About Kathryn Roscoe

Dr. Kathryn Roscoe has 15 years’ experience in flood risk management and hydrology. She obtained her MSc in Hydrology at University of California, Davis, and her PhD in probabilistic modeling of levee systems from the Hydraulic Engineering department at Delft University of Technology. At Deltares, she specializes in probabilistic risk analysis, the reliability of flood defense systems, flood damage modeling for vulnerability and risk assessments, and adaptive planning of flood protection systems to changes in climate, policy, and socio-economic development. She has extensive programming experience, and provides quantitative solutions to typically qualitative analyses. She has worked extensively on the development of models and tools for the risk-based assessment of levees, dunes, and hydraulic structures in the Netherlands, and was involved in numerous contributing research projects. Her research has focused on the use of Bayesian Networks to augment more traditional methods of reliability analysis, particularly the ability to utilize and incorporate the experience and observations of experts and stakeholders in the calculation of flooding probabilities.

Her current work focuses on: (1) supporting flood risk management with the dynamic adaptive policy pathways (DAPP) approach, (2) development and improvement of probabilistic methods for risk assessment, and (3) asset management of flood defenses, in which adaptive and risk-based strategies are developed for optimal investment and stakeholder consensus. She was recently involved in the development of adaptation pathways for managing flood risk in the Little River Canal basin in Miami Dade County, and is leading a follow-up county-scale adaptive planning project for Miami Dade County. She is also currently leading the Deltares contribution to an adaptive planning (DAPP) approach for a 4-year project developing a master watershed plan in Calcasieu Parish in Louisiana.

Kathryn is the regional manager for the United States and Canada.

Share this page.