Adaptation pathways to address climate change-induced flood risk in the Zemgale region, Latvia

A comprehensive understanding of changing climate patterns and flood hazard dynamics is essential for effective climate change adaptation. One of the European Union's Horizon 2020 IMPETUS project demonstrator regions, Zemgale (Latvia), located in the Boreal biogeographic region of Northern Europe, is particularly vulnerable to flooding due to its flat agricultural landscapes, dense river network, and extensive floodplains. Under climate change, spring river flood regimes are being altered by earlier snow and ice melt, as well as by increasing heavy rainfall intensity, which heightens flash flood risks in urban areas. This study aimed to develop climate change adaptation pathways to improve regional preparedness for both river floods and heavy rainfall-induced flash floods through a co-design and co-creation process involving stakeholders from multiple governance levels, with the support of a risk reduction modeling team to establish alternatives for assessment. Adaptation options were identified and evaluated using multi-criteria analysis, considering, among others, effectiveness, costs, co-benefits, feasibility, and additional criteria. A set of structural and non-structural measures, including conventional approaches and innovative nature-based solutions, were selected to address rapid increases in river water levels and intense rainfall events. These measures were subsequently sequenced into adaptation pathways to respond to flood-related hazards and impacts, enhancing adaptive capacity and reducing vulnerability in both rural and urban areas of the Zemgale region under future climate conditions. To identify and evaluate alternative adaptation pathways, the Dynamic Contingency Response Tool (DCRT) was used. Modeling was conducted for three climate change and development scenarios (SSP1, SSP2, and SSP5), utilizing risk projections derived from the Risk Emerging and Problems Inference Engine (REPIE) tool. The outcomes of the risk analysis were discussed with stakeholders and considered in the development of the Zemgale Regional Climate Change Adaptation Plan.

Climate change is widely regarded as one of the most topical scientific and political issues of the 21st century as evidenced by the extensive body of research, the growing focus on climate policy and governance. As stated by the Intergovernmental Panel on Climate Change (Calvin et al., 2023a): “ The cumulative scientific evidence is unequivocal: climate change is a threat to human wellbeing and planetary health. Any further delay in concerted anticipatory global action on adaptation and mitigation will miss a brief and rapidly closing window of opportunity to secure a liveable and sustainable future for all ”. The IPCC, which serves as a bridge between research and policymaking, presents and explains climate change implications and potential future risks, as well as puts forward adaptation and mitigation options (Calvin et al., 2023b). While the IPCC calls for global actions, it is inevitable that actions shall be taken at multiple levels: regional, national, subnational and local (Cradock-Henry and Frame, 2021).

The five Shared Socio-economic Pathways (SSP1 to SSP5) provide narratives that describe alternative scenarios (plausible futures) of socio-economic development which, in turn, can be associated with different trajectories of greenhouse gas emissions. In climate change studies, the SSPs are used globally as a common reference for evaluating future climate impacts, exposure, vulnerability and challenges to adaptation (Van Vuuren et al., 2017; Calvin et al., 2023b). The SSPs are then further explored and downscaled to regional like European Union (Kok et al., 2019; Zandersen et al., 2019), national (Chen et al., 2020), subnational (Pedde et al., 2025) and local (Frame et al., 2018) levels, for mitigation and adaptation planning and decision-making. SSPs are also used as input to develop future scenarios of chemical emissions (Nagesh et al., 2022), or locally extended to cover specific climate challenges like urban heat issues (Rohat et al., 2019). Flooding risk in relation to future scenarios (SSPs) is among the topics that have been explored to estimate impacts on exposure to water resources stress, changes in river flood frequency (Arnell and Lloyd-Hughes, 2014), impacts on population (Porter et al., 2024) and other vulnerability aspects (Vu et al., 2025). The novelty of applying downscaled SSPs in climate risk modeling stems from the ability to couple global climate-socioeconomic scenarios with scale specific analysis, thereby improving the precision, contextual relevance, and policy applicability of future climate risk projections.

The policy relevance of climate risk modeling lies in its ability to inform evidence-based adaptation strategies and long-term resilience planning under conditions of climate-related uncertainty. Floods caused either by intense precipitation or rising water levels resulting from snow and ice melt are among the climate-related hazards of significant policy and societal relevance in the Northern Hemisphere. In the European Union (EU), flood risk management is regulated by the Floods Directive 2007/60/EC (Directive 2007/60/EC of 23 October 2007 on the Assessment and Management of Flood Risks, 2007), which recognizes that floods are natural phenomena; however, various human activities and climate change contribute to an increase in both the likelihood of flood events and their adverse impacts. The likely impacts of climate change on flood occurrence are required to be incorporated into flood risk assessments and flood risk management plans developed and implemented by EU Member States (Priest et al., 2016; Holguin et al., 2021), as well as into comprehensive climate adaptation strategies and associated action plans.

Over time, a wide range of approaches and methods for developing adaptation planning documents containing actions, measures, and interventions have emerged. Nevertheless, a major challenge that remains is the lack of sufficiently strong commitments and the difficulty of translating adaptation strategies into tangible and urgent actions aimed at protecting communities and economies. This challenge was addressed by the EU Horizon 2020 project IMPETUS, which brought together multidisciplinary research teams, policymakers, businesses, and local and regional communities to ensure that knowledge is co-created and shared collaboratively. Adaptation pathways have been recognized as a well-suited planning approach to support decision-making for future climate risk reduction and have been extensively reviewed and discussed in the literature (Cradock-Henry and Frame, 2021; Werners et al., 2021; Haasnoot et al., 2024). The innovation of the IMPETUS project lay in demonstrating, through a case study approach, how adaptation pathways can facilitate improved governance and economic decision-making, while also achieving cost efficiency through the implementation of successful solutions across different spatial scales and thematic areas.

This paper presents the application of adaptation pathways at the regional level through a case study conducted in the Zemgale Planning Region in Latvia, Northern Europe, with the aim of supporting the regional authority in the development of its first climate adaptation plan. The case study also demonstrates a climate adaptation planning framework that enables a shift from ad hoc activities toward a systemic implementation approach for addressing flood risk-related challenges while considering cross-sectoral dimensions such as agriculture, forestry, infrastructure, public health, and biodiversity in the context of long-term climate change scenarios. The key research questions addressed in this study are:

What are the opportunities and limitations of applying climate adaptation pathways to regional flood risk management?

How can the quantification of benefits for climate adaptation pathway support development for regional flood risk management?

