Science Story – May 2019

Breckheimer and Drone

Where all the flowers have gone

If you are old enough to have been lost, you can appreciate how quickly mapping technology is changing field science. It wasn’t that long ago that my trips involved stacks of maps and my shaky sense of location. Now, a swipe of my smartphone keeps my trip stress free.

It is easy to forget how amazing it is. I’ve got a pocket computer that is more powerful than the computers that took astronauts to the moon. Not only is it connected to the larger world, but it triangulates using cell phone towers and satellites to know my precise location. Big data crunching algorithms give me an estimate of arrival time. How can this not change field science? Location is the glue that holds field science together!

The mapping revolution is compounded by rapid changes in technology. Last summer a plane captured the intensities of 420 different bands of visible and non-visible light for every square yard in the valleys around Gothic. This data will be used to identify the species of plants (at least large plants like shrubs and trees) as well as detect pollutants in the soil and leaves of plants.

Throw a little machine learning in with the mapping revolution and emerging technology, and things really take off! Many scientists care about when their study sites melt out. When bare ground emerges sets the start of the growing season for many plants and animals. A different Ian, Dr. Ian Breckheimer (Harvard University), is using machine learning to integrate satellite and drone data, to estimate just that.

Ian’s summer drone work drives home the value of integrating RMBL’s past research and new technology. For 45+ years Dr. David Inouye (RMBL) and his research group have measured the flowering time of plants in a bunch of plots around Gothic. Ian’s drones can cover large areas quickly. Using David’s plots, as well as some additional field-collected data, Ian can “train” computers to identify the species of plants from photographs. Ian’s drone can potentially allow the collection of plant flowering times at much larger spatial scales.

David’s data is critical for allowing us to understand how flowering times have changed since the early 1970’s. It may not be perfect for large spatial scales, but it will be the best we have (at least until field scientists start using time machines)!

Ian’s work is scaling up David’s flowering data, but is the data still only relevant for the Gothic Valley? No!! Using big data and technology (e.g., satellite and plane-based measurements) for applications like predicting flowering times in agricultural systems across the globe will depend upon places like RMBL!

Big data needs to be combined with theory and understanding to make powerful predictions. Places like RMBL, where our unique long-term data and teams of top scientists uniquely position us to build powerful models of ecosystem, will be the places that drive the value of big data! It’s exciting and transformative time for science. You are seeing it in action at RMBL!

The mapping revolution, big data, emerging technology, long-term studies, and one of the largest and longest collaborations of field scientists are the components of a powerful place that will change environmental science, and ultimately have world-wide implications.