Topic: Advanced GIS spatial analyses in environmental studies

Time: 14:00, April 4, 2012

 Place：Ecology Building, Room 608

Lecturer：Chaosheng Zhang, Ph.D.

Prof. Chaosheng Zhang is head of GIS Centre, Ryan Institute, National University of Ireland, Galway. He teaches Geographic Information System (GIS) and statistics courses at School of Geography and Archaeology of the University. Prof. Zhang’s academic background covers both GIS and environmental geochemistry. His research interest focuses on spatial analyses of environmental variables, especially metals in soils and soil organic carbon, using GIS, geostatistics and other spatial statistical techniques. One of the current research directions of Prof. Zhang is spatial analyses of environment and health. Prof. Zhang has published more than 80 papers in peer-reviewed journals. He is a reviewer for more than 30 international journals.

Prof. Zhang has research experience in China, Sweden, the USA, Australia, Jamaica, the UK, and Ireland. Prof. Zhang is chair and organizer of two internationally leading conferences in environment and health (SEGH 2010 International Conference and Workshops on Environmental Quality and Human Health; SESEH 2012 Sino-European Symposium on Environment and Health).

Prof. Zhang holds the following international professional positions:

Councilor of SEGH (Society for Environmental Geochemistry and Health)

Councilor of IMGA (International Medical Geology Association)

Member of Editorial Board of Science of the Total Environment

Coordinating Editor of Environmental Geochemistry and Health

Abstract:

Environmental databases are being constructed at regional, national and international scales. The analyses of large volumes of environmental data become a challenging task. Spatial analyses provide useful tools for interpretation of environmental data which contain both attribute and spatial information. The topics to be covered include spatial outlier identification, spatial variation and spatial modelling. Outliers in a dataset can cause biased statistical results if they are not properly identified and handled. Spatial outliers are indentified based on a comparison with their neighbouring data, and they may imply a different process from the background such as pollution. Spatial variation has been conventionally evaluated using visual interpretation based on maps, but the development of local statistics enables its quantification spatially. Meanwhile, a geographically weighted regression can be applied to model the spatially varying relationships between environmental parameters, making it possible for the spatial modelling of environmental variables. These issues are demonstrated using environmental geochemical data from Ireland. Furthermore, such techniques can be potentially applied to identify the associations between environment and health which is one of the fastest growing areas.