Case study area

Zemgale region (central part of Latvia), located in the boreal region in Northern Europe (Figure 1), covers an area of 10,742 km 2 (16.6% of Latvia). The Zemgale region constitutes a lowland landscape characterized by flat terrain, a dense river network and floodplains. The Zemgale Region is traversed by two major rivers, the Daugava and the Lielupe. Available water resources are sufficient to meet the needs of economic activities and human consumption. The region's drinking water supply is based on groundwater resources. The Zemgale region comprises six municipalities: Jelgava state city municipality and five county municipalities: Dobele, Jelgava, Bauska, Aizkraukle and Jekabpils. Urban areas account for 6% of the territory of the Zemgale region (Central Statistical Bureau, 2026c). The total population was approximately 222,000 inhabitants in 2025 (Central Statistical Bureau, 2026b). Due to its fertile soils, the area has developed into an intensive agricultural region (43% of the territory), dominated by arable land used for grain and rapeseed production, with smaller shares of meadows and pastures. Forests cover about 40% of the territory of the Zemgale Region (Central Statistical Bureau, 2026a).

The region is characterized by a high flood hazard, posing risks to society, settlements, and infrastructure. Spring river floods, driven by snow and ice melt, affect the entire region by increasing water levels in rivers and adjacent floodplains. In addition, the increasing intensity of heavy rainfall presents a greater threat to urban areas due to higher population density, limited urban drainage capacity, and extensive built-up and impervious surfaces that reduce natural infiltration and drainage.

Methodological framework

This study adopts a structured and iterative methodological framework to systematically analyze climate impact chains and develop and evaluate adaptation pathways (Figure 2). The adaptation pathways concept provides an approach for exploring and sequencing portfolios of adaptation actions designed to address future adaptation needs. Adaptation pathways can support planning under changing conditions by “leaving options open to further adapt depending on how the future unfolds” (Di Fant et al., 2025). They are decision-focused tools that can take the form of a decision tree or a roadmap, allowing measures to be implemented progressively as future conditions and knowledge evolve (Werners et al., 2021). Moreover, adaptation pathways are intended to be integrated into the broader adaptation planning process, thereby supporting dynamic adaptation policies (Haasnoot et al., 2013).

Central to adaptation pathways are adaptation tipping points, defined as the conditions under which an action no longer meets clearly specified objectives. Tipping points, identified when system performance drops below a defined level, mark the need to change course along the decision tree (or roadmap) (Haasnoot et al., 2013). The timing of an adaptation tipping point is scenario-dependent, since it depends on climate change trajectories. It also depends on how socio-economic conditions evolve over time, particularly in relation to exposure and vulnerability. Decision points should be set before the critical threshold is reached (Coulter, 2019), to ensure that there is sufficient lead time for the implementation of the selected action. In the present work, we approached the topic considering the occurrence of specific conditions or events (temperature, extreme events, ice cover) as alarm thresholds contributing to a potential impact risk (variables of hazard, vulnerability and exposure type). Hereafter, we do not use tipping points to trigger measures at a specific discrete condition; rather, we allow the optimization process to determine the timing of interventions based on (a) continuous risk-component indicators and their interactions with predefined alarm thresholds, (b) intervention characteristics, and (c) stakeholder criteria. Adaptation pathways culminate with different visuals (Haasnoot et al., 2024), reflecting different degrees of complexity in the decision-making context, and different degrees of uncertainty about how climate change will affect socio-ecological systems and about the effectiveness and lifetime of certain adaptation interventions.

Adaptation pathways provide a means of progressing toward so-called “transformational adaptation”, which implies deep and fundamental changes in socio-ecological systems, involving the transformation of norms, visions, values, rules, and practices (Colloff et al., 2021). In fact, the adaptation pathways approach requires transforming the way decisions are made and actions are implemented under uncertain conditions (Colloff et al., 2021; Bosomworth and Gaillard, 2019; Di Fant et al., 2025). Moreover, adaptation pathways can be used to inform the theories of change, i.e., “methods that explain how a given intervention, or set of interventions, are expected to lead to a specific development change, drawing on a causal analysis based on available evidence” (United Nations Development Group, 2017).

The process of creating adaptation pathways is generally operationalized in several steps to support planners. Planners can first quickly run through the steps to get an idea of the nature of the adaptation needs and solutions, and then repeat the exercise with growing detail and considering different objectives. Ranger et al. (2013) described the approach when deriving lessons learned from the Thames Estuary 2100 project.

Our approach integrates analytical and modeling assessments with participatory processes. The framework comprises six sequential steps (Figure 2). The first step focuses on understanding the climate impact chain. This involves elucidating the causal relationships between climatic hazards, exposure, vulnerability, and the resulting impacts and risks. The objective is to establish a comprehensive representation of the system under study and to identify key factors to be considered when developing adaptation options. In the second step, adaptation options are identified. Based on the defined impact chains, a set of potential adaptation measures is compiled. This process is informed by existing knowledge, expert input, and stakeholder contributions, ensuring that the proposed options appropriately address flood risk management and are suitable for the Zemgale region. The third step entails assessing the proposed adaptation options using a multi-criteria approach (see also section 2.3). A systematic evaluation is conducted using predefined criteria, such as effectiveness, feasibility, cost-efficiency, and co-benefits. Stakeholder consultation plays a critical role at this stage, providing local knowledge when assigning scores in the assessment process. The fourth step involves identifying a preliminary implementation sequence, referred to as an adaptation pathway. In this step, adaptation options are organized into a logical sequence over time and according to potential thresholds under changing climatic conditions. This work is carried out considering stakeholder perceptions and acceptance. In the fifth step, the efficiency of adaptation options is modeled. This step employs analytical tools—REPIE and DCRT—to quantitatively evaluate the performance of individual options and their combinations under different scenarios. The sixth step consists of building and selecting alternative adaptation pathways and assessing them. Multiple pathways are finally proposed to reflect uncertainties in climatic and socio-economic developments and to trigger policy discussion about different strategies available to address the increasing flood risk in the region. These pathways are comparatively assessed in terms of flood risk reduction, an important aspect of climate adaptation planning.

A central feature of the framework is the continuous and iterative engagement of stakeholders throughout all stages of the process, which is further described in the next section. The research team is responsible for conducting analytical procedures, integrating stakeholder inputs, and ensuring methodological consistency. Iterative feedback loops between stages allow for refinement and adjustment of the analysis as new insights emerge.

