Integrating high resolution remote sensing and GIS to build affordable data rich predictive models to better assist environmental planning, policy and mitigation.
Room L, Dep.t Ingegneria Civile Edile e Ambientale Via Marzolo 9. Time: 18:00-19:00
16.06.2016
Speaker: prof. John Radke
University of California Berkeley
The recognition of pattern can lead to the discovery of underlying process.
Planning and policy makers rely on the expertise of scientists to fuel their models
of discovery. Although often conceptually sound, limitations in spatial scale often
lead to model uncertainty and error, leading to incorrect results, bad policy and
flawed planning.
Using high-resolution remote sensing imagery we extract detailed, individual-tree
level vegetation cover and height information, and model its view-blocking effect
in an urban landscape. We incorporate these results into a data-rich multivariate
model to measure and predict its impact on a dependent variable, in this instance
crime. Our methods and findings facilitate proactive policy and planning through
successful integration of exogenous factors on one research platform with a GISbased
methodological framework
