Presentation by Peng Gong, Department of Earth System Science, Tsinghua University.

View the slides.

Abstract: In this presentation, I will report our higher-resolution global land cover mapping efforts beginning in 2010. Our mapping activities are driven by scientific applications.  Initially, a classification system was developed for the purpose of climate modeling needs.  A sampling software was developed to enable efficient sample collection over the world.  With this sample set, we experimented to map the world for 2010, 2015, 2017 using Landsat images.  In 2018, we tested the transferring of 30 meter resolution sample to 10 meter resolution global land cover mapping using Sentinel data.  We find sample collected in a specific year at a certain resolution can be transferable to other years thanks to the robustness of classifiers and large-size and representativeness of the sample.  We also made efforts to map specific land cover classes such as impervious areas, croplands, and water at the global scale.  Different from general-purpose land cover mapping, mapping specific themes has led to more timely-frequent global data sets that enriched the application of global land cover much beyond our original objective.  At the end, I will present our latest research on the use of cloud computing to map seasonal, multiple scheme land cover and land use with a data cube approach.

Looking into the future we are forming a Global Mapping and Application Partnership (GMAP) with collaborators from USGS, Google, Amazon and Shangtang.  In this presentation I shall disclose some of our initial collaborative results through GMAP.

Speaker Bio:  Peng Gong is Professor and Chair of the Department of Earth System Science, Dean of School of Sciences at Tsinghua University (Beijing, China). He had taught at the University of Calgary (Canada) (1991-1994), and the University of California, Berkeley (USA) (1994 – 2015). His major research interests include mapping and monitoring of global environmental change using satellite and ground based sensors, and modelling of environmentally related infectious diseases. He pioneered the monitoring and modeling of urbanization, hyperspectral remote sensing of vegetation, application of wireless sensor network in environmental monitoring and completed the first 30-m and 10-m resolution global land cover map in the world. He also developed and applied regional and global analysis technologies in the modeling of the spatial-temporal transmission of such infectious diseases as schistosomiasis, avian influenza, dengue, malaria, plague and hanta virus.

Lunch will be served.