Statistical tools developed by a Mercer University professor and student are providing new insights into climate patterns throughout Earth’s history.
Economics Professor Dr. Szabolcs Blazsek and Raven Dupree, CLAS ’27, began their work in late 2025 and continued this summer with funding from the Mercer Undergraduate Research Scholar (MURS) Training Initiative. It’s the first business project to receive support from the program.
Their first research paper, which they submitted for publication earlier this year and recently resubmitted with revisions, examines the persistence of temperature changes in different regions of the world over the past 175 years. Their second research paper, which is in progress, analyzes climate shifts across distinct periods over the past 67.1 million years.
Dr. Blazsek’s research specializes in statistics and econometrics, the latter of which uses mathematical and statistical models to interpret economic data. In recent years, he has published several papers involving climate data. His methodological focus has been on score-driven time series models, a statistical approach used to analyze sequential data collected over time.
“It’s a specific class of time series models, which can be used to forecast but also to explain basically the evolution of any kind of time series variable. I am contributing methodically to that, so I provide some theoretical contributions and applications,” he said.
Dr. Blazsek asked Dupree, who is pursuing a Bachelor of Business Administration in marketing, to join his research team after having her in his quantitative business analysis course.
“She is one of my best students. Raven is a very hard worker, a very smart student,” he said. “I know that we will publish these two papers. This may help her if she would like to continue with graduate-level studies … and on the job market.”
Dupree has been involved in data analysis and summary, literature reviews, descriptive statistics, coding and editing for the project. She also presented the research during Mercer’s 2026 BEAR Day.
“I enjoy economics, and (we’re) applying econometric concepts to a real issue that I’m also interested in – climate change,” she said. “I like having the chance to delve deep into those things recreationally. Economics isn’t my actual field study, but it’s something that I think is really important and relevant.”
Their first paper studied National Oceanic and Atmospheric Administration temperature data from 1850-2024 for six global regions: the Arctic, Antarctic, Atlantaic Ocean, Northeast Pacific Ocean, Northern Hemisphere and Southern Hemisphere. The data was provided by Luis Gil-Alana from the University of Navarra in Spain, who is the paper’s third co-author.
Dr. Blazsek created a long memory statistical model, which involves estimating how fast the influence of a climate event disappears over time. Whereas short memory models assume that the effects of an event disappear relatively quickly – less than a decade in the case of climate events – and return to normal, long memory models account for more long-lasting impact and sometimes unstable behaviors. Dr. Blazsek’s model goes a step further by tracking how this long-lasting behavior changes dynamically over time.
“The type of model we propose in that paper implies that … any kind of effect on the climate … has influence not only in the next couple of years but it has a longer influence,” he said. “Basically, it was after the Industrial Revolution when carbon dioxide emissions started to increase, and what we show is that, in the past couple of decades, global warming is not only going on but accelerating.”
From the temperature data, he and Dupree verified if there was an acceleration in global warming in each region. They found the strongest acceleration in the Arctic and Northern Hemisphere, but also measurable acceleration for the Northeast Pacific and the Atlantic oceans, Dr. Blazsek said.
“We basically found that, with the temperature all over the world, generally, the common pattern is that it has become more unstable over time,” Dupree said. “In our era of being on Earth, the temperature has kept going up and not been going down at the rate that it normally would, so that’s what we see as that unstable behavior.”
In essence, the effects of climate disruptions or temperature disturbances are lasting longer than they did in the past, Dr. Blazsek said. These findings have significant implications for environmental policies.
“Sometimes environmental policies are only designed for the short term,” Dr. Blazsek said. “They do not consider longer effects. Our paper indicates that the policy makers should be aware that current actions will not disappear in a few years or decades, but they may have several decades of effects. When someone is setting up an environmental policy plan, it is better to set it up for longer, and that is supported by our paper.”
Their second paper applies a similar statistical approach but looks at a robust set of climate data that goes back nearly 70 million years. Alvaro Escribano from Universidad Carlos III de Madrid provided this data and is the third co-author for the paper.
For this data set, deep sea fossils were collected from all over the world, and specific isotopes – or chemical signatures – were extracted from them. One of the extracted isotopes provided information on carbon dioxide concentration and the other provided information on temperature, indicating subperiods of climate change.
“(We) found that climate persistence has shifted across distinct periods in Earth’s history,” Dr. Blazsek said. “Rather than the climate system repeatedly flipping between stable and unstable states, the key changes appear to be in how strongly past climates held onto their patterns. The model also confirmed known major climate shifts in Earth’s history while uncovering additional, finer-scale transitions.”
Dupree said her involvement in this research has been valuable. She has gained a deeper understanding of complex statistical concepts, learned new statistical tools, and improved her writing and editing skills.
“Outside of the technical side, I better understand the writing process for papers of this level,” she said. “Writing a paper for a journal is something that I’ve never actually done until now. Although my field is marketing, it’s knowing that I have that tool under my belt. It makes me feel stronger in my field of business in general.”
