Veronika is a software engineer with experience in data science, first gained through her participation in Ericsson Nikola Tesla summer camp. For the past three years, she has been working on data science projects, primarily focused on the e-environment domain. Using Python and its ecosystem, her work includes data preprocessing, analysis, and building and evaluating ML models, with focus on Sentinel satellite images. She is also building experience in project deployment and leading a team.

Using satellite imagery to track water quality – how did we implement it?

Leveraging current research and experience, we utilize Sentinel-2 L1C satelliteimages, applying Acolite for atmospheric correction. Following this correction, state-of-art algorithms are applied to estimate the desired parameters (e.g., chlorophyll-a concentration and turbidity levels), while we are also developing machine learning models that aim to surpass said performance for our regions of interest. Throughout the data science phase of the project, various preprocessing and transformation steps are applied. As we transition toward deployment, we’ve chosen to integrate several tools, frameworks, and libraries, including Docker, Flask, MongoDB, Celery, and RabbitMQ. These technologies ensure the application runs smoothly and is scalable for future enhancements.