Data mining to predict climate hotspots: an experiment in aligning federal climate enterprises in the Northwest

Thursday, 18 December 2014
Philip Mote, Oregon State University, Corvallis, OR, United States, Josh George Foster, Oregon State University, CIRC/NW CSC, Corvallis, OR, United States and Steven Brent Daley-Laursen, University of Idaho, Moscow, ID, United States
The Northwest has the nation's strongest geographic, institutional, and scientific alignment between NOAA RISA, DOI Climate Science Center, USDA Climate Hub, and participating universities. Considering each of those institutions' distinct mission, funding structures, governance, stakeholder engagement, methods of priority-setting, and deliverables, it is a challenge to find areas of common interest and ways for these institutions to work together. In view of the rich history of stakeholder engagement and the deep base of previous research on climate change in the region, these institutions are cooperating in developing a regional capacity to mine the vast available data in ways that are mutually beneficial, synergistic, and regionally relevant. Fundamentally, data mining means exploring connections across and within multiple datasets using advanced statistical techniques, development of multidimensional indices, machine learning, and more. The challenge is not just what we do with big datasets, but how we integrate the wide variety and types of data coming out of scenario analyses to create knowledge and inform decision-making. Federal agencies and their partners need to learn integrate big data on climate change and develop useful tools for important stake-holders to assist them in anticipating the main stresses of climate change to their own resources and preparing to abate those stresses.