Abstract

Remote sensing offers an unprecedented view of ecological change, with repeated observations spanning individual organisms to entire continents. But what a sensor observes is not the ecological process itself: satellite measurements are noisy, organisms can be missed or misclassified in high-resolution imagery, and these errors can propagate into ecological conclusions. State-space models provide a way forward by explicitly distinguishing variation in the ecological process from variation introduced by how we observe it. I will show how these statistical models can open new opportunities for integrating field observations, remote sensing, and environmental big data.

Speaker Bio

Trevor Caughlin is an Associate Professor at Boise State University. He grew up in the western United States, including Gunnison CO, and worked in the tropics for a bit before returning home to the sagebrush. His research uses statistical models and remote sensing to aid land management decision-making.