Satellite remote sensing of river discharge (RSQ) algorithms provide a useful source of observations to supplement river gauge records. RSQ algorithms have existed for over a decade yet their widespread use has been impeded by a lack of operational usability and quantitative characterization of uncertainty. Here we present RODEO, an algorithm for estimating river discharge using Landsat observations in near-real time. RODEO is validated with 456 gauges (median Kling-Gupta efficiency = 0.3) and uses a novel quantile rating curve technique that pairs Landsat river widths with discharge estimates from a global hydrologic model. RODEO also characterizes the uncertainty of RSQ estimates (estimated root-mean-square error = +7%), enabling RSQ retrievals to be used for data assimilation into hydrologic models. With the goal of expanding the RSQ user base, the RODEO algorithm is implemented as a freely available, off-the-shelf cloud-based Google Earth Engine application that provides RSQ estimates across North America from 1984-present.

河流流量卫星遥感 (RSQ) 算法提供了有用的观测资料，以补充河流流量记录。RSQ 算法已经存在了十多年，但由于缺乏操作可用性和不确定性的定量表征，它们的广泛使用受到了阻碍。在这里，我们介绍了 RODEO，这是一种使用 Landsat 观测近乎实时地估算河流流量的算法。RODEO 已使用 456 个仪表（中值 Kling-Gupta 效率 = 0.3）进行验证，并使用一种新颖的分位数评级曲线技术，将 Landsat 河流宽度与全球水文模型的流量估计值配对。RODEO 还表征 RSQ 估计的不确定性（估计均方根误差 = +7%），使 RSQ 检索能够用于将数据同化到水文模型中。以扩大 RSQ 用户群为目标，