• Juan M. Restrepo, Mathematics (Statistics and CEOAS) OSU,
  • Ana Spalding, Coastal and Marine Policy (School of Public Policy), OSU,

Topics: Policy analysis, disaster response management, oil spills.


Restrepo has been leading the development of an oil spill transport model, first as a member of the CARTHE Consortia (, and now as part of a team involving scientists at UCLA, LSU, and Oregon State.  The enormous complexity of the task requires the use of advanced computing, data-driven stochastic techniques for dimension reduction, the reduction of model uncertainties via parametrizations, a comprehensive understanding of the policy framework that enables (or hinders) responses, and the socio-cultural context within with disasters are occurring.  At the same time the team is developing strategies for improved landfall determination by incorporating data and models, taking into account their inherent uncertainties.

The NRT project focuses on uncertainty quantification and risk assessments (social, cultural, ecological, and technological), in the realm of evaluating abatement strategies related to ocean oil spills, as well as in the development of novel response strategies to an ongoing oil spill. Specifically, we are developing and testing a response strategy that combines the oil spill model, multiple response strategies/policies, and in-situ measurements as well as human spatial, audio and video reports of a disaster. The novel strategy leads to a significant reduction in the impact of an oil spill, and further, reduces the cost to a community in the recovery process. Our approach combines advanced physics modeling, statistics, public choice modeling, and the capture of dynamic cell phone and passive grid data within the context of socio-cultural characteristics and regional and federal policies around disaster.

Existing Data:

  • The Deep Water Horizon data as well as several other oil spill data.

Data Needs:

  • Response metrics require the inclusion of stakeholder responses as well as the analysis of whether a community is capable of understanding a response strategy and acting on it (i.e. cultural and policy context).
  • We create synthetic data to test scenarios and dimension reduction strategies.
  • We will produce communities via agent-based modeling. Eventually we will link to Transims, the Portland transportation simulator.

Desired Areas of Expertise for Students:

  • Chemical, Physical Oceanography
  • Chemical Engineering
  • Statistics, Mathematics
  • Econometrics, Sociology, Anthropology: decision theory and crowd-derived knowledge.
Print Friendly