Subgrid sampling (SGS) has been increasingly integrated into hydraulic modelling to refine terrain representation without excessively increasing computational demands. This study investigates the effects of SGS and nature-based solutions (NbS), focusing on flow velocities, storage volumes and mass balance stability. Using a rain-on-grid TUFLOW model, we simulate a 1-in-30-year flood event across different NbS interventions, including leaky dams, floodplain planting and bunds with floodable depressions. Results demonstrate that higher SGS frequencies of 11 significantly enhance terrain resolution, improving the representation of flow paths and floodplain activation. However, a lower SGS frequency of three introduces notable numerical artefacts, leading to persistent negative cumulative volume errors and misrepresenting water retention within the system. NbS interventions generally minimise peak flow rates, with bunds and depressions (Option 3) proving the most effective, achieving a storage volume increase of up to 162.83% compared to the baseline scenario. Mass balance assessments show that models without SGS overestimate flood storage, potentially misguiding NbS implementation and compliance with flood storage regulations such as the UK Reservoirs Act (1975).

Study region: The study focuses on Kerala, a state in the southwest of India. Kerala is composed of 14 districts, each characterised by variations in topography, climate, and land use patterns. Study focus: This review critically analyses the literature on flood risk assessment (FRA) in Kerala, particularly after the devastating 2018 floods. Kerala has experienced sporadic floods in the 21st century, caused by localized heavy rainfall, rapid urbanization and improper management of water resources. The 2018 Kerala Floods was one of the most catastrophic recent floods. Nearly all 14 districts were affected, over 480 people lost their lives, and more than a million were displaced. Anthropogenic factors, such as encroachment on wetlands, sand mining in riverbeds, and inadequate drainage systems in urban areas, have worsened the impact of floods. Despite the long history of flooding, flood management in Kerala has struggled to keep pace with the increasing magnitude and frequency of these events. Factors such as outdated infrastructure, uncoordinated dam management, and poor urban planning have exacerbated the impacts of floods. Against this backdrop, this review on flood risk assessment (FRA) in Kerala evaluates and synthesises existing methodologies to improve understanding of current state-of-the-art FRA in Kerala and provide a foundation for more effective flood management strategies. New hydrological insights for the region: The review identifies that research conducted after the 2018 floods can be categorised into three broad methodological themes: Remote Sensing and Geographic Information Systems (GIS), Predictive Modeling (including hydrological and hydraulic simulations), and Analytical Approaches (such as machine learning, statistical methods, and multi-criteria decision-making). The spatial focus of the studies reveals significant disparities, with Allapuzha being the most extensively studied district and Thiruvananthapuram receiving minimal attention. The review identifies critical gaps in the literature, including challenges in translating mitigation strategies, urban flooding stemming from poor land use planning, insufficient integration of various flood sources, and limited research on compound extreme events. Highlighting the urgency of translating the quantification of hazard to mitigation and the integration of climate change projections, the article provides avenues for further research for FRA in Kerala.