Protecting Coastal Infrastructure in a Changing Climate by Integrating Optimization Modeling and Stakeholder Observations

New York, New York
Submitted by:
Yuki Miura
Columbia University, PhD Student
Project URLs:

Project Description

It is crucial to optimize the protective strategy against storm surges and sea level rise. However, given the limited resources and highly non-linear nature of problems, it is not a trivial problem. We propose a methodology framework of finding an optimal solution to the storm surges and sea level rise with a prescribed budget. The methodology uses 2 different flood estimate tools and stakeholders’ inputs. The flood models are GeoClaw and GIS-based model. GeoClaw will be used to finetune the solution after getting a rough estimate with the GIS-based model. The GIS-based model can return a result of below and above ground within less than a second per storm surge event.

We focus on modeling wall-like barriers as our protecting mechanism of choice for simplicity and interest in their more global effect over the coastal region. Modeling such barriers in the geographical scale of a whole region (e.g. the coasts of New York City) requires much refinement near the barriers’ vicinity. This is due to the difference in scale between the barriers and the region. We resolve this problem and make efficient simulations possible by approximating the barrier as a thin interface and using a special modeling method called the “h-box” method. Such methods ensure both accuracy and time efficiency. Currently, the method is being worked out in two dimensions, which will be sufficient for simulating regional flooding.

The methodology also accounts for stakeholder engagement. Interviews with key stakeholders such as the emergency unit, the mayor’s office, and so forth will be an important component of the methodology design and implementation.

Key Successes

- Using highly accurate, physics-based storm models (GeoClaw) to simulate storms
- Using computationally efficient methods to model barriers against storm
- Using GIS-based (faster/efficient) flooding scenarios
- Using multiple storm scenarios to optimize barrier specifications (location, height, length, timing)
- Stakeholders’ firsthand knowledge and feedback.


- Incorporating the computational methods involved in barrier modeling into the overall storm modeling software
- Optimization scheme programming
Any question regarding this posting, email
The Ocean Visions Founding Members
Ocean Visions, Inc.  - 225 Baker Street, Atlanta, GA 30313, USA  - Founded in 2019