Spring 2012 | Robert Morris
When a young Mac Hyman first stepped on Tulane's campus more than 40 years ago, it immediately felt like home. Now — after a career that has put him at the forefront of the world's most pressing scientific problems — Mac has come home again.
"I’ve always sought places where I can have an impact and be engaged in the community. After Katrina, Tulane kept pulling me back to New Orleans." Mac said.
Originally from Lakeland, Fla., Mac was graduating high school in 1968 and considering several southern universities before he visited Tulane for the first time.
"This was the one I didn't want to leave," Mac said. "I felt at home as soon as I walked on campus."
Mac studied math and physics at Tulane; competing with a friend to see how many hours he could pack into a semester. His mentors on the faculty, Jerry Goldstein and Ed Conway, gently advised him to pursue graduate studies at the Courant Institute of Mathematical Sciences at New York University, and from there a summer internship at Los Alamos National Laboratory in New Mexico led naturally into a job offer that became a 35-year career.
"The reason Los Alamos exists is to give scientific guidance on problems of national and international importance," Mac said. "The National Laboratories exist because there are times you need the best interdisciplinary scientific teams possible to spend decades working on a problem."
Los Alamos is best known as the lab where the nuclear bomb was designed, but during the energy crisis of the 1970s, Mac's research was computer modeling to help guide programs for harnessing nuclear fusion power for domestic use. He worked on the issue for 10 years before walking away from it — not because small-scale fusion is impossible, he said, but because he thinks the solutions will remain impractical for sustained power generation.
"I was convinced that fusion was the energy power source of the future. After ten years working on it, I was still convinced that it's the power source of the future, but now believed that it always will be — which means I won't be around to see it," Mac said, laughing. Instead, he wanted to work on a problem with more attainable results.
In the 1980s, a mysterious new epidemic was emerging: HIV/AIDS. All that was known was that it was transmitted sexually, and that it was 100 percent fatal over time. Based on previously known sexually transmitted diseases (STDs) that infected as much as 30 percent of the population at their peak, Hyman became concerned that this new epidemic could pose a major threat to humanity.
"It looked like a third of the planet could be dying in the next couple of decades," Hyman said. "I literally cleaned out my office to learn everything I could about STDs."
Mac joined Stirling Colgate to help guide public health communities in stopping the epidemic. His mathematic models would create the first reliable estimates of how many people were actually infected with HIV and AIDS, and begin offering ideas of how best to combat it. They also discovered why the early epidemic was not growing exponentially, as is typical of newly emerging diseases. The reason was a combination of the wide variation in the number of sexual partners people have, combined with a strong bias towards people selecting partners with similar risky behavior. Their models showed that the time-consuming process of interviewing each patient and everyone they'd been in contact with was a cost-effective way to begin tracking the scope of an early outbreak, but that in populations with more widespread infection such as South Africa, random sampling was more effective.
The team's discoveries have since become common knowledge, but at the time were uncharted territory, and each new development allowed more infected people to be treated, reducing the spread of the disease. “The insights we gain from models are often obvious in hindsight, but don't come to mind when we start the problem," Hyman said.
Fortunately, HIV is much less infectious than syphilis or gonorrhea and the mass fatalities that Hyman had feared did not happen in the United States. However, the HIV incidence is as high as 25% in some southern African communities. Hyman says that we have been lucky, as the government's response to HIV has always been inadequate compared to other emerging epidemics such as SARS or swine flu — diseases he has also worked on. "Our nation has still not responded to AIDS as we should, based on the level of threat," Hyman said.
Epidemiology has continued to be the major focus of Hyman's research since that time, but it is his work on the issues related to modeling itself that he believes will be his real legacy. Mathematicians have worked for centuries to create mathematical equations that preserve the critical properties of the system they model, such as conservation laws, but computers are fundamentally unable to understand them. Where the mathematicians' equations describe a curve, computers can see only a collection of discrete points, and the information lost in that translation from abstract curve to 1's and 0's can become a substantial factor in computer modeling error.
