Feature ArticleDeploying AUVs In Restricted Areas
Balearic Island Coastal Observing and Forecasting System
Balearic Islands, Spain
Autonomous Vehicle Laboratory
Unidad de Tecnología Marina
Consejo Superior de Investigaciones Cientificas
Department of Physiology and Mollusc Culture
Instituto de Investigaciones Mariñas
Consejo Superior de Investigaciones Cientificas
Map of the study area at the Lorbé mussel farm, showing the final track of the AUV across the rafts.
Conducting environmental studies of these sites is essential to determine the conditions for optimal growing rates in order to improve the aquaculture. Key variables in these studies are the amount of food, which is estimated from the fluorescence of the water column, and the current velocity, an index of the water renewal time. In addition, the physicochemical parameters, such as temperature, salinity, oxygen and turbidity, provide water quality information, which could help predict toxic algae blooms that could lead to farm closures and economic losses.
Presently, the primary way to monitor these parameters in mussel farms is to employ sensors on moorings. However, the use of AUVs can improve knowledge of the biogeochemical processes occurring in these sites by obtaining data that complements information gathered from the mooring observatories.
EcoMapper AUV in Galicia Mussel Farm
YSI Inc.'s (Yellow Springs, Ohio) EcoMapper AUV, a small, portable vehicle, was used in an October 2011 study to collect bathymetry and water quality data in a mussel farm in Lorbé, Galicia. The AUV has sensors for measuring CTD, dissolved oxygen, chlorophyll, pH, ORP (oxidation reduction potential) and turbidity, as well as an acoustic Doppler current profiler (ADCP) and Doppler velocity log. A two-man team equipped with the EcoMapper visited the mussel farm to explore the possibility of collecting data by AUV.
An AUV is greatly influenced by environmental conditions. The study was conducted under a worst-case scenario setup, providing an opportunity to test some of the EcoMapper's operational limits. There were several constraints on the vehicle's operations. Mussel farms are dynamic. Rafts are anchored at one point, but they move with the wind and tidal currents. The area of study was a very active mussel farm with several 30-ton service boats operating continuously, which presented the risk of collision with the AUV.
The mussel farm was located on a small bay on the Rías Altas, with a beach surrounded by rocky cliffs. Options for an emergency path were limited, and no previous bathymetry or side scan data of the area existed. The seafloor below the mussel farm was full of debris, including anchors, ropes and mussel lines. In general, this region suffers adverse weather during autumn and winter.
Preparation started with three steps that are general for all mini-AUV campaigns, while the fourth was specifically for studies in restricted areas. The first step was to calibrate the sensors with the manufacturer's standard procedures. Calibration for the fluorescence and turbidity sensors needed to be improved using in-situ samples.
The second step was to calibrate the compass. The EcoMapper relies on a magnetic compass and a Doppler velocity log for navigation, which are usually enough for standard surveys. In this case, it was crucial to have them well-calibrated to minimize navigation errors. The third step was to calibrate the most accurate georeferenced image of the area, i.e., the latest one. In this case, an error of a few meters was found in the image, which was critical due to the narrow navigation margin.
For restricted areas, it is important to know the exact position of the fixed potential risk, which, in this study, was the rafts. The mussel farm had up to 100 rafts, which, despite being anchored to the seafloor, moved with wind and tidal currents and were not in the positions shown on the georeferenced image. To account for this dynamic risk, a new map of the rafts' boundaries needed to be drawn every day.
Each day, the first step of the survey was to perform one tour around the rafts using a small boat, getting as close to them as possible to obtain GPS positioning, create safe passages and inform the AUV planning software for the day. The ideal way to carry out the study would have been to create safe AUV passages before deployment using wind and current historical information, as well as raft dimensions, anchor points and predicted raft drifting, to estimate raft movement without the GPS presurvey. This was not done because it would have been time-consuming. To continue this article please click here.
David Roque is a glider technician at the Balearic Island Coastal Observing and Forecasting System. He has a master's degree in marine sciences and has worked at the Consejo Superior de Investigaciones Cientificas and in the Department of Applied Physics at the University of Cadiz.
Pablo Rodríguez is the technical manager of the Autonomous Vehicle Laboratory at the Unidad de Tecnología Marina (Spanish Research Council). He has a telecommunications engineering degree and worked with acoustical and geophysical instrumentation for 15 years before moving to the AUV world.
Uxio Labarta is a research professor at the Department of Physiology and Mollusc Culture at the Instituto de Investigaciones Mariñas. He has been working for 40 years in ecophysiology and mussel culture management in Galicia, Spain.