/li\u003e\n\u003cli\u003eHolm\u0026eacute;n, J. et al. (2020) Shortening Ambulance Response Time Increases Survival in Out‐of‐Hospital Cardiac Arrest. Journal of the American Heart Association 9, 21. https://doi.org/10.1161/JAHA.120.017048\u003c/li\u003e\n\u003cli\u003eJohnson, S. \u0026amp; Yu, D. (2020) From flooding to finance: NHS ambulance-assisted evacuations of care home residents in Norfolk and Suffolk, UK. Journal of Flood Risk Management 13, 1, e12592. https://doi.org/10.1111/jfr3.12592\u003c/li\u003e\n\u003cli\u003eLi, M. et al. (2021) Measuring emergency medical service (EMS) accessibility with the effect of city dynamics in a 100-year pluvial flood scenario. Cities 117, 103314. https://doi.org/10.1016/j.cities.2021.103314\u003c/li\u003e\n\u003cli\u003eLu, Y., Huang, Y. \u0026amp; Liu, X. (2025) A multi-source data-driven framework for probabilistic flood risk assessment using cascade machine learning models: case study in the Sichuan Basin. Sci Rep 15, 26706. https://doi.org/10.1038/s41598-025-12391-y\u003c/li\u003e\n\u003cli\u003eMathiesen, W.T., Bj\u0026oslash;rshol, C.A., Kval\u0026oslash;y, J.T. et al. (2018) Effects of modifiable prehospital factors on survival after out-of-hospital cardiac arrest in rural versus urban areas. Crit Care 22, 99. https://doi.org/10.1186/s13054-018-2017-x\u003c/li\u003e\n\u003cli\u003eMourad, K.A., Nordin, L. \u0026amp; Andersson-Sk\u0026ouml;ld, Y. (2022) Assessing flooding and possible adaptation measures using remote sensing data and hydrological modeling in Sweden. Climate Risk Management 38, 100464. https://doi.org/10.1016/j.crm.2022.100464\u003c/li\u003e\n\u003cli\u003eMuratoglu, A., Bilgen, G.K., Angin, I. \u0026amp; Kodal, S. (2023) Performance analyses of effective rainfall estimation methods for accurate quantification of agricultural water footprint. Water Research 238, 12001. https://doi.org/10.1016/j.watres.2023.120011\u003c/li\u003e\n\u003cli\u003eNicholl J, West J, Goodacre S, et al. (2007) The relationship between distance to hospital and patient mortality in emergencies: an observational study. Emergency Medicine Journal 24:665-668. https://doi.org/10.1136/emj.2007.047654\u003c/li\u003e\n\u003cli\u003eNoble, I.R., S. Huq, Y.A. Anokhin, J. Carmin, D. Goudou, F.P. Lansigan, B. Osman-Elasha, and A. Villamizar (2014): Adaptation needs and options. In: Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part A: Global and Sectoral Aspects. Contribution of Working Group II to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Field, C.B., V.R. Barros, D.J. Dokken, K.J. Mach, M.D. Mastrandrea, T.E. Bilir, M. Chatterjee, K.L. Ebi, Y.O. Estrada, R.C. Genova, B. Girma, E.S. Kissel, A.N. Levy, S. MacCracken, P.R. Mastrandrea, and L.L.White (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, pp. 833-868.\u003c/li\u003e\n\u003cli\u003eOddo PC and Bolten JD (2019) The Value of Near Real-Time Earth Observations for Improved Flood Disaster Response. Front. Environ. Sci. 7:127. doi: 10.3389/fenvs.2019.00127\u003c/li\u003e\n\u003cli\u003ePanahi, M. et al. (2023) A Country Wide Evaluation of Sweden\u0026apos;s Spatial Flood Modeling With Optimized Convolutional Neural Network Algorithms. Earth\u0026rsquo;s Future 11:11, e2023EF003749.\u003c/li\u003e\n\u003cli\u003ePant, R. et al. (2018) Critical infrastructure impact assessment due to flood exposure. Journal of Flood Risk Management 11, 1: 22-33. https://doi.org/10.1111/jfr3.12288\u003c/li\u003e\n\u003cli\u003ePapilloud, T. et al. (2020) Flood exposure analysis of road infrastructure \u0026ndash; Comparison of different methods at national level. International Journal of Disaster Risk Reduction 47, 101548. https://doi.org/10.1016/j.ijdrr.2020.101548\u003c/li\u003e\n\u003cli\u003ePell, J.P. et al. (2001) Effect of reducing ambulance response times on deaths from out of hospital cardiac arrest: cohort study. BMJ 322:1385. https://doi.org/10.1136/bmj.322.7299.1385\u003c/li\u003e\n\u003cli\u003ePregnolato, M. et al. (2017) The impact of flooding on road transport: A depth-disruption function. Transportation Research Part D: Transport and Environment 55: 67-81. https://doi.org/10.1016/j.trd.2017.06.020\u003c/li\u003e\n\u003cli\u003eShan, X., Scussolini, P., Wang, J. et al. (2023) Deficiency of Healthcare Accessibility of Elderly People Exposed to Future Extreme Coastal Floods: A Case Study of Shanghai, China. Int J Disaster Risk Sci 14, 840\u0026ndash;857. https://doi.org/10.1007/s13753-023-00513-x\u003c/li\u003e\n\u003cli\u003eShi, Y., Qian, Y., Jiahong, W., Xi, J., Li, H. \u0026amp; Wang, Q. (2022) A spatial accessibility assessment of urban tourist attractions emergency response