Sub-daily rainfall observations are vital to help us understand, model and adapt to changing climate extremes. However, gauge records often have quality issues, for example due to equipment malfunctions and recording errors. This paper presents a new, open-source quality control algorithm (GSDR-QC) to identify these issues in hourly rainfall data, along with an application of the algorithm to the Global Sub-Daily Rainfall (GSDR) observational dataset. The algorithm is based on 25 quality checks, which are combined into a simple rule base to remove suspicious data. The quality checks and rule base are adaptable to help incorporate local or regional information. Comparison with manually quality-controlled gauge data shows that the procedure results in an overall improvement to the quality of the GSDR dataset. A UK case study further demonstrates the performance of the GSDR-QC procedure, while showing how region-specific data and understanding can be incorporated into the quality control process.

次日降雨观测对于帮助我们了解、模拟和适应不断变化的极端气候至关重要。但是，仪表记录通常存在质量问题，例如由于设备故障和记录错误。本文提出了一种新的开源质量控制算法 (GSDR-QC)，以识别每小时降雨数据中的这些问题，并将该算法应用于全球次日降雨 (GSDR) 观测数据集。该算法基于 25 项质量检查，这些检查组合成一个简单的规则库以删除可疑数据。质量检查和规则库适用于帮助整合本地或区域信息。与手动质量控制仪表数据的比较表明，该程序导致 GSDR 数据集质量的整体改进。