What is the value of individual sensors in river networks?

There is an increasing interest to install dense networks of sensors in stream networks to help predict flow and water quality - but the questions of how many sensors and where is an open question. This project aims to link the strengths of a deep learning model and a physics-based model to quantify the value of each new sensor in the river network. This work will help inform stakeholders decide where to implement new sensors in a monitoring network to maximize the value for model predictions.