Kevin Winner

The paper compiles a large dataset of biological activity from US weather radar data spanning 1995 to 2025, enabling efficient aeroecology research.

This study evaluates deep learning models for species distribution prediction, finding they match but do not surpass traditional methods, with performance issues for threatened species and data leakage concerns.

This paper introduces a framework to derive population and organismal-level ecological niches from individual-level niches, showing how higher-level niche properties can be partitioned into individual contributions and how species-level niche shifts can be estimated from individual responses.