A cluster of recent floods in the UK has prompted significant interest in the question of whether floods are becoming more frequent or severe over time. Many trend assessments have addressed this in recent decades, typically concluding that there is evidence for positive trends in flood magnitude at the national scale. However, trend testing is a contentious area, and the resilience of such conclusions must be tested rigorously. Here, we provide a comprehensive assessment of flood magnitude trends using the UK national flood dataset (NRFA Peak Flows). Importantly, we assess trends using this full dataset as well as a subset of near-natural catchments with high-quality flood data. While headline conclusions are useful for advancing national flood-risk policy, for on-the-ground flood-risk estimation it is important to unpack these local changes to determine how climate-driven trends compare with those from the wider dataset that are subject to a wide range of human disturbances and data limitations. We also examine the sensitivity of reported trends to changes in study time window using a ‘multitemporal’ analysis. We find that the headline claim of increased flooding generally holds up regionally to nationally, although we show a much more complicated picture of spatio-temporal variability. While some reported trends, such as increased flooding in northern and western Britain, appear to be robust, trends in other regions are more mixed spatially and temporally – for example, trends in recent decades are not necessarily representative of longer-term change, and within regions (e.g. in southeast England) increasing and decreasing trends can be found in close proximity. While headline conclusions are useful for advancing national flood-risk policy, for flood-risk estimation it is important to unpack these local changes, and the results and methodological toolkit provided here could provide such supporting information to practitioners.

英国最近发生的一系列洪水引起了人们对洪水是否会随着时间的推移变得更加频繁或严重的问题的极大兴趣。近几十年来，许多趋势评估已经解决了这个问题，通常得出的结论是，有证据表明全国范围内的洪水规模呈积极趋势。然而，趋势测试是一个有争议的领域，必须严格测试此类结论的弹性。在这里，我们使用英国国家洪水数据集（NRFA 峰值流量）对洪水规模趋势进行了全面评估。重要的是，我们使用这个完整的数据集以及具有高质量洪水数据的近自然集水区的子集来评估趋势。虽然总体结论有助于推进国家洪水风险政策，对于实地洪水风险估计，重要的是要解开这些局部变化，以确定气候驱动的趋势与受广泛人类干扰和数据限制的更广泛数据集中的趋势相比如何。我们还使用“多时态”分析检查报告的趋势对研究时间窗口变化的敏感性。我们发现，尽管我们展示了更为复杂的时空变化图景，但洪水增加的总体主张通常在区域和全国范围内都成立。虽然一些报告的趋势，例如英国北部和西部的洪水增加，似乎很强劲，但其他地区的趋势在空间和时间上更加复杂——例如，近几十年来的趋势不一定代表长期变化，而在地区（例如 在英格兰东南部）附近可以发现增加和减少的趋势。虽然标题结论对推进国家洪水风险政策很有用，但对于洪水风险估计，解开这些局部变化很重要，这里提供的结果和方法工具包可以为从业者提供此类支持信息。