in Shanghai. International Journal of Disaster Risk Reduction 74, 102919. https://doi.org/10.1016/j.ijdrr.2022.102919\u003c/li\u003e\n\u003cli\u003eShi, H., Zhou, M., Kong, N., Zhang, Y., \u0026amp; Li, X. (2023) A Study on the Accessibility of the Emergency Medical Services for Urban Kindergartens and Nursing Homes Based on Urban Pluvial Flooding Scenarios. Sustainability, 15(13), 10443. https://doi.org/10.3390/su151310443\u003c/li\u003e\n\u003cli\u003eShi, J., Wang, H., Zhou, J., \u0026amp; Zhang, S. (2024) Assessment and Improvement of Emergency Rescue Service Accessibility under Urban Waterlogging Disasters. Water, 16(5), 693. https://doi.org/10.3390/w16050693\u003c/li\u003e\n\u003cli\u003eSohn, J. (2006) Evaluating the significance of highway network links under the flood damage: An accessibility approach. Transportation Research Part A: Policy and Practice 40, 6: 491-506. https://doi.org/10.1016/j.tra.2005.08.006\u003c/li\u003e\n\u003cli\u003eSold\u0026agrave;, G., Molsted-Alvesson, H., Montalti, M., Reali, C., Gori, D., Von Schreeb, J. \u0026amp; Ljungman, P. (2025) Responder perspectives on preparedness for healthcare needs of vulnerable populations during floods and heatwaves: a qualitative study in Emilia-Romagna, Italy. BMJ Public Health. 3(2):e002459. doi: 10.1136/bmjph-2024-002459. PMID: 41099035; PMCID: PMC12519328.\u003c/li\u003e\n\u003cli\u003eSwedish Civil Contingencies Agency (MSB) (2023a) MSB Research Agenda 2024\u0026ndash;2028. Research for a safer society. Pub. no: MSB2228 \u0026ndash; October 2023. ISBN: 978-91-7927-414-6.\u003c/li\u003e\n\u003cli\u003eSwedish Civil Contingencies Agency (MSB) (2023b) Metod f\u0026ouml;r skyfallskartering av t\u0026auml;torter. Publikationsnummer: MSB2260 \u0026ndash; november 2023. ISBN: 978-91-7927-435-1.\u003c/li\u003e\n\u003cli\u003eSwedish Transport Administration (Trafikverket) (2018) Regeringsuppdrag om Trafikverkets klimatanpassningsarbete. Publikationsnummer: 2018:195. ISBN 978-91-7725-365-5.\u003c/li\u003e\n\u003cli\u003eTsang, M. \u0026amp; Scott, D.M. (2020) An integrated approach to modeling the impact of floods on emergency services: A case study of Calgary, Alberta. Journal of Transport Geography 86, 102774. https://doi.org/10.1016/j.jtrangeo.2020.102774\u003c/li\u003e\n\u003cli\u003eVerschuur, J. et al. (2024) Quantifying climate risks to infrastructure systems: A comparative review of developments across infrastructure sectors. PLOS Climate 3, e0000331. https://doi.org/10.1371/journal.pclm.0000331\u003c/li\u003e\n\u003cli\u003eVieira Passos, M., Kan, JC., Destouni, G. et al. (2024) Identifying regional hotspots of heatwaves, droughts, floods, and their co-occurrences. Stoch Environ Res Risk Assess 38, 3875\u0026ndash;3893. https://doi.org/10.1007/s00477-024-02783-3 \u003c/li\u003e\n\u003cli\u003eVieira Passos, M., Kan, JC., Destouni, G. et al. (2025a) Estimation of cascading hydroclimatic hazard impacts on supply systems and associated economic shocks. npj Nat. Hazards 2, 81. https://doi.org/10.1038/s44304-025-00129-9\u003c/li\u003e\n\u003cli\u003eVieira Passos, M., Barquet, K., Kan, J. C., Destouni, G., \u0026amp; Kalantari, Z. (2025b) Hydrometeorological resilience assessment of interconnected critical infrastructures. Sustainable and Resilient Infrastructure, 10(3), 267\u0026ndash;283. https://doi.org/10.1080/23789689.2024.2446124\u003c/li\u003e\n\u003cli\u003eXie, X., Huang, L., Marson, S.M. et al. (2023) Emergency response process for sudden rainstorm and flooding: scenario deduction and Bayesian network analysis using evidence theory and knowledge meta-theory. Nat Hazards 117, 3307\u0026ndash;3329. https://doi.org/10.1007/s11069-023-05988-x\u003c/li\u003e\n\u003cli\u003eYazdani, M. et al. (2022) A modelling framework to design an evacuation support system for healthcare infrastructures in response to major flood events. Progress in Disaster Science 13, 100218. https://doi.org/10.1016/j.pdisas.2022.100218\u003c/li\u003e\n\u003cli\u003eYin, J., Yu, D., Lin, N. \u0026amp; Wilby, R.L. (2017) Evaluating the cascading impacts of sea level rise and coastal flooding on emergency response spatial accessibility in Lower Manhattan, New York City. Journal of Hydrology 555: 648-658. https://doi.org/10.1016/j.jhydrol.2017.10.067\u003c/li\u003e\n\u003cli\u003eYu, D., Yin, J., Wilby, R.L. et al. (2020) Disruption of emergency response to vulnerable populations during floods. Nat Sustain 3, 728\u0026ndash;736. https://doi.org/10.1038/s41893-020-0516-7\u003c/li\u003e\n\u003cli\u003eZhang, Y., Zhou, M., Kong, N., Li, X., \u0026amp; Zhou, X. (2022a) Evaluation of Emergency Response Capacity of Urban Pluvial Flooding Public