In an era of massive biodiversity loss, the greatest conservation success story has been the growth of protected land globally. Protected areas are the primary defense against biodiversity loss, but extensive human activity within their boundaries can undermine this. Using the most comprehensive global map of human pressure, we show that 6 million square kilometers (32.8%) of protected land is under intense human pressure. For protected areas designated before the Convention on Biological Diversity was ratified in 1992, 55% have since experienced human pressure increases. These increases were lowest in large, strict protected areas, showing that they are potentially effective, at least in some nations. Transparent reporting on human pressure within protected areas is now critical, as are global targets aimed at efforts required to halt biodiversity loss. One-third of global protected land is under intense human pressure Kendall R. Jones1,2,*, Oscar Venter3, Richard A. Fuller2,4, James R. Allan1,2, Sean L. Maxwell1,2, Pablo Jose Negret1,2, James E. M. Watson1,2,5 See all authors and affiliations Science 18 May 2018: Vol. 360, Issue 6390, pp. 788-791 DOI: 10.1126/science.aap9565

The Global Artificial Land Surface in 30 meters resolution (GlobeLand30-ATS2010 for short) was developed based on the data mining methodology by integrating and analyzing the 9907 scenes of the USA Landsat TM5, ETM+ data and 2640 scenes of the China environment disaster mitigation satellite (HJ-1) data in 2010 （±1）. Since the Artificial Land Surface is mostly a mosaic of, for example, buildings, trees, roads, small-water bodies, and grasslands that are frequently combined, it makes data mining for identifying the artificial land surface more difficult. The Pixel-Object-Knowledge (POK）methodology was applied in this study and data development. The 30m dataset shows where and how many residents there are in cities and villages, as well as industrial lands, airports, and roads worldwide. Data citation: CHEN Jun et al. : 2014.Global Artificial Land Surface Dataset in 30m Resolution (2010) ( GlobeLand30_ATS2010 ) ,Global Change Research Data Publishing & repository, DOI:10.3974/geodb.2014.02.02.V1 Available at: http://www.geodoi.ac.cn/WebEn/doi.aspx?Id=163

2010 estimates of total number of people per grid square across Africa South America and Asia, with national totals adjusted to match UN population division estimates, 2012 revision (http://esa.un.org/wpp/)

The GHSL Landsat is a spatial raster dataset that is mapping human settlements globally based on the Landsat satellite data collection. The GHSL Landsat uses the Global Land Survey (GLS) collection of Landsat imagery, which is a carefully coordinated collection of high resolution imagery for global modelling and is produced by the Global Land Cover Facility (www.landcover.org). This allows the mapping of settlements back in time until the year 1975. In addition, Landsat GHSL uses recent Landsat-8 from 2013/2014 for the latest coverage. The aggregated set has the coordinate Reference System: Spherical Mercator (EPSG:3857) and the spatial resolution of 38 m. This data is provided as single GEOTIFF file.

The Global Rural-Urban Mapping Project, Version 1 (GRUMPv1) consists of estimates of human population for the years 1990, 1995, and 2000 by 30 arc-second (1km) grid cells and associated data sets dated circa 2000. A proportional allocation gridding algorithm, utilizing more than 1,000,000 national and sub-national geographic units, is used to assign population values (counts, in persons) to grid cells. This data set is produced by the Columbia University Center for International Earth Science Information Network (CIESIN) in collaboration with the International Food Policy Research Institute (IFPRI), The World Bank, and Centro Internacional de Agricultura Tropical (CIAT).

The GHSL Landsat is a spatial raster dataset that is mapping human settlements globally based on the Landsat satellite data collection. The GHSL Landsat uses the Global Land Survey (GLS) collection of Landsat imagery, which is a carefully coordinated collection of high resolution imagery for global modelling and is produced by the Global Land Cover Facility (www.landcover.org). This allows the mapping of settlements back in time until the year 1975. In addition, Landsat GHSL uses recent Landsat-8 from 2013/2014 for the latest coverage. GHS BUILT-UP GRID These data contain a multitemporal information layer on built-up presence as derived from Landsat image collections (GLS1975, GLS1990, GLS2000, and ad-hoc Landsat 8 collection 2013/2014). The data have been produced by means of Global Human Settlement Layer methodology in 2015. 250m of resolution - World Mollweide (EPSG:54009)

World Urban Areas represents the major urban areas of the world as shaded polygons.

2010 estimates of total number of people per grid square across Africa South America and Asia, with national totals adjusted to match UN population division estimates, 2012 revision (http://esa.un.org/wpp/)

A global map of built-up presence derived from backscattered information of Sentinel1 images. Both the GHS BUILT-UP GRID (LDS) as derived from Landsat image collections and the GlobeLand30 (GLC30) were used for training of the Symbolic Machine Learning (SML) classifier. 20m of resolution - Spherical Mercator (EPSG:3857) Dataset name (size): GHS_BUILT_S12016NODSM_GLOBE_R2016A_3857_20 Legend: 0 = no built-up 1 = built-up

The developed approach outputs a global raster layer representing both the spatial distribution and density of built-up areas, for the year 2010. The information about the presence of built-up is expressed as the percentage of built-up area respect to the total surface of the cell. Values are expressed in the range [0 to 100]. The layer is made available as a grid having a spatial resolution of 30-arc seconds (approximately 1 km at the equator), in the WGS84 coordinate system. Being available as a quantitative, continuous raster dataset significantly increases its value by facilitating integration with other spatial datasets for analysis or modeling The method uses machine learning techniques to understand the best population thresholds translating population densities to built-up densities. In the proposed methodology the MODIS Urban Land Cover (ULC) 500 m (C5) made by satellite data of the year circa 2001-2002 is used as training set for classification of the LandScan 2010 Global Population Database (LS). Similar techniques are described in Pesaresi et al. (2013) and Gueguen (2014) for the purpose of finding best rescaling parameters translating remote sensing image-derived features to a high-level-abstraction semantic as “built-up areas”.