Atmospheric rivers (ARs) are elongated narrow bands of enhanced water vapor that can cause intense rainfall and flooding. ARs only appeared in the literature the last 30 years, and there has been much debate about how to define ARs and how to identify them. As a result, a wide range of AR identification algorithms have been produced with variations in the conditions required for an object to be classified as an AR and differences in the input data. One of the key conditions in most AR identification algorithms is a minimum threshold of water vapor flux, along with geometric criteria. The aim of this study is to explore uncertainties in global AR identification based on a single integrated water vapor transport (IVT)‐based identification method. We conduct a sensitivity analysis under one algorithmic framework to explore the effects of different IVT thresholds, input data resolutions, and regridding methods during the Years of Tropical Convection operational analysis (May 2008 to April 2010). We found that the resolution and regridding method affects the number of ARs identified but the seasonal cycle is maintained. AR identification is highly sensitive to the choice of IVT threshold; importantly, the commonly used 250 kg m⁻¹ s⁻¹ IVT threshold is not appropriate for global studies with detection methods that also include a restrictive geometric condition as this combination can lead to the strongest systems failing to be identified. The uncertainties within a single AR detection method and input data parameters may be as large as uncertainties across AR detection methodologies.

中文翻译：

---

大气河流识别对综合水汽输送阈值，分辨率和重新划分方法的敏感性

大气河流（ARs）是细长的增强水蒸气的窄带，可能引起强降雨和洪水。AR仅在最近30年才出现在文献中，关于如何定义AR和如何识别AR一直存在很多争论。结果，已经产生了各种各样的AR识别算法，其中将对象分类为AR所需的条件的变化以及输入数据的差异。大多数AR识别算法中的关键条件之一是水蒸气通量的最小阈值以及几何标准。这项研究的目的是探索基于单一集成水蒸气输送（IVT）的识别方法在全球AR识别中的不确定性。我们在一个算法框架下进行敏感性分析，以探讨热带对流年（2008年5月至2010年4月）期间不同IVT阈值，输入数据分辨率和重新网格化方法的影响。我们发现，分辨率和重新网格化方法会影响已识别的AR数量，但会保持季节性周期。AR识别对IVT阈值的选择非常敏感；重要的是，常用的250公斤米^-1  s ^-1 IVT阈值不适用于还包括限制性几何条件的检测方法的全局研究，因为这种组合可能导致无法识别最强的系统。单个AR检测方法和输入数据参数中的不确定性可能与AR检测方法中的不确定性一样大。