A team of researchers from a pair of California-based universities believe they can predict the next influenza outbreak using a mathematical analysis.

Researchers from the University of California and Stanford University have found that applying empirical dynamic modeling techniques to heat and humidity readings over a period of several years can indicate some of the factors that lead to flu outbreaks, according to a phys.org article.

According to the study, it is relatively easy to predict flu season in temperate countries, which are correlated to the advent of winter. However, scientists have struggled to predict flu season in tropical climates with weaker annual climate cycles and the outbreaks are less seasonal and more difficult to explain.

The research group was able to use convergent cross mapping—a test for causality that does not require correlation—to test alternative hypotheses about the global environmental drivers of influenza outbreaks. The team tested four factors over the course of the last 18 years:  absolute humidity, temperature, relative humidity and precipitation.

Using this method, the scientists found that despite differences in outbreak patterns between temperate and tropical locales, absolute humidity and, to a lesser extent, temperature drive flu outbreak on a global level.

With absolute humidity being the most important factor, the study reveals that in more temperate climates with drier air when temperature drops below about 70 degrees F and the humidity level drops, the flu season begins; while in tropical climates flu season actually starts when the humidity and temperature rises.

The study was published in the Proceedings of the National Academy of Sciences.