Co-design and co-creation process

The adaptation pathways were developed through a co-design and co-creation process involving stakeholders from multiple levels of governance, including local communities and all six municipal authorities in the Zemgale region, regional institutions such as the Energy Agency and the Administration of the Zemgale Planning Region, farmers, environmental NGOs, and national institutions responsible for water and flood risk management in Latvia. Participants were invited based on their knowledge, experience, and role in climate adaptation planning and flood risk management in the region and the country. These stakeholders participated in a series of workshops conducted between 2023 and 2025 (see Supplementary Material 1 for details). The iterative nature of the engagement process, involving multiple consultation stages, helped to validate and refine the adaptation pathways through regular stakeholder feedback. The stakeholder assessment of the adaptation option was used to support the identification and evaluation of adaptation pathways rather than to provide statistically generalizable estimates.

The stakeholder engagement process for developing the adaptation pathways was divided into three major stages: preparation, development of adaptation options, and modeling (Figure 3). The first stage focused on identifying climate change risks and impacts, priority policy areas, and ongoing or planned climate change adaptation actions in the Zemgale region. This stage was conducted in parallel with, and in close synergy with the development of the Zemgale Regional Climate Change Adaptation Plan (adopted in March 2026). While the adaptation plan had a broader scope, addressing a wider range of climate-related impacts associated with temperature, wind, and precipitation, the present study focused specifically on flood risk. The second stage aimed to develop adaptation pathways for reducing flood risk through the identification and refinement of adaptation options. A range of facilitation methods was employed to encourage interactive stakeholder participation and maintain participants' engagement throughout the process. Moderated working group discussions and World Café sessions were organized during stakeholder workshops to identify, discuss, and refine potential adaptation options. In the third stage, the focus shifted to performance-based pathway development through the application of the PATH framework (REPIE and DCRT models) to quantitatively assess the effectiveness of adaptation pathways and individual measures for flood risk reduction. The multidisciplinary research team, comprising IMPETUS project partners, including the case study implementation partner (Baltic Environmental Forum-Latvia) and research partners specializing in climate adaptation and modeling (Thetis SpA and the National Technical University of Athens), led this stage. Stakeholders were primarily consulted to provide feedback on the modeling results and the proposed pathways.

Climate risks' conceptual model

The cause-effects relationships among different variables and processes of the system are systematized in a conceptual model. This model links key components of risk [Intergovernmental Panel On Climate Change (IPCC), 2023]: hazard, vulnerability, exposure, impacts. Climate risks arise from the interactions of all these elements.

The model schema is derived using the impact chain approach (GIZ and EURAC, 2017; Zebisch et al., 2022, 2023), which was developed to support the creation of a shared understanding of the elements and processes contributing to climate risks. The approach facilitates stakeholder engagement and participatory processes, enables the integration of local data and knowledge, and provides an entry point for discussing potential adaptation options.

To build the impact chain, background information was collected on past trends and future projections of climate variables such as precipitation (including extreme events) and temperature, as well as the likely dynamics of socioeconomic factors (e.g., urban expansion, population dynamics etc.). Climate data from the Latvian Environment, Geology and Meteorology Center (Latvian Environment Geology Meteorology Centre, 2025) were used to substantiate the conceptual model and served as input to the REPIE tool (Section 2.6).

Assessment of adaptation options by criteria

Adaptation options were assessed through a combination of quantitative and qualitative criteria and indicators, according to data and model availability. Eight criteria were applied:

Effectiveness: Relevance of the measure for the achievement of the objective selected in Step 1 and the associated indicator(s). A measure is highly relevant if it is expected to significantly contribute to the achievement of the objective, while it has low relevance if it can marginally contribute to the selected objective. If possible, this evaluation should also consider the systemic implementation of the intervention at the regional scale, not only the individual implementation of a pilot or small-scale intervention. Effectiveness may be referred to the contribution in the reduction of risk acting on vulnerabilities or exposure factors.

Lifetime/ sell-by date/sell-by condition: How long the intervention is expected to last and be effective (lifetime) and/or critical conditions that make the action no longer effective. Lifetime (also often referred as sell-by-date, use-by date) can depend on the infrastructure physical features and/or on the changing environmental conditions (sell-by condition) that affect its durability.

Feasibility - TRL Technology readiness level. TRL expresses the progression of science innovation from the original idea to its commercial usability.

Feasibility-SRL Social readiness level. SRL addresses the readiness of society to adopt and use an innovation.

Economic Costs. Total monetary costs needed to implement the measure. Unit costs (e.g., cost per square meter) are also relevant in assessing the scaling potential of the intervention. Capital (CAPEX) and operational (OPEX) are considered.

Co-benefits. Beyond the main adaptation result, co-benefits include environmental benefits (e.g., biodiversity enhancement, better water quality), societal benefits (e.g., human wellbeing, recreational opportunities) and economic benefits (e.g., increased revenues, reduced operational costs).

Unintended effects/hidden risks. Hidden risks refer to possible negative impacts associated to the implementation of the intervention. They also refer to the environment (e.g., habitat disruption), society (e.g., gentrification) and economy (e.g., increased maintenance costs). A careful assessment of cascading effects help avoiding maladaptation.

Implementation time. Time needed to implement the action. This information is needed to sequence various adaptation interventions in the pathways and to set decision points before the critical threshold is reached (to leave enough time for the implementation).

Modeling of efficiency of adaptation options with REPIE and DCRT tools

In this study, we used two tools developed within the IMPETUS project, namely the Risk and Emerging Problems Inference Engine (REPIE) and the Dynamic Contingency Response Tool (DCRT) (Nikolopoulos et al., 2026). REPIE operationalizes the steps (i) to (iii) of the methodological framework (Figure 2), while DCRT operationalizes steps (v) and (vi). The modeling workflow is illustrated in Figure 4.

The tools operate under the assumption that the climate change risk results from complex interactions between three components, i.e., climate-related hazard, vulnerability and exposure (HVE), following the conceptual model described by the impact chain. REPIE uses as inputs future projections of climatic and socio-economic variables that form the components of risk, as identified through desk studies, stakeholder engagement and expert knowledge of the users (Supplementary Material 4). In REPIE, each variable belongs to the hazard, exposure, or vulnerability category. The user sets four different “alarm thresholds” to classify different levels of threat, ranging from the least concern to critical values (LC: least concern, MT: medium threat, HT: high threat, CT: critical threat). The thresholds vary depending on the specific variable context (and can be estimated from historical records) and data availability, making REPIE flexible for application in diverse context settings and regions. The user may also assign different weights to the variables, according to the different perceived or estimated relevance in determining the risk. The HVE components are expressed as rescaled weighted estimations of the associated variables. Risk for each timestep is estimated as the geometric mean of the three weighted HVE components, as it scales better with small values of components than regular simple multiplicative relations. A risk curve is produced for each scenario formulated using the SSPs narratives and other projections. As such, risk curves express the behavior of the system under a future scenario in a stress-testing manner, like other established resilience assessment methodologies (Makropoulos et al., 2018; Nikolopoulos et al., 2019).

