A Scientist's Life: Julie McClean

Julie McClean is a research oceanographer at Scripps Institution of Oceanography at UC San Diego. She earned a master’s degree in physical oceanography from the University of Sydney in 1987 and then a PhD from Old Dominion University in Virginia in 1993. After receiving her doctorate, she joined the Naval Postgraduate School in Monterey, Calif. as a postdoctoral researcher before becoming research faculty there. She joined Scripps Oceanography in 2005.

explorations now: What do you do for a living?

Julie McClean: I’m an ocean modeler and I simulate the global ocean and the sea ice caps as realistically as possible; for the ocean, this entails simulating relatively small-scale processes such as eddies and frontal structures like you would see in the Gulf Stream. They're important in the global system when we're thinking about climate questions because they transport heat and salt in various directions around the planet.

en: What are some of the major questions in your field?

JM: An overarching question is how to reduce uncertainty in climate model projections. 

Most recently, I have  been working to understand the impact of ice melt from both land ice and sea ice on the ocean. When land ice melts, it adds mass to the ocean leading to sea-level rise. Land ice melt has not been represented in climate models used by the IPCC [Intergovernmental Panel on Climate Change] but that melt, especially in Antarctica, represents the largest portion of uncertainty when it comes to sea-level rise, so its realistic representation in climate models is a first-order problem.

Other missing or under-represented physical processes also contribute to the uncertainty. By simulating the ocean at ten times the horizontal resolution of standard ocean climate models, which have resolutions of nominally one degree latitude and longitude, a broader range of ocean processes are simulated. However, these simulations are much more computationally demanding so they are necessarily shorter, which is also a limitation. There are only so many computer resources available, so how do you use them effectively? 

Some climate modelers might argue that you should use the resources by increasing the capability, and hence realism, of the standard resolution ocean model.  Others would increase the number of ensemble members so that you enhance the strength of the climate signal and reduce noise. Regardless, the overarching goal here is to reduce uncertainty in climate projections, so multiple approaches are valuable.

en: What are some of the tools you use in your research?

JM: I rely on supercomputers to be able to do this work. From agencies such as the Department of Energy (DOE), the Office of Naval Research (ONR), and the National Science Foundation (NSF), you first get funded to do the science and then you need to apply for computer resources to do the simulations. These computing centers have standard allocations that you can use. In my case, I need sustained resources with a large processor count to be able to run the models viably, so I apply for what are called grand challenge resources. If you're successful in being awarded them, that typically gives you both more computing hours and higher queue priority so you can get through your simulations more quickly. I've had a number of grand challenge awards over the years from multiple agencies.

For 2020-21, for example, I received  a DOE Office of Advanced Scientific Computing Research’s Leadership Computing Challenge (ALCC) award that provided priority hours at the National Energy Research Scientific Computing Center. For the Arctic Ocean and sea-ice prediction project funded by ONR, I  also received two  “Pathfinder” computing awards. These awards are tremendously helpful as the simulations are computationally demanding so having priority time makes it much more possible to carry the simulations in a reasonable period of time. 

en: What got you into this field?

JM: A field trip! As an undergraduate, I majored in applied math at Monash University in Australia and then did a second major in marine sciences at Sydney University. At the end of my first term at Sydney, our cohort went up to northern Queensland, a tropical environment, for a ten-day class field trip. The faculty leading the trip had a broad base of expertise in marine science, so we collected data for subsequent laboratory studies in marine geology, bio-sedimentology, and physical and chemical oceanography. We camped and shared meals in the evening after doing all these neat fieldwork experiments. It was wonderful—all of that effort to train us. 

What keeps me going is being able to produce increasingly realistic high-resolution ocean models to both study the dynamics of the ocean and to use them to understand the importance of the ocean to climate. I caught the vision for the use of fine-resolution ocean models from my postdoc advisor and it was he who introduced me to his connections at some of the major national laboratories who were and still are carrying out model developments and computations.  

It can be frustrating work sometimes as there are any number of issues that can come up when setting up and running the models such as in the communication software that is not my area of expertise, but you still have to figure out what's going on to move forward.The supercomputing center can change the available software options when they do an operating system upgrade and everything breaks. That can be really stressful because you've got to get the code working again quickly. You can't just have the computing time and not use it. You have to use it consistently or you lose it. Also, it can take you several years to have something that's ready for analysis. So there are multiple challenges, but to have been able to look at new simulations and see some of these ocean processes and features that are being produced has been exciting.

en: Why did you want to come to Scripps Oceanography?

JM: One of the reasons I wanted to come to Scripps was because of the expertise here in observational oceanography. At Scripps, people are constantly collecting new observational data sets. As part of my work, I use observations to determine the degree of realism of the simulations that I'm producing. By collaborating with the observationalists, often by co-advising students, it's a chance to both validate the models and use the observations and model output together to advance the research.