Cross’ research takes him to the treetops of the Costa Rican rainforest

Imagine spending two weeks in the Costa Rican rain forest, waking up to the sounds of howler monkeys, spending your days hiking through the forest and eating dinner while the sun sets. Matthew Cross, a civil engineering PhD student, did just that this summer, but it wasn’t a tropical vacation. Cross was stationed at La Selva Biological Station to collect data for his dissertation, “Improved Understanding of the Carbon Cycle through Accurate Above-Ground Biomass Measurements of Central American Forests.” La Selva Biological Station is situated on 3,900 acres of tropical forests in northern Costa Rica and is managed by the Organization for Tropical Studies. La Selva is part of a network of ecological research stations in Costa Rica that represent some of the most important sites for original research on neotropical ecology. 

The goal of this research is to better characterize the forest by tree type to get a better idea of the trees’ biomass. Currently, this is determined only by crude estimates of the forest canopy and density. The biomass is the living portion of the tree, both above and below the ground. An accurate measure of the biomass, says Cross, is critical in realizing the effect of forests on global climate through the carbon cycle.

“Trees uptake carbon, and this uptake varies by tree type,” he explains. “An enhanced process for the identification of tree types from satellite imagery and generating accurate biomass measure could lead to an improvement in estimates of carbon storage and uptake related to variations in the earth’s forests, assisting the scientific community in determining the impact that variations in forest cover have on the global climate.”

Cross’s trip to Costa Rica was the beginning of three phases of research. The first is to develop a taxonomy specification for trees using the known plantation sites at La Selva to refine the process of identifying tree type. This will be achieved using high-resolution satellite imagery, along with spectrometer measurements taken in the field during his trip. Spectrometer data was collected both at ground level and also above the canopy. Ted Scambos from the Cooperative Institute for Research in Environmental Sciences at the University of Colorado Boulder assisted in this process.

“This ‘ground truth’—looking at the tree canopy from the top down—will verify what the satellite imagery is measuring for specific tree types,” Cross says.

The second phase is to develop biomass estimations on the plantations using the results of phase one and LiDAR data, which measures tree height. Finally, Cross will take his processes into the real forest, where the trees aren’t in specific locations by groups like on the plantations.

Although he deems the trip a success, Cross still has a lot of work to do. Right now, his focus is on compiling all of the spectrometer data collected, verifying the tree types measured, and using this ground truth data to build a satellite-based identification schema for the tropical rain forest. This will be done using plantation sites that provide homogeneous tree types for a large, defined area. It is anticipated that he will need to return to La Selva next spring.

“One of the things that quickly became evident is the complexity of the forests,” says Cross. “It is likely that species identification will not be realistic. However, a preliminary look at the data suggests that differentiation by family is achievable, which will have a significant impact on the accuracy of the overall biomass calculation.”