Project Details

  • Client: ESA - European Space Agency
  • Date: June 2020 - ongoing
  • Manager: Theodora Papadopoulou
  • Features: EnvironmentR&D

Project Description

The project introduced a breadboard for end-to-end (E2E) Marine Litter Optical Performance Simulations (ML-OPSI), designed in the frame of the ESA Discovery Campaign. ML-OPSI supports Earth observation scientists with the design of computational experiments for operations research. Thus, it allows EO scientist to create a model of the acquisition of an optical and IR signal by a remote sensing sensor and use it to run simulations to characterize that signal under varying environmental conditions, such as sea state, etc., to assess the optimal characteristics required for a sensor designed with an aim to detect aggregates of marine litter.

The ML-OPSI breadboard will estimate marine litter signal at Top-Of-Atmosphere (TOA) or at bottom-of-atmosphere (BOA), representing the various case studies (e.g., windrows, frontal areas, river mouths, sub-tropical gyres), coming from synthetic data or from real observations. It is a modular and extensible framework, promoting re-use and being adapted for different missions, sensors and scenarios. ML-OPSI breadboard is based on a component-based architecture and is fundamentally built over several modules, both of the Simulator and of the Model, that take care of various aspects of the E2E simulation. The Simulator (OPSI) modules are responsible for the execution of the Model over varying initial conditions, these comprise the orchestrator, that controls the running of the processes, along with configuration management and the proposed GUI model-builder front end. A Domain, in this instance Marine Litter (ML) detection, can be represented by one or more EO Models, that are made up of one or more EO Components.

Through a process of functional decomposition these top-level generic components have been refined to specify lower level modules, that are themselves components that can be modularized and the definition of inputs, outputs, parameters and interfaces of these components can be performed. The top level identified modules are: scenario/model builder (OPSI), performance assessment module (OPSI), scene generation module, atmospheric propagation module, instrument detection module and retrieval module. A standalone demonstrator, built as a rapid prototype implementing three of the modules, was constructed.