I.Q. Project Highlight: Predicting Climate Change

How will marine organisms adapt and survive under extreme climate stressors, specifically, rising ocean temperatures and their extremes?

This is an important question motiving ECE Professor Jennifer Dy and her research that would see the integration of statistics (with Prof. Adam Ding), computation, climate science (with Prof. Auroop Ganguly), and marine ecology (with Prof. Tarik Gouhier) to develop a predictive model from nonlinear and relatively non-stationary systems, giving better insight into complicated what-if scenarios.

What if the ocean rises by N degrees? How does that impact sea life?

There’s a problem with how little people currently understand climate change and the effect on the environment when a variable, like the temperature of the ocean, is altered. There is a major knowledge-gap in the ability of a climate change science researcher to develop credible projections of crucial variables and extremes at impact-relevant scales.

To help close that gap, Professor Dy is developing a data-driven solution that works towards creating better projections by proposing novel statistical dependence measures that capture nonlinear dependencies and non-stationary properties common in extremes and spatio-temporal applications.

Marine organisms are sensitive to changes in mean and extreme temperatures at spatial scales of meters and yet the current generation of global climate models can only produce credible insights at 100 kilometers. The focus on extremes and the large volume of data make creating accurate projections particularly challenging.

Professor Dy is working to increase both the accuracy and scope of projections addressing the need for new developments in computational and data-driven representations and models of marine ecosystems, which in turn motivates novel computational and data science methods in spatio-temporal associations and predictive models.

This research will create novel computational methods on feature selection and prediction, enabling the discovery of associations and prediction of climate extremes at finer resolutions, making it relevant for marine ecology survivorship prediction.

Some of the immediate benefits of this research will see the creation of a repository of climate and ecological data along with new and traditional algorithms that will be developed to generate a benchmark for future researchers. Also, data sets and methods developed from the project will be made widely available through open source codes. Despite being developed specifically for climate science and marine ecology, these developments will be beneficial to other application domains as well, such as bioinformatics and medicine where the new algorithms can help discover the features predictive of disease.

Related Faculty: Jennifer Dy

Related Departments:Electrical & Computer Engineering