Global hydrological models are essential for managing water resources and predicting hydrological events. However, the local-scale usability of global models challenges big-data management, communication, adoption, and validation. Validation is the biggest challenge bercause of the need for large-scale data management and model calibration, which requires extensive and often inaccessible observed data. This study assesses the GEOGLOWS-ECMWF Global Hydrologic Model, revealing systematic biases that impact its accuracy. We propose a bias-correction methodology using flow duration curves to align non-exceedance probabilities of simulated and observed streamflow, significantly improving the GEOGLOWS model. Unfortunately, this approach does not inherently improve simulations in ungauged locations. The methodology not only enhances the GEOGLOWS model's accuracy but also stands as a versatile solution applicable across various hydrological models. This bias correction approach provides a tool for improving hydrological predictions and gives users the confidence to use global models for local water resource management and decision-making processes.

中文翻译：

---

GEOGLOWS ECMWF 流流水文模型的历史模拟性能评估和月流速曲线分位数图 （MFDC-QM）


全球水文模型对于管理水资源和预测水文事件至关重要。然而，全球模型的本地规模可用性对大数据管理、通信、采用和验证提出了挑战。验证是最大的挑战，因为需要大规模数据管理和模型校准，这需要大量且通常无法访问的观测数据。本研究评估了 GEOGLOWS-ECMWF 全球水文模型，揭示了影响其准确性的系统偏差。我们提出了一种偏差校正方法，使用流持续时间曲线来调整模拟和观测的溪流的非超额概率，从而显着改进 GEOGLOWS 模型。遗憾的是，这种方法本身并不能改善未测量位置的模拟。该方法不仅提高了 GEOGLOWS 模型的准确性，而且还是适用于各种水文模型的通用解决方案。这种偏差校正方法为改进水文预测提供了工具，并让用户有信心使用全球模型进行本地水资源管理和决策过程。