Drones in the Service of Our Forests

If you've spent any time alive on this planet, you've probably noticed that things here work in cycles – the sun passes overhead daily, we complete one revolution around the sun every year, and the slight tilt of our planet's axis subsequently yields seasonal changes in time with these astronomical movements. The lives of plants here on earth are guided by these cycles. In deciduous forests like the ones here in Central Massachusetts, trees awaken from a long period of dormancy in the spring to push out the leaves that will allow them to conduct photosynthesis in the summer. The leaves change color and fall from branches in autumn, leaving trees lying dormant again for another winter. The study of these events in plant life cycles is a subject known as plant phenology.

We know today that climate change is making a significant impact on these cycles. The timing of the emergence of leaves in the spring and their abscission (or detachment) in the fall is dependent on patterns of temperature and precipitation that are changing as a result of increased global average temperatures. The impact here in our forests is that spring bud-break and fall abscission are happening earlier and later in the year, extending the annual forest growing season.

We want to keep track of changes in the length of forest growing seasons to understand how critical forest ecosystem services like carbon sequestration and water filtration are being impacted. But how do we do this? Well, there are many methods to choose from. People have kept track of the timing of phenological events for thousands of years through records of direct visual observations. Supplementing those records in more recent years are digital photographical records made by things like stationary cameras directed at the canopy and by satellites orbiting the Earth. The most recent addition to our phenological monitoring toolbox, however, is a kind of technology known as a UAV. UAV stands for unmanned aerial vehicle, but most people know UAVs colloquially as drones.

This summer, I work with my mentor to use drones in taking time series photography of the forest canopy. We mount a digital camera to the bottom of a drone and send it on a pre-programmed flight path over the trees. It covers an area equivalent to one satellite image pixel, and after conducting several of these flights over time, we can put together a timeline of what the canopy in this region looks like during seasonal transitions.

This time series imagery is no good, however, if we don't know exactly what we're monitoring in it. That's where my independent project comes in. The goal of my work is to nail down just what part of leaf physiology the greenness of our time series imagery correlates to. To do this, I will predict dates for the beginning, middle and end of spring based on the greenness of my time series imagery, and do the same with leaf development data collected by someone who has directly observed the forest for years. I'll then compare the two sets of dates, and hopefully figure out what metric for leaf development the drone imagery tracks most accurately! Once we know this, we'll be better able to employ drones as ecological monitoring tools, and in turn, be better equipped to track changes in our forests.