Dataset Information

Mapping land cover on Reunion Island in 2017 using satellite imagery and geospatial ground data.

ABSTRACT: We here present a reference database and three land use maps produced in 2017 over the Reunion island using a machine learning based methodology. These maps are the result of a satellite image analysis performed using the Moringa land cover processing chain developed in our laboratory. The input dataset for map production consists of a single very high spatial resolution Pleiades images, a time series of Sentinel-2 and Landsat-8 images, a Digital Terrain Model (DTM) and the aforementioned reference database. The Moringa chain adopts an object based approach: the Pleiades image provides spatial accuracy with the delineation of land samples via a segmentation process, the time series provides information on landscape and vegetation dynamics, the DTM provides information on topography and the reference database provides annotated samples (6256 polygons) for the supervised classification process and the validation of the results. The three land use maps follow a hierarchical nomenclature ranging from 4 classes for the least detailed level to 34 classes for the most detailed one. The validation of these maps shows a good quality of the results with overall accuracy rates ranging from 86% to 97%. The maps are freely accessible and used by researchers, land managers (State services and local authorities) and also private companies.

SUBMITTER: Dupuy S

PROVIDER: S-EPMC6926296 | biostudies-literature | 2020 Feb

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Mapping land cover on Reunion Island in 2017 using satellite imagery and geospatial ground data.

Dupuy Stéphane S Gaetano Raffaele R Le Mézo Lionel L

Data in brief 20191205

We here present a reference database and three land use maps produced in 2017 over the Reunion island using a machine learning based methodology. These maps are the result of a satellite image analysis performed using the Moringa land cover processing chain developed in our laboratory. The input dataset for map production consists of a single very high spatial resolution Pleiades images, a time series of Sentinel-2 and Landsat-8 images, a Digital Terrain Model (DTM) and the aforementioned refere ...[more]

PMID: 31890790

Similar Datasets

Project description:The Lower Mekong Basin (LMB) is biologically diverse, economically important, and home to about 65 million people. The region has undergone extensive environmental changes since the 1990s due to such factors as agricultural expansion and intensification, deforestation, more river damming, increased urbanization, growing human populations, expansion of industrial forest plantations, plus frequent natural disasters from flooding and drought. The Mekong river is also heavily used for human transportation, fishing, drinking water, and irrigation. This paper discusses use of pre-existing LULC maps from 1997 and 2010 to derive a LMB regional LULC change map for 9 classes per date using GIS overlay techniques. The change map was derived to aid SWAT hydrologic modeling applications in the LMB, given the 2010 map is currently used in multiple LMB SWAT models, whereas the 1997 map was previously used. The 2010 LULC map was constructed from Landsat and MODIS satellite data, while the 1997 map was from before the MODIS era and therefore based on available Landsat data. The 1997-2010 LULC change map showed multiple trends. Permanent agriculture had expanded in certain sub-basins into previously forested areas. Some agricultural areas were converted to industrial forest plantations. Extensive forest changes also occurred in some locations, such as areas changed to shifting cultivation or permanent crops. Also, the 1997 map under classified some urban areas, whereas the 2010 LULC map showed improved identification of such areas. LULC map accuracy were assessed for 213 randomly sampled locations. The 1997 and 2010 LULC maps showed high overall agreements with reference data exceeding 87%. The LULC change map yielded a moderately high level of overall agreement (78%) that improved to ~83% once LULC classification scheme specificity was reduced (forests and agriculture were each mapped as singular classes). The change map regionally showed a 4% decrease in agriculture and a 4 % increase in deciduous and evergreen forests combined, though deforestation hot spot areas also were evident. The project yielded LULC map data sets that are now available for aiding additional studies that assess LMB LULC change and the impacts such change may pose to water, agriculture, forestry, and disaster management efforts. More work is needed to map, quantify and assess LULC change since 2010 and to further update the 2010 LULC map currently used in the LMB SWAT models.

Dataset Information

Mapping land cover on Reunion Island in 2017 using satellite imagery and geospatial ground data.

Publications

Mapping land cover on Reunion Island in 2017 using satellite imagery and geospatial ground data.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets