Landscape mapping is crucial for monitoring and understanding a wide range of ecological and social processes, 
as well as for addressing major issues such as land use planning, biodiversity assessment, health-environment 
interactions and monitoring the impacts of environmental and resource management policies. 
In order to make the landscape approach operational, it is necessary to have robust methodological frameworks 
and specific tools for characterising and mapping landscapes at different scales.



In this context, our objective is to develop a map of the major landscapes of the Kerguelen archipelago using 
only satellite time series (MODIS SITS). This work is based on the use of the MOD13Q1 product (NDVI, 250 m) 
covering the period 2003–2022, which is used to identify and characterise the main landscape structures.

The dataset provided includes:

a methodological diagram describing the steps of the approach adopted;
all the data used, structured around four Essential Landscape Variables (ELVs) derived from the NDVI time series;
the final map of the five major types of radiometric landscapes;



This approach illustrates the potential of satellite time series for mapping landscapes in sub-Antarctic 
environments and has resulted in the first radiometric landscape map of the archipelago. It thus provides a 
valuable reference base for future environmental analyses, such as the study of landscape trajectories, 
monitoring vegetation dynamics, and assessing the impacts of climate change.


 This work was supported by the Centre National d’Études Spatiales (CNES) through the TOSCA EVOLKER project. Louise Lemettais PhD grant was funded both by university of Lille and CNES.

QGIS, 3.14.16

ArcGIS, 10.8.1