The Earth System Science research team at CASUS tackles problems at the interface spatial ecology, hydrology, and anthropogenic change to understand the dynamics of biodiversity in the Anthropocene. We combine sophisticated models of spatial ecological processes with multispecies datasets from around the world by leveraging the latest techniques in data science, machine learning, and high-performance computing.

Overview of Earth System Science research

Spatial ecological processes strongly influence the species’ distributions and the biodiversity of ecological communities. These processes are mediated by the geometry of habitats and the movement of animals, both of which can be heavily impacted by anthropogenically driven global change. Teasing apart the intricate, multiscale interactions among these key components requires adopting a systems science perspective. Additionally, the large volumes of complex and often noisy data required to understand spatial ecological processes and their interactions necessitate the latest data science and high-performance computing (HPC) tools. To tackle these challenges, the Earth System Science research team brings together expertise in Spatial Ecology, Animal Movement, Hydrology, Data Science, Geostatistics, Computer Science, Epidemiology.

Research themes for Earth System Science research

Animal Movement Ecology. The movement of animals is a fundamental ecological process with effects that range from driving local population dynamics to shaping biodiversity patterns at continental scales. Moving animals can transport energy, materials, and nutrients across ecosystem boundaries, disperse plant seeds to new environments, and spread diseases over great distances and across species. Animal movement can thus have major effects on the structure, function, and services provided by ecosystems, and is a key, but underappreciated, component of the Earth system. We develop cutting-edge statistical methods and software to facilitate rigorous inferences from growing stockpiles of animal tracking data. In particular, we focus on techniques for quantifying encounters between moving individuals, con- and heterospecific animals, areas or objects of interest (e.g., particular habitat types), and human infrastructure (e.g., roads and vehicles).

Biodiversity in River Networks. Conserving biodiversity amid accelerating global change is a grand challenge in ecology. While biodiversity loss affects many ecosystems, rivers— which harbor a disproportionately large fraction of global biodiversity—are among the most imperiled ecosystems on Earth. Riverine fish biodiversity is influenced by a tapestry of factors including geometric constraints, environmental, anthropogenic effects, and the biological properties of constituent species. We are teasing apart this complexity by combining hydrologically-driven, dendritic neutral biodiversity models with global datasets on river geometry, hydrology, and species occurrence. When combined with forecasted hydrology under climate change scenarios, our models allow large-scale projections of biodiversity loss.

Epidemiology and Optimal Control. Traditional compartment-type epidemiology models are essential tools for understanding and mitigating disease outbreaks. The recent COVID-19 pandemic showcased both the strengths and limitations of this venerable modeling framework. To be better prepared for the next pandemic, we are extending compartment modeling approaches along two axes to facilitate deeper understanding of the controllability of infectious diseases. First, … comparative framework… This generalized “age of infection” model allows us to perform meaningful, side-by-side comparisons across infectious diseases to isolate the factors that govern controllability. Second, we extend traditional epidemiological models to account for behavioral and economic considerations. These coupled models, when combined with multifaceted datasets across nations, then facilitate more balanced insights into the effectiveness of different non-pharmaceutical intervention strategies.