Seeing the Forest and the Trees

Trying to figure out how tropical forests work is fun…and also a pain in the ass. They’re remote, they’re complicated, they’re subtly (or starkly) different everywhere you go, they’re full of things that can bite, sting or otherwise inconvenience you. Many of them are a reasonable place to do research assuming you don’t mind these challenges, but other big swaths are in politically dicey areas to say the least. And did I mention they’re complicated?

Let’s say, for example, that you’re a card-carrying ecosystem ecologist who wants nothing more than to learn what nutrient most limits the growth of tropical trees. OK, cool, you know what to do — mark off some plots, add some fertilizer in various combinations, measure how fast the trees grow, write the papers. Oh but wait….the trees in this hectare are totally different from the trees in that hectare and that one over there is different still and dammit what to do about all these mounds of dirt from the leafcutter ants and ohshitohdear how do I wrap a dendrometer band around this thing:


to measure how much bigger the trunk might get?!

It’s not a trivial set of problems. The very wonder of tropical forests is their extraordinary diversity, and that diversity is not just biological. It’s biogeochemical. Some are warm and some are cool. Some are flat and some are ohmygod steep. Arguably, the “state factors” that can predict a lot of ecosystem dynamics (and from whence this blog derives its name) are more variable across tropical forests than any other slice of the planet. And yet, in texbooks and beyond, tropical forests are often subject to The Great Lumping. “They’re phosphorus limited.” “They’re on poor soils.” “They’re incredibly biodiverse.” Etc. For every one of these generalizations, the right response is typically: well, yes…and no.

Or so we think. Truth is, for the most part, we still don’t truly understand where it’s yes and where it’s no. Too many forests, too few studies. And mind-bendingly complex landscapes to sort through, almost anywhere you go. Problem is, all that complexity is not just a fun academic challenge – what tropical forests do on a grand scale matters to all of us. Why? Because they not only cradle an astounding fraction of Earth’s life, they are a major engine for Earth’s climate. When the earth breathes, a hell of a lot of that breath is tropical forest. They move more energy and carbon dioxide in and out of the air every year than any other ecosystem type. Cut ’em down, and not only does a lot of carbon hit the air, the climate of the region – and possibly the world – can change for the long-term. So we have to figure out what makes them tick, ideally in ways that allow confident predictions of what they may do in the face of a fast-changing global environment.

Enter the eye in the sky. Following years of development, the Carnegie Institution’s Greg Asner and team are now mapping tropical forests in remarkable new ways — ones that I suspect will be a game changer for finally getting a better grasp on what remains a true ecological frontier. The Carnegie Airborne Observatory (CAO) combines LiDAR and hyperspectral sensors on a single plane, from which it can image forests (or any other landscape) with astonishing detail. It’s like sticking big chunks of forest through an MRI tube – what comes out the other side can reveal wonders (and flaws) you never knew existed. Do yourself a favor and go poke around the CAO website, where you can “fly” through a slice of Amazon forest in a video based entirely on this new imagery. (As an aside, the newly operational Airborne Observatory Platforms that are part of NEON use the same technology, meaning we’ll be “seeing” landscapes in the U.S. in a brand new way very soon. Also very cool.)

Critically, what comes out of those sensors is not just a buncha pretty maps. It’s a mountain of quantitative information. From detailed ledgers of tree leaf chemistry on a grand scale, to estimates of carbon storage that even thousands of people could not do as well from the ground, to – perhaps most astonishingly of all – the potential to map biodiversity from the air, those little toys on the CAO plane really do let us see the forest and the trees. Hell, the technology has even been used to quantify differences in how male and female lions hunt!

More on this to come in the near future. We teamed up with the CAO gang to fly our long-term field sites in southwest Costa Rica last year, and are about to embark on a suite of new projects aimed to take maximum advantage of those extraordinary data. This image shows the flight lines we chose and their spatial coverage – in effect, mega-transects designed to cover major axes of variation in the Osa Peninsula region:


Out of those flight lines, here’s just a taste of what’s to come – LiDAR derived estimates of canopy height:


If you don’t know this slice of Osa forest (and pasture and mangroves and…), it’s just pretty pictures. But after years of walking around the place, I can quickly see things I know from the ground (yep, there’s that ridge with the really big trees)…and things I had no idea were there. The LiDAR lets you slice down to any level you want, so the 3D structure of the landscape – whether that be the forest itself or the ground on which they grow – is revealed in astonishing detail. That can be seen quite readily in the CAO website examples linked above. For achieving real progress in figuring out the mysteries of tropical forests, that detail may be everything – from such a wealth of structural data should flow truly new understanding of function, especially when combined with the additional insights the hyperspectral sensor can provide.

Phil Taylor is down there right now, mixing new and old school. As in, using these new data still means getting out there in the woods and making measurements. Phil is setting out plots in the flight lines that can serve as calibration / validation anchors for the imagery-derived estimates of things ranging from canopy height to carbon storage to leaf chemistry. Here’s a shot from yesterday, staring down the barrel of a 57 meter (yep, that’s damn tall…) tree:


And, apropos to hazards mentioned in the open of this post, here’s a lil’ friend Phil encountered along the way…


For many of us trained with our boots on the ground, framing questions around the CAO data represents a new way of thinking, and I’ve had moments of faux depression (“damn, that plane just measured infinitely more in two hours than we’ve done in ten years…“), but truth is it’s all a candy store of possibility. At least in the ecosystem world, most of us have always wanted to scale up whatever we might measure in a given spot. Problem was, past remote sensing options typically couldn’t provide the kind of local-scale resolution that could make your standard-issue tropical field ecologist comfortable…and yet still apply that high resolution over big areas.

No longer. Like iPhones and twitter, these new school remote sensing platforms are likely to revolutionize our (ecological) lives.  Ready or not.


One thought on “Seeing the Forest and the Trees

  1. Pingback: Asner elected to NAS | Townsend Lab at CU Boulder

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s