Environmental Science and Engineering Seminar

Why do trees take certain structural forms, and why are there significant structural differences within and across species? This question has fascinated generations of scientists, including Da Vinci, who observed a pattern in tree branch sizes, and Darwin, who documented the leaf movement on over 200 species. Vegetation structure is a multifaceted concept that includes, for example, the height of a tree, crown size, and the arrangement of leaf and woody parts on the canopy. These are essential metrics for understanding how plants compete for light and withstand environmental stresses. However, we were limited by what we could measure and the details our measurements could provide.

New remote sensing technologies, such as Terrestrial Laser Scanning and ultra-high-resolution remote sensing images, provide new perspectives into how plants work from a single leaf to the ecosystem scales. In this talk, I provide two examples: I start with an example in which Terrestrial Laser Scanning (TLS) provides detailed information on leaf angles of hundreds of thousands of leaves within a few minutes. A fundamental challenge in optimizing photosynthesis is to adjust leaf angles to efficiently use the intercepted sunlight under the constraints of heat stress, water loss, and competition. Despite the importance of leaf angle, until recently, we have lacked data and frameworks to describe and predict leaf angle dynamics and their impacts on leaves. We developed an algorithm that uses TLS data to estimate leaf angles with unprecedented detail. It opens new windows into the rarely-measured intraspecific, interspecific, seasonal, and interannual variations of leaf angles, which have a nontrivial impact on carbon, water, energy fluxes, and many remote sensing observations. In a second example, we move from the leaf scale to the ecosystem scale, where recently available ultra-high-resolution remote sensing data and machine learning techniques provide an opportunity to examine ecological theories and ecosystem resilience at the individual plant's scale. A recent debate is whether big or small trees are more resilient to drought. By mapping over 1.8 million individual trees in the Sierra Nevada during the 2012-2016 California Drought, we show that over 40% of trees were killed during the drought, and tree height is among the most effective predictors of tree mortality – for a severe drought like this one, larger trees are more likely to be killed. We are now using a similar approach to understand the impact of sea-level rise on coastal forests. I conclude by discussion the pros and cons of such types of new remote sensing observations.