The output of REPIE is used as input for DCRT (Supplementary Material 5), which identifies the optimal adaptation pathway, given the emerging climate risk conditions, and a selection of suitable measures and interventions that stem from stakeholder engagement, desk studies, already developed masterplans, and expert judgement from users. The tool considers the multiple characteristics of the measures and interventions presented in Section 2.5 (assessment criteria) and requires as input also considerations concerning eventual mutual exclusivity or precedent/antecedent conditions for the sequencing of interventions, to account for relationships between them. In a simulation run, DCRT and REPIE operate in unison, with feedback loops, with DCRT generating a sequence (pathway) of interventions and passing their effects on the REPIE tool to estimate the reduced risk curve, while DCRT calculates costs, feasibility and externalities. DCRT applies the NSGA-II algorithm, which is suitable for multi-objective Pareto optimization (Deb et al., 2002). Users can further filter the resulting pathways using a weighted approach or the kneedle algorithm for the best compromise solution.

Understanding climate impact chain in the Zemgale region

The impact chain model (Zebisch et al., 2023) was used to assess the risk of loss and damage associated with floods in the Zemgale region. The climate impact chain for the region was developed, applying a participatory approach involving stakeholders and experts at national, regional, and local level. During stakeholder workshops, regional and local data and knowledge, as well as previous experiences related to flood risk assessment—hazards, exposure, vulnerability and damage—were collected. The cause-effect chain represents main factors and processes leading to flood-related loss and damage in agriculture, infrastructure, and settlements (Figure 5). Climate data from the Latvian Environment, Geology and Meteorology Center (Latvian Environment Geology Meteorology Centre, 2025) were used to inform the discussions and explore interlinkages among key variables (Table 1) and processes.

Several hazards were identified to characterize the impacts of climate change in the Zemgale region, particularly within the Lielupe River Basin and its largest urban center, Jelgava (Table 1). Flood risks in the Zemgale region are primarily associated with river overflow, which is most commonly triggered by ice jams and the melting of snow and ice. These processes are often induced by rapid temperature increases during the winter and spring seasons (Klavins et al., 2009; Apsite et al., 2024). Consequently, average seasonal temperature was identified as a key climate indicator for assessing flood risk.

Under the climatic baseline conditions typical of the Zemgale region, average winter temperatures (January–February) remain below 0 °C, indicating that watercourses are generally ice-covered and precipitation predominantly occurs as snow. During the transition to spring (March–April), rising temperatures promote snow and ice melt, which can lead to the formation of drift ice and increase the likelihood of ice jam events.

During the summer season, the region is exposed to several hazards, primarily associated with intense precipitation events. Heavy rainfall can lead to pluvial (flash) flooding by overwhelming drainage systems in both urban and rural areas. When exceptionally high precipitation occurs within a 24-h period, significant increases in water levels in the Lielupe River have been observed at monitoring stations. Such events may become increasingly significant if precipitation intensity continues to rise, particularly when urban flooding coincides with elevated river discharge.

For the Zemgale region, four principal exposure indicators have been identified, each closely linked to corresponding vulnerability indicators (Table 2). Flooding poses significant risks to agricultural production, including crop losses and yield reductions, as well as to public and private infrastructure, particularly in flood-prone areas adjacent to rivers. Flood impacts can also reduce accessibility by damaging or inundating road networks, thereby limiting mobility and disrupting essential services, including electricity supply and public and private transport. In addition, flood events place increased pressure on emergency response and social care systems.

The severity of flood-related impacts is strongly influenced by the effectiveness of drainage infrastructure, surface permeability, and the capacity for water retention within the landscape. Outdated or poorly maintained drainage systems, the prevalence of impervious surfaces, and the insufficient implementation of water retention measures—including nature-based solutions such as constructed wetlands-contribute to increased vulnerability. In the agricultural sector, extensive monoculture practices (predominantly grain production) and the limited uptake of flood risk insurance among farmers and landowners further constrain adaptive capacity. Moreover, low public awareness of flood risks and limited knowledge of appropriate preventive actions, combined with insufficiently developed regional early warning systems, reduce overall preparedness across sectors in the Zemgale region.

Adaptation options

Based on the identified climate impact chains, a set of potential adaptation options was identified and developed through an iterative process. These options (Table 3) were classified according to the type of measure, distinguishing between non-structural (soft) and structural (investment-intensive) measures. The latter were further grouped based on their primary area of implementation, namely urban or rural contexts. The selection process was informed by experience from recent regional and local infrastructure and capacity building projects that generated relevant knowledge and fostered stakeholder acceptance of specific options, including within the research team. Consequently, several initially proposed measures were not pursued further within the scope of this study. At the same time, several proposed options were innovative and relatively ambitious, considering the specific characteristics of the Zemgale region. Establishing a clear and consistent description of each adaptation option was a critical step prior to undertaking the assessment process (Supplementary Material 2).

The assessment of the 11 selected adaptation options, based on the eight criteria and corresponding definitions described above (section 2.5), was conducted during Stakeholder Workshop 6 (Supplementary Material 1). Each workshop participant received a printed evaluation template containing all 11 adaptation options together with the respective assessment criteria. In total, 18 stakeholders anonymously completed the evaluation forms. The participants carried out a semi-quantitative assessment using a three-point scale, where 1 indicated a low score, 2 a medium score, and 3 a high score. The respondents were invited to provide comments reflecting their experience and opinions regarding the proposed adaptation options (Supplementary Material 3).

In addition to the stakeholder-based assessment, the research team collected supplementary data and information to support the semi-quantitative evaluation with quantitative evidence and further justification for the final selection of adaptation options. Average scores were calculated to derive the overall semi-quantitative assessment (Table 4). The following threshold values were applied for interpretation: scores below 1.75 were classified as +/-; scores between 1.75 and 2.5 as –/++; and scores above 2.5 as —/+++.

