LULC - Special topic

Land Use and Land Cover I

6107.2 - Synergy between Sentinel-1 radar time series and Sentinel-2 optical for the mapping of restored areas in Danube delta

Thursday, July 6
8:50 AM - 9:10 AM
Location: Maryland B

Wetlands are important and valuable ecosystems, yet, since 1900, more than 50% of wetlands have been lost worldwide. An example of altered and partially restored coastal wetlands is the Danube Delta in Romania. Over time, human intervention has manifested itself in more than a quarter of the entire Danube surface. This intervention was brutal and has rendered ecosystem restoration very difficult. Remote sensing offers accurate methods for detecting and mapping change in restored wetlands. Vegetation change detection is a powerful indicator of restoration success. The restoration projects use vegetative cover as an important indicator of restoration success. To follow the evolution of the vegetation cover of the restored areas, satellite images radar and optical of last generation have been used, such as Sentinel-1 and Sentinel-2. Indeed the sensor sensitivity to the landscape depends on the wavelength whatever radar or optical data and their polarization for radar data. Combining this kind of data is particularly relevant for the classification of wetland vegetation, which are associated with the density and size of the vegetation. In addition, the high temporal acquisition frequency of Sentinel-1 which are not sensitive to cloud cover allow to use temporal signature of the different land cover. Thus we analyse the polarimetric and temporal signature of Sentinel-1 data in order to better understand the signature of the different study classes. In a second phase, we performed classifications based on the Random Forest supervised classification algorithm involving the entire Sentinel-1 time series, then starting from a Sentinel-2 collection and finally involving combinations of Sentinel-1 and -2 data. The supervised classifier used is the Random Forest algorithm available in the OrfeoToolbox (version 5.6) free software. Random Forest is an ensemble learning technique, and builds upon multiple decision trees particularly relevant when combining different kind of indicators. The results of this study relate to combinations of data from different satellite sensors (multi-date Sentinel-1, Sentinel-2) in order to improve the accuracy of recognition and mapping of major vegetation classes in the restoring areas in Danube Delta. First, the data from each sensor are classified and analyzed. Results show quite good classification performance for only one Sentinel-2 data (87.5% mean accuracy) contrary to the very good results obtained using the Sentinel-1 time series (95.7% mean accuracy) in this first step. The combination of Sentinel-1 time series and optical data Sentinel-2 improved the performance of the classification (97.1%). The very good indices kappa have been obtained: for the multi-data radar was Kappa index 0.96 and for multi-sensor data integration was Kappa index 0.97. The reliable Producers Accuracy and K coefficient results prove the complementarity of the two satellites for the observation, analysis and spatial representation of the deltaic plant ecosystems. The Producers Accuracy analysis by class shows that the Sentinel-2 optical sensor has its limits as concerns the detection of similar plan classes, like for example the different classes of reed. It detects these classes but the mapping precision is not always great (on some occasions, it is about 55% for the Reed on compact plaur class). By contrast, the use of a Sentinel-1 time series shows how interesting the C band radar time signature is in the Danube Delta ecosystem. Information from different sensors may assist in the variable retrieval by limiting potential ambiguities. The temporal resolution of optical sensor Sentinel-2 does not allow having temporally frequent products of vegetation characteristics due to the cloud coverage. A multi-temporal radar, multi-sensor approach applied to a temporal sequence of data acquired by different sensors can improve mapping and monitoring of vegetation state variables over time.

Niculescu Simona

Dr HDR
Institut Universitaire Européen de la Mer

Dr. Simona Niculescu received the Ph.D. degree in Geography from the Université de Paris IV-Sorbonne (France) in 2002. Since 2005 she is with the LETG-Géomer UMR 6554 CNRS lab within the European Institute for Marine Studies in Brest. The researcher’s current research topic deals with coastal areas, and hence deltas. She focuses on methodologies for exploiting the opportunities offered by the new satellite images.

Presentation(s):

Send Email for Niculescu Simona

Ammatzia Peled

Prof.
University of Haifa

Dr. Ammatzia Peled is a professor for GIS&RS at the University of Haifa, Israel. Ammatzia Served as the Chair of the ICA Commission on Incremental Updating and Versioning of spatial Databases. He served also as ISPRS Treasurer and Second Vice President and as ISPRS president of Commission VIII. In 2010, Prof. Peled was awarded the Eduard Tsiolkovsky Memorial GOLD Medal by the Russian Academy for Cosmonautics for “Outstanding contribution to Cosmonautics”. In 2013 he was awarded as a Professor Honoris Causa by the Siberia State Academy for Geodesy & Cartography and last July he was awarded as ISPRS Fellow.

Presentation(s):

Send Email for Ammatzia Peled


Assets

6107.2 - Synergy between Sentinel-1 radar time series and Sentinel-2 optical for the mapping of restored areas in Danube delta

Slides Handout

Attendees who have favorited this

Please enter your access key

The asset you are trying to access is locked. Please enter your access key to unlock.

Send Email for Synergy between Sentinel-1 radar time series and Sentinel-2 optical for the mapping of restored areas in Danube delta