LAUSR

Abstract Human-mediated climate change over the past century has resulted in significant impacts to global ecosystems and biodiversity including accelerating redistribution of the geographic ranges of species. In mountainous regions, the transition zone from continuous closed-canopy montane forests to treeless alpine tundra areas at higher elevations is commonly referred to as the “alpine treeline ecotone” (ATE). Globally, warming climate is expected to drive ATEs upslope, which could lead to negative impacts on local biodiversity and changes in ecosystem function. However, existing studies rely primarily on field-based data which are difficult and time consuming to collect. In this study, we define an ATE-detection index (ATEI) to automatically identify the ATE positions from 2009 to 2011 in the western United States using geospatial tools and remotely sensed datasets provided by Google Earth Engine. A binomial logistic regression model was fitted between standardized ATEI components and a binary variable of pixel status of 141 sampled Landsat pixels manually classified with high-resolution imagery in Google Earth. The average model accuracy was around 0.713 (±0.111) and the average Kappa coefficient was approximately 0.426 (±0.221) based on a 100-time repeated 10-fold cross-validation. Furthermore, the ATEI-estimated elevation is highly correlated (Pearson's r = 0.98) with a published set of field-collected ATE elevations at 22 sampling sites across the region. The detection metric developed in this study facilitates monitoring the geographic location and potential shifts of ATEs as well as the general impact of climate change in mountainous areas during recent decades. We also expect this approach to be useful in monitoring other ecosystem boundaries.