Digital Governance and Hotspot Geoinformatics of Detection and Prioritization with Environmental Policy Case Studies

• Activity Number: 398
• Type: Invited
• Date/Time: Wednesday, August 5, 2009 : 8:30 AM to 10:20 AM
• Sponsor: Environmental and Ecological Statistics
• Author(s): Sharad Joshi*+ and Ganapati P. Patil and Rainer Bruggemann and Rajesh E. Koli and Sanjay P. Pawde
• Companies: Slippery Rock University and Penn State University and Leibniz Institute of Freshwater Ecology and Inland Fisheries and Moolji Jaitha College and Moolji Jaitha College
• Keywords: hotspot detection ; scan statistic ; progressive upper level set ; genetic algorithm ; partial order ; environmental policy

Abstract

A hotspot is an unusual phenomenon, aberration, elevated cluster, or critical area. Geospatial temporal hotspot detection and prioritization is extremely important for digital governance for monitoring/etiology/management/early warning, etc. At risk are societal issues (e.g., public health, pollution, and national security). We present a new geospatial and spatiotemporal scan statistic, progressive upper level set (PULSE), for hotspot detection. PULSE, a refinement of the upper level set (ULS) scan statistic, carries out a much more exhaustive search for a hotspot using a genetic algorithm in a much larger search space. The algorithm also employs a prioritization technique based on the theory of partially ordered sets to determine 'fitness' of potential hotspots based on their likelihood and compactness. A couple of case studies involving environmental policy decisions are presented.