Cartography of landscape units made from satellite data in Madagascar between 2016 and 2020

Updated at: 13/10/2022
The first part of the TOSCA project: determining methods to delineate landscapes.

Landscape mapping is crucial for monitoring and understanding a wide range of processes. It is also important for responding to challenges such as land use planning, biodiversity assessment, health-environment interactions, and monitoring/evaluating the impacts of environmental and resource management policies. To develop operational "landscape approach recognition", we need to develop methodological frameworks and tools dedicated to landscape characterization and mapping.

The Landscape Scientific Expertise Centre (SEC) of the Theia Data and Services centre for continental surfaces to which this TOSCA project of the same name is linked aims to develop products and services that make landscape analysis robust and transferable within a variety of fields. This requires human activity in given environments to be taken into account by working with managers who already use this integrating object for land planning. This project consists of developing an automated method to delimitat Landscape Units (LU) from Essential Landscape Variables (ELV) on the basis of Copernicus data.

 

This first part of the TOSCA project aims to determine methods to delimit landscapes. A first method is proposed in this dataset. You can find the following elements :

  • a general schema summarizing the method ;
  • the data set used with a MODIS NDVI time series (MOD13Q1 product during the period 2016-2020) and Sentinel-2 data (2019);
  • the final landscape unit cartography created using the method on the selected test area;
  • a scientific poster presented at the INTECOL-2022 conference (see related publication).

 

 

Lemettais, Louise; Bégué, Agnès; Luque, Sandra; Laques, Anne-Elisabeth; Alleaume, Samuel, 2022, "Cartographie des unités paysagères à partir de données satellitaires à Madagascar entre 2016 et 2020", https://doi.org/10.23708/LBGQQU, DataSuds, V1.