Speaker: Jie Tian, Assistant Professor in GIS at Clark University
Title: Modeling the Environment Using Modern Geospatial Approaches (75 MB mp4)
Abstract: Accurate information on the spatial-temporal distributions of environmental variables (e.g., air pollution, soil moisture) at a regional scale is crucial for understanding environmental processes, as well as to impact studies on, for example, public health. Presently, many environmental variables are observed by both modern remote sensing and ground monitoring networks. However, the synthetic use of environmental information from these two sources poses a fundamental challenge because remotely sensed data differs very much from ground data in terms of spatial resolution, accuracy, coverage, and temporal frequency, due to the distinct means of their acquisition. It is therefore very important, in environmental monitoring and modeling, to be able to effectively integrate these two types of data for a maximum and appropriate use of the available information. This presentation will demonstrate an innovative research framework that can complement the advantages of remotely sensed data and ground-based data in modeling the spatial-temporal dynamics of an environmental variable. The Bayesian Maximum Entropy approach is used to process the “hard data” measured on the ground and the “soft data” estimated from remote sensing in a statistical and rigorous manner. The working research procedure will be illustrated by a case study of mapping air pollution at a regional scale using MODIS aerosol optical depth data and ground-based measurements of PM2.5 concentration. The presented research framework is believed to have a general value in research and practices of modeling various environmental variables, where the integration of remotely sensed information and ground survey data is a necessity.
no links
no file attachments