Using multi-scale remote sensing to study the spatial dynamics of piñon-juniper woodland biomass

Piñon-juniper (PJ) woodlands are the most widespread dryland woodland ecosystem in the United States, with a range spanning one-quarter of the conterminous US. Throughout this range, codominant piñon and juniper trees represent a major source of biomass in the semiarid American Southwest. This biomass varies greatly over space, driven by environmental gradients, and time, driven by fluctuations in recent historical climatic conditions. To precisely understand PJ's role in the global carbon cycle requires an accurate, spatially-explicit accounting of current aboveground biomass - a task ideally-suited for remote sensing. However, mapping biomass in dryland woodland ecosystems, with their short trees and open canopies, poses challenges not faced in other forested environments and requires tailored analytical approaches. In this study, we present one such approach, linking plot-level estimates of biomass to the local scale using airborne lidar, and linking lidar-level estimates of biomass to the regional scale using Landsat. We demonstrate how this multi-scale analysis is ideally-suited to the unique structural characteristics of PJ woodlands, and provide an analytical framework for mapping biomass in similarly-structured ecosystems around the globe. We compare mapped biomass estimates to a range of abiotic environmental factors (e.g. climate, terrain, soil variables) to identify the ecological conditions that promote or hinder PJ biomass with the intent of better understanding how this widespread and important, yet threatened ecosystem will respond to future changes in climate.