Stakeholders considered non-structural measures to be moderately efficient, but they raised concerns about Technology Readiness Level (TRL) and Social Readiness Level (SRL), which may constrain the acceptance and uptake of the innovation associated with these measures. The measure involving development and building restrictions in flood-prone areas, included in approximately half of the pathways generated by the tools, was questioned by stakeholders. Stakeholders highlighted that flood risk maps should better reflect observed conditions rather than relying on modeled outputs and suggested that providing information on safe and appropriate construction practices may be preferable to prohibiting development entirely.

Costs of non-structural measures were generally assessed as lower compared with structural measures, except for flood risk early warning systems, where implementation costs were considered comparable to those of structural interventions. Despite these costs, stakeholders considered optimization of early warning systems highly relevant and necessary. At the same time, concerns were raised regarding personal data protection, and the need for stronger involvement of local governments in developing and improving such systems was emphasized.

Regarding urban structural measures, stakeholders rated improvements in the capacity of open ditches and culverts and increased rainwater collection capacity higher than permeable surfaces and NBS for urban water retention. Ensuring the functionality of open ditches and culverts, which was included in most pathways generated by the tools, was considered a relatively simple solution; however, its applicability is constrained by the density of urban development. Therefore, this option may be more feasible and effective along main streets rather than in densely built-up city centers.

By highlighting the need to increase stormwater retention capacity, stakeholders also identified potential implementation challenges related to alignment with municipal spatial plans and high investment costs. Stakeholders were skeptical about the use of permeable pavements and urban NBS, despite their frequent inclusion in pathways generated by the tools. Concerns were mainly related to maintenance requirements and limited available space in historically developed urban areas, suggesting that these measures may be more suitable for smaller towns with less dense urban structures. In rural areas, increasing the functionality of drainage systems, which was one of the most frequently selected measures in the pathways, was highly appreciated by stakeholders, although concerns were raised regarding costs and implementation time. NBS for water retention and riverbed cleaning were generally perceived as routine small-scale activities, which may explain their lower frequency of inclusion in the pathways generated by the tools.

The overall results indicated that adaptation options that are well established and have been implemented in practice over a longer period were generally assessed as the most effective across most criteria. Measures such as open ditches in urban settlements and drainage and land reclamation systems have been implemented for decades and are therefore well perceived by local stakeholders as flood risk management measures. Cleaning of riverbeds (dredging and sediment removal) has also been implemented for more than fifty years in Latvia, with more recent measures supported by EU funding schemes aimed at reducing flood risks. In contrast, the option to restrict development in flood-risk zones, particularly in floodplains, has the lowest scores in terms of technical and social readiness. This is attributed to stakeholder resistance and limited political willingness to implement restrictive regulatory instruments.

Alternative pathways and their assessment

Set of alternative pathways

Various sequence of measures (interventions) implemented at different time-steps can be applicable in numerous combinations to reduce flood risk. Applying the PATH framework (REPIE + DCRT models) a set of alternative adaptation pathways (solutions) was developed to reduce the flood-related risks in the Zemgale region and support local and regional decision makers in selecting appropriate strategies. The modeling analysis was performed for three climate scenarios: SSP1-RCP 2.6, SSP2-RCP 4.5 and SSP5-RCP 8.5, using projected risk trajectories retrieved by the Euro-Mediterranean Center on Climate Change Earth System Model (CMCC-ESM). As shown in Figure 6, a comparison of risk trends under the SSP1, SSP2, and SSP5 scenarios for the Zemgale region indicates relatively limited changes in risk under SSP1 (0.0036 risk index units per year) and SSP2 (0.0101 risk index units per year), while the most pronounced increase occurs under SSP5 (0.0181 risk index units per year). The projected increase in risk of approximately 6% compared with the base year (2021) under SSP1 may indicate that existing adaptation measures could be sufficient to manage climate-related risks under this scenario. In contrast, the projected increases of approximately 17% under SSP2 and 31% under SSP5 indicate substantially higher levels of risk, potentially requiring stronger adaptation interventions. As SSP5 represents the scenario with the most rapid increase in potential flood-related socioeconomic risk values (Buskop et al., 2025), the analysis in this paper focuses on adaptation pathways under SSP5 conditions.

In total, 37 Pareto-optimal alternative adaptation pathways, representing different combinations and sequences of measures (interventions) were identified for climate adaptation and flood risk reduction. These pathways represent non-dominated alternative solutions that satisfy, to varying degrees, the different optimization criteria (risk reduction, costs, feasibility, etc.) considered by the algorithm. Figure 7 presents the starting year of 11 different non-structural, urban structural and rural structural measures across the pathways. Of the 37 alternative adaptation pathways, 24 (65%) comprise all three types of measures, 11 (30%) combine two types—non-structural and urban structural measures—and 2 (5%) consist exclusively of urban structural measures.

Some adaptation measures, such as permeable surfaces, improving rainwater collection capacity, improving drainage functionality, and cleaning riverbeds, can be implemented in two phases due to the lifetime and effectiveness of these options or the scale of their implementation. Thus, an intervention refers to a single implementation phase of an adaptation measure, and the number of interventions is used to describe and assess individual adaptation pathways. Accordingly, the number of interventions ranges from 13 in P_1 to 2 in P_7, P_11, P_13, P_19, P_26, P_27, and P_34.

The implementation of adaptation measures is distributed unevenly over the period to 2050, with implementation activity peaking in 2029 and 2030. This indicates that, on average across the assessed pathways, proactive interventions to enhance system resilience should be initiated as early as possible, provided that sufficient funding and budget allocations are available.

Modeled risk by REPIE + DCRT

Performance-oriented adaptation pathways for quantifiable flood risk reduction were developed by combining the qualitative selection of measures with targeted quantitative analyses. Flood risk in the Zemgale region under the SSP5 scenario was modeled for the period 2021–2100 using available input data and a range of flood risk reduction measures (Figure 8). The analysis compared a baseline scenario, in which no additional measures were implemented, with multiple combinations of 1 to 11 measures. As some measures are implemented in phases, the total number of interventions considered exceeds the number of individual measures.

