Phil Brodrick

Understanding interactions using optimization and statistical learning

Understand interactions using optimization and statistical learning

Earth Systems

Combining data from satellites, aircraft, and fieldwork helps further our basic understanding of ecology and allows for insight into conservation practices.  I work across a large number of scientific fields and study locations, and consequently much of my work in this space is highly collaborative.  My emphasis is on nonlinear statistical models, with particular focus on bridging the gaps between different scales of observations.  Below are some select recent publications.

Uncovering Ecological Patterns with Convolutional Neural Networks

Despite their prevalence in image processing, CNNs have been relatively slow to be adopted in the context of ecology. This, however, is rapidly changing, and in this review we discuss emerging trends, good practices, and what lies ahead for CNNs in ecology.

Forest drought resistance at large geographic scales

As hot droughts become more frequent and more severe, it is important for long‐term forest management to understand which forests are at risk and which have the capacity to withstand these events. The findings in this California-based study are an important first step towards assessing spatially explicit forest stability in the face of a changing climate.

Imaging spectroscopy predicts variable distance decay across contrasting Amazonian tree communities

Accurately characterizing the spatial distribution of species composition in the dense, Amazonian tropics has remained an open challenge. Here, we demonstrate that imaging spectroscopy can be used to map beta diversity, examine spectral similarity over large distances, and test ecological theory of species distributions.

Mapped aboveground carbon stocks to advance forest conservation and recovery in Malaysian Borneo

Forest carbon stocks in Malaysian Borneo are some of the highest observed on the planet.  This mapping effort demonstrates not only the extreme values, but also the relatively high carbon of  logged forests, indicating a significant carbon sink potential with new conservation efforts.

Remotely sensed predictors of conifer tree mortality during severe drought

Changes in canopy water content can be used as an advance indicator of tree mortality.  We explore this relationship in a variety of conifer communities throughout the Sierra Nevada in California.

What mediates tree mortality during drought in the southern Sierra Nevada?

Tree mortality during drought is sensitive to a variety of environmental variables.  This study explores these impacts along an elevation gradient in Sequoia National Park in California.

Scale-dependence of environmental and socioeconomic drivers of albizia invasion in Hawaii

Socioeconomic factors combine with environmental drivers to shape the distribution of albizia, one of the most invasive tree species in Hawaii.

Progressive forest canopy water loss during the 2012–2015 California drought

California underwent a drought of historic proportions between 2012 and 2015, and the effects still resonate today.  This study explores trends in canopy water content in forests throughout the state.