Biophysical Characterization of Protected Areas Globally through Optimized Image Segmentation and Classification

Abstract: Protected areas (PAs) need to be assessed systematically according to biodiversity values and threats in order to support decision-making and funding allocation processes. For this assessment, PAs must be characterized according to their species, ecosystems and threats, but such information is often difficult to access and usually not comparable across regions. There are currently over 200,000 PAs in the world and assessing these systematically according to their ecological values remains a huge challenge. Linking remote sensing with ecological modelling can help to overcome some typical limitations of ecological studies, such as sampling bias of biodiversity inventories, and thus to address important conservation challenges. It is the purpose of this paper to introduce eHabitat+, a habitat modelling service supporting the European Commission’s Digital Observatory for Protected Areas (DOPA), and specifically to discuss a specific ecological segmentation tool designed as a component of eHabitat+. This mapping tool systematically stratifies PAs into different habitat functional types based on remote sensing data. It uses a procedure of automatic image segmentation based on several environmental variables to identify the main biophysical gradients in each PA. This allows systematic production of key indicators on PAs that can be compared at the global level. With a few case studies illustrating the ecological segmentation, we discuss the benefits and limitations of the method. eHabitat+ is an example of free and open source software (FOSS) tools for geoscientific tasks.

MARTINEZ LOPEZ Javier;  BERTZKY Bastian;  BONET-GARCÍA Francisco Javier;  BASTIN Lucy;  DUBOIS Gregoire; 

2016-12-20

MDPI AG

JRC101892

2072-4292,   

http://www.mdpi.com/2072-4292/8/9/780,    https://publications.jrc.ec.europa.eu/repository/handle/JRC101892,   

10.3390/rs8090780,