Service Based on Scenario Simulation. International Journal of Environmental Research and Public Health, 19(24), 16542. https://doi.org/10.3390/ijerph192416542\u003c/li\u003e\n\u003cli\u003eZhang, Y., Li, X., Kong, N., Zhou, M., \u0026amp; Zhou, X. (2022b) Spatial Accessibility Assessment of Emergency Response of Urban Public Services in the Context of Pluvial Flooding Scenarios: The Case of Jiaozuo Urban Area, China. Sustainability, 14(24), 16332. https://doi.org/10.3390/su142416332\u003c/li\u003e\n\u003cli\u003eZiya, O. \u0026amp; Safaie, A. (2023) Probabilistic modeling framework for flood risk assessment: A case study of Poldokhtar city. Journal of Hydrology: Regional Studies 47, 101393. https://doi.org/10.1016/j.ejrh.2023.101393\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":true,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"climatic-change","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"clim","sideBox":"Learn more about [Climatic Change](https://www.springer.com/journal/10584)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/clim/default.aspx","title":"Climatic Change","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Flood risk management, critical infrastructure, emergency medical services (EMS), climate-resilient health systems, GIS analysis, climate change and health","lastPublishedDoi":"10.21203/rs.3.rs-8600985/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8600985/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Flood occurrence and intensity changes may increasingly threaten the accessibility of emergency medical services. To investigate this threat, we develop a data-driven framework for assessing flood effects on time-critical ambulance routes, using suspected cardiac arrests Sweden as a test case. We integrate two key datasets: over 117,000 ambulance dispatch records, and seven years of high-resolution modelled hydrological risk data. First, we establish a strategic health access network through algorithmic routing of the empirical dispatch locations and intersect these routes with historical flood risk to derive their long-term exposure profile. We find that the most vulnerable emergency corridors face up to 17% annualised flood risk frequency, or approximately 62 flooded days per year. Second, we perform a risk-centric performance screening to identify specific operational degradation, isolating ambulance routes where real-time flood risk coincided with measurable response delays that exceeded local performance baselines. This screening localises critical operational delay hotspots in emergency services and quantifies delays for flood-exposed routes, with an average delay rate of 82 seconds/kilometre over the municipal median. These delayed ambulance route segments occurred most frequently within urban centres in west and central Sweden (e.g. the cities of Gothenburg and Karlstad) as well as in smaller, geographically diverse and rural areas, such as northern Sweden. Overall, the approach developed and tested here shifts risk assessment from static hazard mapping to dynamic service discontinuity, offering a general tool for prioritising operational planning measures and infrastructure investment through a public health lens to enhance emergency medical response capacity and resilience.","manuscriptTitle":"Assessing and screening flood exposure risks for emergency response","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-03-11 14:23:00","doi":"10.21203/rs.3.rs-8600985/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"reviewerAgreed","content":"","date":"2026-03-11T12:37:42+00:00","index":0,"fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-03-06T07:16:03+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-01-17T07:51:40+00:00","index":"","fulltext":""},{"type":"submitted","content":"Climatic Change","date":"2026-01-16T05:40:45+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"climatic-change","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"clim","sideBox":"Learn more about [Climatic Change](https://www.springer.com/journal/10584)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/clim/default.aspx","title":"Climatic Change","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"091e0b57-c1d6-45f7-9e3f-2014bc340dc9","owner":[],"postedDate":"March 11th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-03-12T21:30:35+00:00","versionOfRecord":[],"versionCreatedAt":"2026-03-11 14:23:00","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8600985","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8600985","identity":"rs-8600985","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}