At the beginning of the modeling period (2021), the flood risk index under the baseline scenario (by a linear fit) was estimated at a moderate level of approximately 4.5 on a scale from 1 to 10. Under this scenario, the risk is projected to increase to 6.2 by 2100 (Figure 6). It should be noted that this scale represents a normative expression tailored to the specific problem formulation and the selected interventions to address the identified challenges, rather than a calculation based on hydrological/hydraulic numerical evaluations of risk (e.g., flood depth/velocity/extent maps). Flood risk reduction can be achieved through different combinations of non-structural measures, as well as structural measures implemented in urban and rural areas. Different levels of flood risk reduction are achieved depending on the strategy applied, including combinations of 2, 5, or 13 risk-reduction interventions, as presented in Figure 8. It is important to consider that each measure requires an implementation period (a build-up period of 1–5 years) before reaching its full effectiveness and remains effective for a defined lifetime. The comparison clearly indicates that a greater number of implemented measures leads to more pronounced benefits, reducing flood risk to approximately 2 by the end of the modeling period.

Taking into consideration the increase in climate-related risks under the SSP2 scenario, a set of alternative adaptation pathways for the Zemgale region was modeled. Similar risk-reduction patterns were observed under the SSP2 and SSP5 scenarios, although differences between the two scenarios are evident in the effectiveness of alternative adaptation pathways in reducing risk (Figure 9). Comparing the risk index in 2050 with its initial value in 2021, it is evident that under SSP2, all pathways except for two, each comprising a single intervention, result in a risk index below the initial value. This indicates that the implementation of measures included in these adaptation pathways may be sufficient to mitigate climate-related risks. Moreover, the implementation of more extensive adaptation pathways, comprising 10–13 interventions, can maintain the risk index below half of its initial value. Under SSP5, 11 pathways, comprising 2–3 interventions each, fail to maintain the risk index at or below its initial 2021 value, while only the most extensive adaptation pathways, comprising 12 and 13 interventions, are sufficient to reduce the risk index to below half of its initial value by 2050. This indicates that, under a more severe climate change scenario, regional and local authorities may need to plan for a greater number of adaptation measures (interventions) to effectively mitigate climate-related flood risks.

Assessment of alternative pathways

The example in Figure 8 illustrates that only limited reduction in risk can be achieved through the implementation of a small number of measures (P_27, purple line). In contrast, five selected interventions (P_16, blue line) provide efficient short-term risk reduction; however, their effectiveness decreases over time as the lifetime of structural measures expires. Applying a comprehensive set of 13 interventions (P_1, green line) reduces flood risk to approximately 2 by 2050 and maintains this level of risk reduction throughout the modeling period until 2100.

However, achieving such an efficient flood risk reduction pathway requires the implementation of both lower-cost and high-cost measures (see Table 5 and Supplementary Material 5), resulting in a substantial financial burden. Risk reduction benefits and estimated total cost based on the simple mapping of low, medium, high costs to value for each pathway are presented in Table 5. As expected, higher levels of risk reduction are associated with higher implementation costs. For example, pathway P_1 provides a risk reduction benefit of 235.2 units at a total estimated cost of 73.1 million EUR, whereas pathway P_13 provides a risk reduction benefit of 15.4 units at a total estimated cost of 5.8 million EUR. Therefore, considerations of cost-efficiency can provide additional support for decision-making when selecting an adaptation pathway for implementation.

To assess the 37 modeled alternative pathways under the SSP5 scenario, two indicators were compared. Figure 8 demonstrates that complete risk reduction to zero level is not achievable through the set of measures considered for implementation. The flood risk reduction benefit, calculated as the integral of risk reduction over the 80-year modeling period for each pathway, represents the climate adaptation performance aspect. The second indicator relates to the implementation costs of measures, including capital and operational maintenance costs. A greater number of implemented measures generally requires higher costs, however, the flood risk reduction benefit per 1 million EUR is not directly proportional to the number of interventions. This indicates that cost-efficiency of an adaptation pathway depends on the type and combination of interventions selected, including both non-structural and structural measures. Therefore, by linking risk reduction benefits with total costs (Table 5), the benefit units per 1 million EUR were used as an indicator of cost-efficiency. The distribution of the 37 alternative pathways is presented by comparing the flood risk reduction benefits and the benefit units per 1 million EUR (Figure 10).

The range of pathways can be grouped into four categories:

Low risk reduction benefits—low cost-efficiency. This group includes nine adaptation pathways involving 2–3 measures that results in limited flood risk reduction (see Figure 8, purple line representing pathway P_27). Although, these pathways require relatively low implementation cost, they provide limited benefits per 1MEUR allocated/ invested. Therefore, this approach is not recommended due to its low efficiency.

Low risk reduction benefits–high cost-efficiency. This group includes 10 adaptation pathways involving 3–5 measures. Although the achieved risk reduction benefits are below average, cost-efficiency is above average due to the selection of predominantly low-cost measures.

High risk reduction benefits—high cost-efficiency. This group includes 14 adaptation pathways involving a range of 5 to 10 measures. Both risk reduction benefits and the cost-efficiency are above average. The strategy for selecting of measures in this group relies on a combination of low-cost and cost-intensive interventions.

High risk reduction benefits—low cost-efficiency. This group includes four adaptation pathways that achieve high flood risk reduction benefits through implementation of a comprehensive set of 11–13 interventions. However, the inclusion of costly measures reduces the benefit gained per 1 million EUR invested, resulting in below-average cost-efficiency.

Figure 11 presents the intermediate risk-reduction results for 2030 and 2050, together with the total reduction achieved by 2100. Out 37 alternative pathways, 28 were included in the comparison, while 9 pathways characterized by low-risk reduction and low cost-efficiency were intentionally excluded from further consideration as they were not recommended. The highest flood risk reduction, amounting to 4.1 units, is expected by the implementation of the Adaptation Pathway P_1 (Figure 11C) which includes the largest number of interventions (13). Other combinations of intervention result in lower level of risk reduction units achievable by 2100. Overall, the selected combinations of interventions achieve up to 55% of the potential flood risk reduction benefit by the end of the modeling period in 2100.

Comparing flood risk reduction benefits at the end of the modeling period (2100) reflects the maximum achievable long-term outcome, while intermediate modeling results provide insights into the temporal development of each pathway. In the shorter-term prediction for 2030 (Figure 11A), the highest achievable flood risk reduction, amounting to 1.8 units, is obtained through the Adaptation Pathway P_16. This pathway includes four measures, with the last measure implemented in two phases or interventions (both phases included in the pathway): non-structural measures focused on increasing public awareness and improving the efficiency of authorities; an urban structural measure aimed at increasing the capacity of culverts and open ditches; and a rural structural measure aimed at improving drainage capacity. The Adaptation pathway P_16 achieves the highest benefit per 1 million EUR invested (Figure 10).

