Gilberto Câmara: Big geospatial data analytics for global land change monitoring

CamaraLand use change is one of the most immediate consequences of humanity’s capability to transform the Earth’s ecosystems and landscapes to be able to feed 8.5 billion people. Informed public policies need objective appraisal of human actions on the land system such as deforestation, pasture intensification, and food replacement by biofuels. Geoinformatics methods and techniques are essential to support global land research, especially tools that deal with Earth Observation images, the only source that provides a continuous and consistent set of information about the Earth’s land masses. Fortunately, many Earth Observation images are now available publicly, making an unprecedented large amount of data available for research and operational usage. To make good use of these archives, we need methods to analyze big EO data sets. This talk will discuss the theoretical and practical challenges of an IT infrastructure capable of handling multitemporal, multisatellite data sets and analysing such data effectively. We will also present some results from an on-going project at IFGI that uses the SciDB open source data management and analytics coupled with the R open source statistical environment for doing big data analytics for Earth Observation data.

CV: Gilberto Câmara is a researcher on Geoinformatics at Brazil's National Institute for Space Research (INPE), where he was General Director (2006-2012). Under his leadership, INPE made major advances in forest monitoring by satellites. He is the currently Brazil Chair at the University of Münster (WWU) in Germany (2013-2015) and is a Guest Professor at WWU’s Institute for Geoinformatics. Gilberto advised 22 PhD dissertations and published 150 papers that have been cited more than 5500 times. He received a Dr. Honoris Causa from the University of Muenster (Germany), the Global Citizen Award from the Global Spatial Data Infrastructure Association, and the Pecora Award from USGS and NASA for "leadership to the broad and open access to remote sensing data".