Simulation of Sentinel-2 Images

In 2015 European Space Agency (ESA) is planning to launch Sentinel-2 earth observation mission. It will provide continuity of high resolution satellites such as LANDSAT, SPOT, and IRS P6 with improved spatial and/or spectral resolution for services such as forest monitoring, complex land cover and land (LC-LU) use and changes detection or natural disasters management, thus supporting the needs of Copernicus – The European Earth Observation Programme.

ESA’s satellite Sentinel-2 will be used for the observation of land cover-land use.

Using advanced airborne hyperspectral remote sensing IES will simulate Sentinel-2 data to continue the development of LC-LU classification algorithms specifically for Latvian user needs and to perform a feasibility study on fusion of Sentinel-2 and Landsat-8 TIR satellite data with airborne remote sensing. The algorithm will help to use satellite data for analysis of land cover-land use, spread of invasive species, dynamics of forest resources and vitality, identification of damaged areas, wetlands, flood risk areas and degraded ecosystems and evaluate various aspects of biodiversity loss. The results could be used by the Rural Support Service, the Latvian Rural Advisory and Training Centre, Ministry of Agriculture, Ministry of Environment and Regional Development, Latvian Environment, Geology and Meteorology Centre and other institutions.

The Objectives

1. To develop a LC-LU classification scheme to support EU and national decision making processes and efficient policy accounting (CAP, subsidies, biodiversity conservation, regional development, climate adaptation, flood risk management, ecological corridors, etc.)
2. To adapt IES’s flying laboratory to perform a simulation of Sentinel-2 imagery;
3. To develop a LC-LU classification algorithm for Sentinel-2 data and implement into a prototype by using experience and products developed in the frame of other EU-funded projects (e.g. Geoland 2, HLANDATA);
4. To verify the classification results with high-resolution airborne RGB images, obtained simultaneously with hyperspectral data and in-situ ground measurements for difficult or questionable sites or classes;
5. To perform a feasibility study on fusion of Sentinel-2 satellite images (simulated by airborne sensors in 2015, and obtained from the satellite after it is launched) with  airborne hyperspectral data and Landsat-8 Thermal Infrared (TIR) bands to extract more sub-classes;
6. After the satellite’s launch in 2015, to assess the influence of atmosphere on satellite data signal by comparing the real Sentinel-2 data and results of its simulation by airborne sensors.

Our role

IES is the main contractor. The airborne remote sensing data will be collected with IES’s Environmental Surveillance and Monitoring System ARSENAL. For verification of LC-LU classes an image of high spatial resolution (0.2-4m), covering the same swath, will be acquired simultaneously with the hyperspectral data.

The observation will be fulfilled by two hyperspectral airborne imaging sensors (infrared light) and a high resolution RGB camera (visible light) which are embedded in the ARSENAL. The spectral band of the two hyperspectral sensors will be applied to simulate the Sentinel-2 bands and the maximum swath and required spatial resolution both for hyperspectral and RGB sensors will be reached by careful data collection planning.

As a result of the project, IES will develop a LC-LU classification algorithm for the Sentinel-2 data.

Project duration: May 2015 – April 2017

Project Manager: Dr.Phys. Dainis Jakovels

The Project is financed by: European Space Agency’s programme The Plan for European Co-operating States