The modeling assumes that the implementation of all selected interventions begins by 2030, with their effects maintained throughout the modeling period until 2100. No measures are scheduled to begin after 2050, as implementation at later stages was considered beyond the practical planning horizon of the adaptation framework. It is notable that all interventions under Adaptation Pathway P_16 are initiated by 2030, while the efficiency of flood risk reduction declines over the longer term due to the limited lifetime of the interventions' effectiveness (Figure 8). Nevertheless, some of the efficiency is maintained, resulting in a flood risk reduction benefit of 37% by 2100 (Figure 11C).

Adaptation pathways in the Zemgale region were initially developed qualitatively using participatory planning methods to incorporate stakeholder inputs and gather local knowledge. A more targeted quantitative evaluation and pathway refinement using the REPIE and DCRT models was conducted at a later stage, complementing quantitative data with socio-economic considerations.

Uptake of adaptation pathways in climate change adaptation policy and decision making

A series of adaptation pathways was developed to support governance and economic decision-making for effective flood risk management in the Zemgale region. Different combinations of non-structural and structural measures were evaluated using the PATH framework (REPIE and DCRT models) to estimate their efficiency in reducing flood risk.

Opportunities and limitations of applying climate adaptation pathways to regional flood risk management

Efficient flood risk management requires identifying the most beneficial combinations of risk reduction measures and implementing them in a sequential order within the budget constraints of municipalities. Decision-makers often need to make difficult choices, particularly when costly structural alternatives are being considered for financing. Comparing the quantified outcomes achievable through different pathways of measures provides evidence of their potential effectiveness and supports informed decision-making. The PATH framework presented in this study offers significant advantages for evaluating adaptation pathways by enabling the conversion of qualitative or semi-quantitative input metrics into quantitative estimates of flood risk reduction. This capability is particularly important for assessing the effectiveness of non-structural socio-institutional measures, such as strengthening the capacity of authorities and increasing public awareness of appropriate actions during flood events. Assessing a wide range of adaptation pathways enables decision-makers to develop informed strategies for flood risk reduction.

The PATH framework also provides intermediate modeling results, allowing for the assessment of short-term progress toward long-term objectives. By 2030, the Adaptation Pathway P_16 delivers the greatest reduction in flood risk, clearly demonstrating its short-term benefits. However, transitioning to Adaptation Pathway P_1 would provide greater long-term benefits by 2050 and 2100 (Figure 11). Moreover, stakeholder feedback (Workshop 8, Supplementary Material 1) indicated a preference for P_16 as the most suitable option for implementation to reduce flood risk in the Zemgale region.

The limitations of climate adaptation pathways are primarily related to the long modeling time horizon applied in this study (80 years) and the inherent uncertainty in estimating real-world implementation outcomes. Nevertheless, the simplified approach adopted by the PATH framework serves as an effective screening tool, providing rapid estimates of potential outcomes while allowing a wide range of adaptation options to be evaluated. To support implementation decisions, it is advisable for decision-makers to undertake a more detailed assessment of both the efficiency of flood risk reduction and the associated financial costs for the selected set of measures.

Quantifying the benefits of climate adaptation pathways for regional flood risk management

By targeting flood risk reduction, the adaptation pathways developed in this study adopt a performance-oriented approach based on quantitative modeling. Using the PATH framework, the effectiveness of individual adaptation measures is integrated into an overall estimate of flood risk reduction by the end of the modeling period (2100). From a flood risk management perspective, the objective is to maximize flood risk reduction while maintaining high cost-efficiency in the selected adaptation measures. As stated above, for a given year, risk is calculated based on the hazard, vulnerability, and exposure indicators. The benefit value of a pathway is calculated as the integral of the difference between the baseline risk curve for a given scenario and the corresponding pathway risk curve. Therefore, if the baseline risk remained at the maximum value of 10 throughout all 80 simulation years for a given scenario, an ideal pathway that reduced the risk to 0 throughout the entire period would have a benefit value of 800 (80 × 10). Thus, the benefit value essentially quantifies the overall flood risk reduction achieved by a pathway. The stakeholder- and expert-based assessment of the proposed measures indicated that most interventions are associated with medium implementation costs, with the exception of improving the functionality of rural drainage systems and gray infrastructure for urban rainwater collection, which were classified as high-cost measures (Table 4). Comparing flood risk reduction benefits with the corresponding benefit per 1 million EUR invested (Figure 10) provides valuable information to support investment planning and decision-making. The results indicate that pathways incorporating more costly interventions generally achieve greater flood risk reduction but often exhibit lower cost-efficiency, as reflected by below-average benefit per 1 million EUR invested.

Stakeholder process and added value in adaptation pathway development

Stakeholder involvement within IMPETUS framework was a key element in the development of tools to support climate change adaptation in the Zemgale region. The Regional Climate Change Adaptation Plan was developed as a strategic planning instrument, whereas Adaptation Pathways were developed as a decision-support framework to assist regional and local institutions in reducing flood-related risks.

In Latvia, climate change adaptation (CCA) measures are recommended to be implemented across all governance levels—local, regional, and national. The Ministry of Climate and Energy coordinates the implementation of national climate change adaptation policies and measures. National climate adaptation objectives are binding for planning regions and municipalities within the scope of their respective responsibilities. Planning regions are required to integrate CCA objectives, priorities, and implementation conditions into their sustainable development strategies and development programmes, while also defining the actions and measures needed to achieve these objectives. Similarly, local governments are expected to incorporate CCA objectives, priorities, activities, and investment projects into their sustainable development strategies, development programmes, and other territorial development planning documents, in accordance with the objectives and priorities established by the respective planning region (Klimatnoturibas un ekonomiskās ilgtspējas likums, 2025).

According to the project implementation schedule, the preparatory phase for both tools coincided, allowing the outcomes of stakeholder co-design and co-creation activities to serve two complementary purposes. The development of adaptation pathways was supported by essential knowledge of local conditions provided by stakeholders in the Zemgale region. During the preparation stage (Figure 3), this study benefited from the involvement of experts from six municipalities in the region, who shared their perspectives on vulnerabilities, governance gaps and challenges related to flood impacts. While stakeholder-driven adaptation pathway development is well-established concept, combining the participatory identification of adaptation options with PATH framework modeling enabled the integration of qualitative and quantitative metrics for evaluating these adaptation options.

