Analysis of enhanced observations made by a SeaExplorer glider deployed in the central Mediterranean

Abstract

As the Earth’s atmosphere continues to heat up and change the environment in unpredictable ways, it is necessary to not only have tools that are able to accurately predict future trends, but also to have an exact baseline of the studied properties to be able to accurately measure how these parameters may vary. The aim of this study is to verify existing oceanographic models and satellites in the Mediterranean Sea through the use of primary data collection on temperature, salinity, and chlorophyll-a through the use of an ocean glider. This ocean glider was deployed on a summer mission and a winter mission from areas along the Maltese Islands headed towards the northern coast of Africa, diving to depths of up to 700 metres. Ocean gliders are a branch of Underwater Autonomous Vehicles used for oceanographic measurements on a wide range of properties. During this time, satellite and model products were downloaded from an online service known as Copernicus Marine Environmental Monitoring Service which is widely relied upon across the world for policy making, climate monitoring, forecast predictions, and a realm of other services. All three datasets were uploaded to Matlab where the program could loop through to find the nearest model and satellite measurement in time and space to the nearest glider data point. A series of statistical analyses were performed to test the accuracy of the model and satellite products in comparison to the measured glider data. The accuracy between Missions 1 and 2 were also tested. These measurements included error bar plots, Mann-Whitney tests, correlations and Wilcoxon Signed Rank tests. A unique component of this study was the measure of correlation under certain conditions, this was done to determine if there were certain times or locations that the model or satellite products may be less accurate. These specific conditions measured were the correlation with the duration of the mission, with depth, and with the presence of backscattering. The key findings of this study included the importance of recalibration of the ocean glider perhaps earlier than what was previously predicted. Additionally, the reliability of using an ocean glider rather than model or satellite data when measuring any of the three parameters within the first 200 metres from the surface was realized. Finally, it was found that 95.5 percent of chlorophyll-a data points measured by the ocean glider were greater than the data points measured by the model, indicating there is likely an error with the model used for the chlorophyll-a product.