Jordan Landers

Biography

I am a PhD candidate in Earth Sciences at the University of Southern California, advised by Julien Emile-Geay. My background spans mathematics, physical oceanography, and data science, with research stints at Woods Hole Oceanographic Institution and several years in industry as a data scientist before returning to graduate school. My work sits at the intersection of paleoclimate and nonlinear time series analysis, with a recurring focus on the question of whether apparent relationships in the climate record reflect genuine dynamical influence.

Research Interests

The central problem motivating my research is causality in nonlinear systems. In Earth’s climate, correlation is not only an unreliable guide to causal structure — it can be actively misleading: two causally linked variables can exhibit intervals of positive, negative, or no correlation depending on system state, while two unrelated variables can appear coherent because of a shared driver. Standard associative approaches, including spectral analysis, cannot resolve this ambiguity.

I use Convergent Cross Mapping (CCM) to approach these questions more carefully. CCM tests whether the state space of one variable carries recoverable information about another — a condition that, under the theory of dynamical systems, is only satisfied in the presence of genuine coupling. My dissertation applies this framework to questions in Holocene and Pleistocene paleoclimate.