Abstract The Rio de la Plata is a micro-tidal estuary located in Southeast South America. With an annual mean flow of 26,500 m 3 /s, it receives 160 million tons/yr of suspended sediments. The high content of cohesive fine sediments in the estuary generates high turbidity levels in its inner and intermediate zones, which can be clearly seen in color satellite images. In this work, an image-based algorithm was successfully implemented to remotely detect the turbidity front of the Rio de la Plata, based on the top of the atmosphere (TOA) reflectance in the red band of MODIS-Aqua satellite. The algorithm finds the reflectance level that ‘best’ separates two water classes: turbid fluvial and clear ocean waters. The front dynamic was studied combining remotely sensed information and data of river discharge, winds, salinity and sea level time series, in the four-year period 2014–2017. River discharge was identified as the main external forcing, revealing a solid general pattern of behavior: when discharge was high (low) the front tended to be located in the outer (intermediate) zone of the estuary. Sea level seemed to be a secondary forcing, presenting higher correlations along the center of the estuary than near both coasts. Local winds needed to have a relatively persistent (2-day) component in a given direction to affect the location of the front. Additionally, results of an already implemented numerical model of the Rio de la Plata were evaluated in terms of spatio-temporal performance, considering turbidity and salinity fronts locations. New strengths and limitations of the model were identified, and an improvement in the parameterization of sediments’ settling velocity was tested. Model results revealed the relative importance of bottom shear stress on the general location of the front, and of salinity on the flocculation process of cohesive sediments. This work provided new insights for the understanding of the Rio de la Plata estuarine dynamics through the combination of three complementary tools – remote sensing, in situ data, and numerical modeling, – which may be extended to other systems around the world.

中文翻译：

---

Río de la Plata 浊流锋的时空动态；将遥感与原位测量和数值建模相结合

摘要 拉普拉塔河是位于南美洲东南部的一个微潮河口。它的年平均流量为 26,500 m 3 /s，接收悬浮沉积物 1.6 亿吨/年。河口内粘性细粒沉积物含量高，导致其内部和中间带的浊度较高，在彩色卫星图像中可以清楚地看到。在这项工作中，基于 MODIS-Aqua 卫星红带中的大气顶部 (TOA) 反射率，成功实施了一种基于图像的算法来远程检测拉普拉塔河的浊度前沿。该算法找到了“最佳”区分两个水类的反射率水平：浑浊的河流和清澈的海水。结合遥感信息和河流流量、风、盐度和海平面时间序列数据，对锋面动态进行了研究，在 2014-2017 年的四年期间。河流流量被确定为主要的外部强迫，揭示了一个坚实的一般行为模式：当流量高（低）时，锋面往往位于河口的外（中）带。海平面似乎是次要强迫，沿河口中心呈现出比两个海岸附近更高的相关性。局部风需要在给定方向上具有相对持久（2 天）的分量才能影响锋面的位置。此外，考虑到浊度和盐度前沿位置，已经实施的拉普拉塔河数值模型的结果在时空性能方面进行了评估。确定了模型的新优势和局限性，并测试了沉积物沉降速度参数化的改进。模型结果揭示了底部剪应力对前缘一般位置的相对重要性，以及盐度对粘性沉积物絮凝过程的相对重要性。这项工作通过结合三种互补工具——遥感、原位数据和数值建模——为理解拉普拉塔河口动力学提供了新的见解，这些工具可能会扩展到世界各地的其他系统。