A new digital information layer describing global human settlements from 40-years Landsat satellite imagery

Global human settlement information is required by a number of institutions operating globally, and also it will be essential for developing indicators, for international framework agreements including the Sendai framework for Disaster Risk Reduction or Sustainable Development Goals. These indicators should be action oriented, global in nature and universally applicable. Human settlements can be mapped with Remote Sensing data which are independent, globally-consistent, updated regularly, providing a synoptic overview and can be considered objective. Regular provision of remote sensing data may be one of the few ways to gather standardized information globally. In this context, a new global information baseline describing the spatial evolution of the human settlements in the past 40 years is presented. It is the most spatially detailed global data on human settlements available today, also with the greatest temporal depth. The core processing methodology relies on a new supervised classification paradigm based on symbolic machine learning, which has been developed for its application in big remote sensing data scenarios. The information is extracted from Landsat image records organized in four collections corresponding to the epochs 1975, 1990, 2000, and 2014. The method described here is the first known attempt to exploit global Multispectral Scanner data for historical land cover assessment. The Landsat-made Global Human Settlement Layer (GHSL) reports about the presence of built-up areas in the different epochs at the spatial resolution allowed by the Landsat sensor. The performed tests show that the quality of the information on built-up areas delivered by GHSL is an improvement over that of other global information layers extracted by automatic processing from Earth Observation data.

PESARESI Martino;  EHRLICH Daniele;  FLORCZYK Aneta;  CARNEIRO FREIRE Sergio Manuel;  JULEA Andreea Maria;  KEMPER Thomas;  SOILLE Pierre;  SYRRIS Vasileios; 

2016-10-17

International Society for Digital Earth

JRC101305

https://publications.jrc.ec.europa.eu/repository/handle/JRC101305,