Ecosystem services are essential for the livelihoods of marginalized communities; however, their supply can be severely disturbed when natural hazards impact on low-lying coastal regions. To understand potential disaster impacts on sustainable livelihoods, ecosystem resilience needs to be assessed before and after a disaster. We examined ecosystem resilience to a tropical cyclone in a southwestern coastal region of Bangladesh in terms of its impacts on the ecosystem’s service supply proficiency (ESSP). We adopted a two-tiered methodology commencing with landcover mapping using object-based image analysis (OBIA) to identify five distinct landcovers (cropland, mangrove forest, riparian forest, tidal flat, and sandy beach). Then, Likert scores were used to reflect human perceptions of the supply proficiency of three types of ecosystem services (provisioning, regulating, and cultural services) associated with each landcover. The resilience of each landcover type was assessed by integrating maps of damage and recovery with the Likert scores to calculate ESSP scores. A substantial post-cyclone reduction in ESSP scores was revealed. Significant recovery (both short- and long-term) was observed in most of the affected ecosystems, albeit at somewhat different rates for different ecosystems. This study provides a multi-faceted evaluation of cyclone impacts on existing ecosystems and insight into ecosystem behaviour at different temporal phases of a disaster event. Accordingly, the study offers a novel way to study ecosystem resilience to support effective coastal planning within a vulnerable coastal community.

生态系统服务对于边缘化社区的生活至关重要。但是，当自然灾害影响低洼的沿海地区时，它们的供应就会受到严重干扰。为了了解灾难对可持续生计的潜在影响，需要在灾难发生之前和之后评估生态系统的复原力。我们研究了孟加拉国西南沿海地区的生态系统对热带气旋的适应力，它对生态系统服务供应能力（ESSP）的影响。从采用基于对象的图像分析（OBIA）进行土地覆盖制图开始，我们采用了两层方法，以识别五个不同的土地覆被（农田，红树林，河岸森林，滩涂和沙滩）。然后，利克特分数用于反映人类对与每个土地覆被相关的三种类型的生态系统服务（供应，调节和文化服务）的熟练程度的认识。通过将破坏和恢复图与李克特（Likert）得分相集成，以计算ESSP得分，评估每种土地覆被类型的复原力。发现旋风后ESSP分数大幅降低。在大多数受影响的生态系统中都观察到了显着的恢复（短期和长期），尽管不同生态系统的恢复速度有所不同。这项研究提供了对飓风对现有生态系统的影响的多方面评估，并深入了解了灾害事件不同时间阶段的生态系统行为。因此，