514
entire Sentinel-1 time series, proceeding through a Sentinel-2 collection and finally
involving combinations of Sentinel-1 and -2 data. The supervised classifier used is
the Random Forest algorithm that is available in the OrfeoToolbox (version 5.6) free
software. Random Forest is an ensemble learning technique that builds upon multiple decision trees and is particularly relevant when combining different types of
indicators. The results of this study relate to the use of combinations of data from
different satellite sensors (multi-date Sentinel-1, Sentinel-2) to improve the accuracy of recognition and mapping of major vegetation classes in the restoring areas
of the Danube Delta. First, the data from each sensor are classified and analyzed.
The results obtained in the first step show quite good classification performance for
only one Sentinel-2 data (87.5% mean accuracy), in contrast to the very good results
obtained using the Sentinel-1 time series (95.7% mean accuracy). The combination
of Sentinel-1 time series and optical data from Sentinel-2 improved the performance
of the classification (97.1%).
Keywords Coastal wetlands • Danube delta • Alteration and remediation of ecosystems • Remote sensing • Synergy of radar time series Sentinel-1 and optical
image Sentinel-2
17.1 Introduction
According to United Nations Environment Programme research, 40% of the global
economy depends on the adequate functioning of ecosystems. Many ecosystems
have been sacrificed in the name of economic development in various fields such as
agriculture and other production systems and as a result of urbanization, industrialization, resource extraction, transportation and other infrastructures. Given the
increasing number of anthropogenic perturbations of natural and semi-natural ecosystems worldwide, the mere preservation of these ecosystems is no longer sufficient. Ecological restoration may thus prove to be an essential complement to their
preservation. The dramatic depletion of biological diversity and the degradation of
ecosystem services allowed legitimizing a marked evolution of the international,
European and Romanian political context. As a result, the preservation and restoration of natural ecosystems has been recognized to be essential and constitutes part
of Romania’s future political commitments, more precisely of the National
Biodiversity Strategy 2013–2020.
Over the past few decades, efforts to raise awareness of the environmental, functional and patrimonial interest of natural areas, as well as the evolution of protection
policies, with a view in particular to the implementation of the “Habitats-FaunaFlora” European Directive of 1992, have led to the carrying out of numerous ecological restoration operations. Ecosystem restoration is becoming a global priority
at various levels of decision-making, with the goal of achieving specific political
and technical objectives (Aronson and Alexander 2013).
S. Niculescu et al.
entire Sentinel-1 time series, proceeding through a Sentinel-2 collection and finally
involving combinations of Sentinel-1 and -2 data. The supervised classifier used is
the Random Forest algorithm that is available in the OrfeoToolbox (version 5.6) free
software. Random Forest is an ensemble learning technique that builds upon multiple decision trees and is particularly relevant when combining different types of
indicators. The results of this study relate to the use of combinations of data from
different satellite sensors (multi-date Sentinel-1, Sentinel-2) to improve the accuracy of recognition and mapping of major vegetation classes in the restoring areas
of the Danube Delta. First, the data from each sensor are classified and analyzed.
The results obtained in the first step show quite good classification performance for
only one Sentinel-2 data (87.5% mean accuracy), in contrast to the very good results
obtained using the Sentinel-1 time series (95.7% mean accuracy). The combination
of Sentinel-1 time series and optical data from Sentinel-2 improved the performance
of the classification (97.1%).
Keywords Coastal wetlands • Danube delta • Alteration and remediation of ecosystems • Remote sensing • Synergy of radar time series Sentinel-1 and optical
image Sentinel-2
17.1 Introduction
According to United Nations Environment Programme research, 40% of the global
economy depends on the adequate functioning of ecosystems. Many ecosystems
have been sacrificed in the name of economic development in various fields such as
agriculture and other production systems and as a result of urbanization, industrialization, resource extraction, transportation and other infrastructures. Given the
increasing number of anthropogenic perturbations of natural and semi-natural ecosystems worldwide, the mere preservation of these ecosystems is no longer sufficient. Ecological restoration may thus prove to be an essential complement to their
preservation. The dramatic depletion of biological diversity and the degradation of
ecosystem services allowed legitimizing a marked evolution of the international,
European and Romanian political context. As a result, the preservation and restoration of natural ecosystems has been recognized to be essential and constitutes part
of Romania’s future political commitments, more precisely of the National
Biodiversity Strategy 2013–2020.
Over the past few decades, efforts to raise awareness of the environmental, functional and patrimonial interest of natural areas, as well as the evolution of protection
policies, with a view in particular to the implementation of the “Habitats-FaunaFlora” European Directive of 1992, have led to the carrying out of numerous ecological restoration operations. Ecosystem restoration is becoming a global priority
at various levels of decision-making, with the goal of achieving specific political
and technical objectives (Aronson and Alexander 2013).
S. Niculescu et al.