Stakeholder involvement was important for providing feedback on alternative adaptation pathways, which were compared based on flood risk reduction efficiency and cost-efficiency. By presenting stakeholders with alternative pathways, the process supported more informed decision-making regarding actions to address flood risk. However, the implementation of adaptation measures remains primarily at the local level and will largely depend on the priorities and decisions of local governments. From another perspective, priorities for climate change adaptation, including flood risk mitigation, can be revised through updates to the midterm planning document Zemgale Regional Climate Change Adaptation Plan, incorporating the latest developments and approaches. The focus should be directed toward adaptive pathways planning decision making as conditions change (Haasnoot et al., 2024), integrating slow-onset impacts (such as environmental degradation) and sudden climate—related disasters in a Resilience Pathway (Zorita et al., 2025), while accounting for multi-risk interactions across various sectors (Schlumberger et al., 2025).

Data availability, assumptions

A simplified approach was used for estimating the effect of interventions on the studied system and the associated risk. For a more comprehensive analysis, the use of hydrological and hydraulic models would be required. Such a modeling approach would have allowed the incorporation of more detail in the analysis and would lead to more accurate results. Nevertheless, this would have been accompanied with a larger computational burden. Running elaborated models for simulation periods as large as 80 years and considering the large number of runs carried out in the context of the generic algorithm, could have rendered the application impractical. As opposed to this, a simpler methodology (which does not incorporate relevant dynamics in depth) allows for fast evaluations. In other words, there is a trade-off between accuracy and computational requirements. Comparing the current results with those of a methodology leveraging more sophisticated modeling approaches would be an interesting step forward.

Uncertainty in the modeling results is also associated with the limited availability of data on the environmental effectiveness of adaptation measures. While gray infrastructure measures have been implemented for several decades and can therefore provide directly measurable evidence of their effectiveness, nature-based solutions and retention measures have only recently been operationalized in the Zemgale region. As a result, available datasets for these approaches cover shorter time periods, and robust assessment of their effectiveness and co-benefits requires more complex and multi-dimensional data sources. Collecting and assessing evidence on the role of nature-based solutions for climate change adaptation is an important prerequisite for further operationalizing these measures in planning documents (Cohen-Shacham et al., 2019; Debele et al., 2023).

Outlook for future

Whereas this paper demonstrates the application of the adaptation pathways approach to support climate change adaptation planning, the framework has considerable potential for application in other policy domains that require long-term, adaptive decision-making. For example, the approach is highly relevant to biodiversity conservation and ecosystem restoration, where the implementation of restoration measures must remain responsive to evolving climate and ecological conditions, monitoring results, stakeholder priorities, and financial constraints. The adaptation pathways framework enables the identification of flexible sequences of interventions, reducing the risk of maladaptation or inefficient investments while maintaining progress toward long-term restoration objectives.

By incorporating climate change adaptation measures into local planning documents, municipalities in the Zemgale region demonstrate a commitment to implementing flood risk reduction strategies. Medium-term planning documents are regularly reviewed and updated to reflect emerging scientific evidence, improved knowledge, and technological advancements. This iterative planning process enables the modification and optimization of adaptation measures, thereby enhancing the effectiveness of flood risk reduction over time. In this context, the PATH framework, which generates alternative adaptation pathways, represents a valuable decision-support tool. It facilitates the identification and selection of the most appropriate and cost-effective adaptation measures within existing financial constraints.

The proposed multidisciplinary approach has a strong potential for supporting planning and decision making. A systemic perspective is strictly embedded in the process of analyzing all component of risks, encompassing surveys that take into consideration climate projections and physical-ecological processes as well as trends in land use, urban development, population and housing dynamics. The capacity to address the mutual relationships between risk factors and the effects of multiple adaptation solutions, included their implications from the socio-economic perspective, is one of the main strengths of adaptation pathways conceptual approach.

This study explored the application of climate adaptation pathways to regional flood risk management in the Zemgale region, with particular attention to the opportunities and limitations of co-designed and co-creative pathway-based planning and the role of benefit quantification in supporting climate change adaptation decisions. The findings demonstrate that adaptation pathways can provide a useful framework for addressing the long-term and uncertain nature of climate-related flood risk, while also highlighting several methodological and practical limitations.

The integration of stakeholder assessment with model-based information demonstrates the importance of combining quantitative and qualitative evidence in adaptation planning. While the modeled pathways incorporating a broader set of adaptation interventions generally indicated greater potential for long-term risk reduction and cost-efficiency, stakeholder preferences, based on local knowledge and experience, did not necessarily correspond to the pathways with the highest modeled risk reduction. This highlights that adaptation decisions are influenced not only by technical effectiveness and economic costs, but also by stakeholders' priorities, the social readiness and perceived feasibility of implementing particular measures.

The identification of alternative pathways enables spatial and land-use planners and decision-makers to compare different combinations, timing and sequencing of measures and to consider trade-offs between flood-risk reduction, costs and stakeholder preferences. The approach also supports flexible, long-term planning by allowing adaptation strategies to be adjusted as risks and conditions tend to change. However, regional-scale pathways may mask differences between municipalities, urban and rural areas, and different flood mechanisms. Their applicability is also constrained by data availability, modeling assumptions and uncertainties in climate scenarios and socio-economic conditions. Therefore, regional pathways should be viewed as a strategic framework that can be complemented by more detailed local assessments.

Quantifying the benefits of adaptation pathways enables a more transparent comparison of alternative strategies in terms of their risk-reduction potential and costs. In the Zemgale case, it supported the identification of pathways offering different balances between investment and flood-risk reduction and helped to illustrate how the benefits of adaptation evolve over time. However, quantified benefits should be considered together with stakeholder preferences, feasibility and local conditions, as the pathway with the highest quantified benefit is not necessarily the most preferred or feasible option. Thus, benefit quantification can strengthen regional flood risk management by providing an evidence base for prioritizing and sequencing adaptation measures while retaining flexibility under uncertain future circumstances.

Overall, the study demonstrates that combining climate adaptation pathways with benefit quantification can strengthen regional flood risk management by making future adaptation choices more transparent, comparable and adjustable. The approach supports a transition from selecting individual adaptation measures toward managing sequences of interventions that can evolve with changing climate risks. For Zemgale, this suggests that regional flood risk management should focus on establishing feasible, cost-effective, near-term actions while retaining the flexibility to introduce additional measures as risks increase. Future research should further improve the spatial differentiation of pathways and strengthen the quantification of multiple adaptation benefits, thereby enhancing the robustness of the developed pathways across a wider range of climate, socio-economic, and implementation conditions.