In other words, the inaccurate approximation of a mathematical model on the computer can give as poor a prediction as inaccurate equations describing the underlying process. An inaccurate computer simulation of a physical process can be far worse than no model at all, because it can deceive scientists into believing that they understand a complex system when they don't, leading to unexpected and unwanted consequences. “It is far worse to have a model that is wrong than to not have a model at all,” Hyman said.
Hyman's pioneering research helped create a wave of mathematicians dedicated to creating a discrete calculus that automatically preserves the same properties as the original differential equation models. As part of this research, he has devoted time to improving the tools for quantifying the uncertainty in mathematical model predictions. He uses sensitivity analysis to determine which factors in a model create the most substantial variations in its predictions, so researchers will know which variables need more precise data. For example, sensitivity analysis has shown that the path and strength of a hurricane prediction is very sensitive to the ocean-surface temperature. The predictions about the path of hurricanes have become more accurate, even in the relatively few years since Hurricane Katrina, because scientists have improved the models and data for the ocean-surface temperatures.
Throughout his career, Mac remained in love with New Orleans, returning to visit when he could and stay in touch with friends. After Hurricane Katrina, Mac hosted the Goldstein family (and their dogs) in Los Alamos and started returning regularly to Tulane as a member of Dean Nicholas Altiero’s School of Science and Engineering Advisory Board.
In 2009, when he called home from a Board meeting, Mac's wife, Debby, told him she could hear the passion he had for helping Tulane recover and urged him to consider working here. She knew that he had accomplished what he had set out to do in Los Alamos and that he loved teaching.
Mac took a visiting position at Tulane for the fall semester, and says that he "immediately became engaged with the students and faculty and never looked back. I was home again.”
"I’m delighted he joined the faculty. He brings a wealth of experience in applied mathematics," said Morris Kalka, chair of the mathematics department, noting that he and Hyman attended graduate school together. "It was exceptional for the department to be able to recruit someone of his caliber."
As the Evelyn and John G. Phillips Distinguished Chair in Mathematics, Mac says he sees his new role as much about preparing students for the next step of their careers as it is about teaching or research. Students know, he said, that if they take any of his courses, they will be making presentations, writing papers and doing independent research — all skills that every scientists need in addition to learning the math.
"These students all have the smarts needed to go on," Mac said. "But how do they learn how to pick the right problem to work on? How do we teach them how to write an award-winning proposal? It is our responsibility, as their professors, to teach them these skills."
Further, Hyman says he is captivated by Tulane's community outreach efforts in New Orleans, and hopes to see more innovative contributions from SSE. It's not just teaching math, he says; it's getting people to understand the value of learning it. One of his goals would be to spread the message to young students that math is the language to address the most complex problems facing the world, from fighting emerging epidemics, mitigating climate change, developing clean energy sources, managing water resources, and stabilizing the international financial markets.
“The world needs a new generation of scientists with the best mathematical skills possible to help the world unravel these enormous threats and opportunities,” Hyman said. "Scientists describe the world around us with the language of mathematics. Like all languages, it's best if you start learning it when you're young. If, in our elementary schools, we can convey the impact that mathematics and science can have in solving these world-changing challenges, then we might be able to better entice more students to learn the tools that can change the world.”
Mac's passion for his work is so expansive it may come as quite a surprise to learn about his many, many hobbies. In Los Alamos, he spent up to 50 days a year on the ski slopes, but also had a tradition of picking up a new hobby every year, such as playing piano, clay sculpting, mountain biking, poetry, pastel painting, Afro/Haitian dance and growing exotic fruits.
This year he is focused on learning to play jazz piano by ear. "I can hear when I play a wrong note, but I can't sit down and play without at least the lead line and chord changes for the music," Mac said, then paused. "There are just not enough hours in the day."
School of Science and Engineering, 201 Lindy Boggs Center, New Orleans, LA 70118 504-865-5764 firstname.lastname@example.org