Water Quality Monitoring Using Unmanned Aerial Systems Imagery and a Novel Autonomous Surface Vessel

Although the science and technology for retrieval of key water quality parameters from satellite data has matured significantly, the resolution and quality of remote sensing data can be limited by technical constraints, such as the sensor's spatial and spectral resolution and the unavailability of usable satellite data for extended time periods due to cloud coverage. In comparison, unmanned aerial systems (UASs) based multispectral remote sensing applications have the potential to provide critical information needed for water quality monitoring of water bodies at a local scale. However, there are a few challenges to the feasibility of using UAS-based remote sensing for water quality monitoring. Some of those challenges were addressed in this study, a set of algorithms were developed for a suite of water quality parameters, and a time-series analyses was performed. Several field campaigns were undertaken to collect field data from the water over the Merill Shell oyster reef in the Western Mississippi Sound. A multispectral sensor on a drone to collect low altitude aerial imagery and simultaneously, an autonomous surface vessel (ASV) with a plethora of sensors were deployed over the same sites. A series of automated procedures were developed and implemented for the geometric and radiometric corrections of the UAS imagery and algorithms were developed for a suite of water quality parameters using empirical and machine learning approaches. The algorithms were applied to the image mosaics and a time-series analysis of the water quality parameters with discharge was performed to determine the relative influence of the major rivers on the water quality of the study area. All the developed algorithms are promising with reasonably high R ² and low RMSE. Use of ASV and UAS data together helped develop robust algorithms, using which the relative influence of several rivers on the water quality on the oyster reef could be evaluated.