Projects / Programmes
M3Sat - Methodology of Multitemporal Multisensor Satellite Image Analysis
Code |
Science |
Field |
Subfield |
2.17.00 |
Engineering sciences and technologies |
Geodesy |
|
Code |
Science |
Field |
T181 |
Technological sciences |
Remote sensing |
Code |
Science |
Field |
2.07 |
Engineering and Technology |
Environmental engineering
|
remote sensing, satellite images, time series
Researchers (12)
Organisations (2)
Abstract
Satellite Earth observation has become an important tool in a wide variety of studies and applications – from archaeology, biology and ecology, geography, spatial planning, meteorology, geology, agriculture and forestry to geodesy. Increased number of satellites, caused by the emergence of small satellite technologies and the Copernicus initiative of the European Union, coupled with the open access to data has led to a revolution in the field of Earth observation.
The increasing volume, regular and periodical acquisition and variety of remote sensing data have encouraged the exploitation of satellite imagery with time series approaches and increased expectations to provide monitoring globally and consistently. Time series raised huge challenges in handling datasets and extracting relevant information. Despite the enormous increase of available data, temporal density is still an issue in many spatio-temporal analyses. To overcome the missing data challenges (caused by cloud cover, data availability, etc.) time series are being generated by combining data from two or more sensors. Harmonization of data, bulk processing, and development of advanced raster time series processing is currently a topic of the intense research.
The main objective of the proposed project is to assess and improve the consistency of medium and high-resolution multi-sensor time series, to achieve data harmonisation, to facilitate time series analysis, and to improve the results in selected applications.
The proposed project is organised in four work packages (WP 1-WP 4) that follow a sequence of time series processing, as well as into the work package for project management (WP5):
WP 1: Data collection and analysis of multi-sensor fusion and calibration approaches
WP 2: Generation of long and dense time series
WP 3: Time series analysis
WP 4: Evaluation of the time series processing methodology for selected applications
WP 5: Project management
In WP 1 we will prepare a collection of satellite images (presumably Landsat 8, Sentinel-2, PlanetScope), analyse their characteristics and organize them into a database. Data calibration will be implemented to produce optimal input datasets for time series harmonisation. In WP 2 we will combine time series from different satellite sensors and ensure their radiometric consistency. We will use some existing procedures for harmonization, however the main goal is to develop a generic harmonization procedure. Harmonized data will be analysed in WP 3 to find optimal time series analysis tools considering different applications requirements. In WP4 we will summarize all the analyses and suggest the best practice procedures. All algorithms and programs developed within the project will be open source; with this, we would like to contribute to the wider use of time series in Earth observation.
The obtained methodological knowledge and solutions will be applied in various case studies that are important on global and national levels (e.g. forests stress response, monitoring the effects of urbanization on the natural environment, the potential of farming activities improvements by real-time observation of crop conditions, etc.). Project outcomes will support the needs of national stakeholders we have been already collaborating with. In addition, the research is coherent and directly linked to the activities and goals of the European and international space programmes.
The group of researchers proposing the project includes prominent researchers, young researchers and PhD students, that have innovative ideas, versatile knowledge and programming literacy, and have the necessary hardware, software, and data needed for the successful realization of the project.
Significance for science
Analysis of satellite image time series and derivation of meaningful and reliable products are currently the major research challenges in remote sensing and Earth observation. The substantial increase and availability of historic and new-coming satellite data increasing expectations to provide monitoring globally and consistently. A major factor to ensure monitoring reliability is the combination of multi-sensor observations into a comparable and consistent, long-term observation. Our major contribution will be a study whether simpler generic procedures can be used to solve fine-tuning, temporal radiometric adjustment for multi-sensor data time series generation.
The result of the proposed project will be an open source library (Python, own or/and a contribution to existing ones), which will facilitate the preparation and analysis of different satellite data time series and will support the multi-sensor datasets integration. The library will increase the potential to study changes in the environment, allowing numerous applications of long-term (also multi-decadal) comparison, thereby contributing to global environmental studies (e.g. climate change variables monitoring) and other spatial data (e.g. national spatial data infrastructure).
Innovative contribution of the expected results will thus reflect in:
improvements of approaches for joining time series data derived from different satellite sensors, while ensuring the best radiometric calibration and consistency, and
improvements in the knowledge (practical guidelines) on satellite image time series analyses due to in-depth assessment of the impact of time series data preparation strategy to time series data analysis performance within the scope of different applications.
The research is coherent and directly linked to the activities of the US (NASA) and the European (EU, ESA) space programmes as well as to the guidelines of the Committee on Earth Observation Satellites (CEOS) to increase joint and globally focused activities for the long-term acquisition of consistent and calibrated space-borne observations. The objectives are further consistent with the European guidelines for the establishment of a common spatial data/information infrastructure, including the INSPIRE Directive.
The obtained methodological solutions will be applied to several case studies: forests stress conditions response, monitoring of the effects of urbanization on the natural environment, and potential of farming activities improvements by regular monitoring of crop conditions. Therefore, the research will also have a direct contribution for national stakeholders we are already cooperating with and to the preparation of environmental indicators at the national and international level.
Significance for the country
Analysis of satellite image time series and derivation of meaningful and reliable products are currently the major research challenges in remote sensing and Earth observation. The substantial increase and availability of historic and new-coming satellite data increasing expectations to provide monitoring globally and consistently. A major factor to ensure monitoring reliability is the combination of multi-sensor observations into a comparable and consistent, long-term observation. Our major contribution will be a study whether simpler generic procedures can be used to solve fine-tuning, temporal radiometric adjustment for multi-sensor data time series generation.
The result of the proposed project will be an open source library (Python, own or/and a contribution to existing ones), which will facilitate the preparation and analysis of different satellite data time series and will support the multi-sensor datasets integration. The library will increase the potential to study changes in the environment, allowing numerous applications of long-term (also multi-decadal) comparison, thereby contributing to global environmental studies (e.g. climate change variables monitoring) and other spatial data (e.g. national spatial data infrastructure).
Innovative contribution of the expected results will thus reflect in:
improvements of approaches for joining time series data derived from different satellite sensors, while ensuring the best radiometric calibration and consistency, and
improvements in the knowledge (practical guidelines) on satellite image time series analyses due to in-depth assessment of the impact of time series data preparation strategy to time series data analysis performance within the scope of different applications.
The research is coherent and directly linked to the activities of the US (NASA) and the European (EU, ESA) space programmes as well as to the guidelines of the Committee on Earth Observation Satellites (CEOS) to increase joint and globally focused activities for the long-term acquisition of consistent and calibrated space-borne observations. The objectives are further consistent with the European guidelines for the establishment of a common spatial data/information infrastructure, including the INSPIRE Directive.
The obtained methodological solutions will be applied to several case studies: forests stress conditions response, monitoring of the effects of urbanization on the natural environment, and potential of farming activities improvements by regular monitoring of crop conditions. Therefore, the research will also have a direct contribution for national stakeholders we are already cooperating with and to the preparation of environmental indicators at the national and international level.
Most important scientific results
Interim report
Most important socioeconomically and culturally relevant results
Interim report