We utilize various types of geospatial data (e.g. remote sensing images, social media data and mobility data) and machine learning techniques to detect human dynamics in disasters at multiple scales. We develop quantitative models to assess disaster resilience and explain socio-economic factors that affect the resilience. A suite of modeling frameworks and tools are developed to assess community and infrastructure resilience in various disasters and at multiple spatio-temporal scales.
This research aims to address the long-standing challenges of multi-scale spatio-temporal analysis in GIS. This research subverts the traditional views of space (flat layers) and time (linear intervals) and create new data models and analytical tools based on the Triangle and Pyramid Model. Our goal is developing a unified modeling framework to measures patterns and relationships across spatial and temporal scales.
Learn more about the NSF-funded CroScalar project.
We utilize GIS and spatial analysis to study environmental injustice and social disparities related to disaster risk and urban planning. We conducted national assessments of population and infrastructure exposure to flooding hazards. We also applied GIS to evaluate the inequalities in the exposure to scenic landscapes. The goal of these studies are to detect population groups that are disproportionally exposed to flooding hazards. Our studies revealled the systematical inequalities and injustice associated with hazard exposure and share of environmental resources.
Interactions between the systems are typically dynamic, non-linear and nested. Understanding the dynamics of coupled natural and human (CNH) systems is critical for evaluating resilience and sustainability of social and eco-systems. To this end, we apply artificial intelligence (AI) and geosimulation to model dynamic interactions in CNH systems. Such models have been used to predict land cover change, population movement and eco-system degradation in scenarios of natural disasters and climate change.