Quantifying Carbon Storage in Illinois Restored Ecosystems with Remote Sensing Techniques

Abstract

The Champaign County Forest Preserve District (CCFPD) manages diverse properties throughout Champaign County, Illinois, balancing habitat restoration with public access to ecological services. While some preserves prioritize recreation, others focus on conservation, with CCFPD actively expanding its holdings to enhance ecosystem services like carbon storage. This report evaluates aboveground carbon pools using field measurements and remote sensing, providing actionable insights for future land management.

In this report, our primary goal is to quantify the sequestered aboveground carbon of CCFPD restored and remnant forests using field and remote sensing techniques. Storage of carbon is an ecosystem service provided by forests as they remove carbon dioxide from the atmosphere and store it as stable biomass. These carbon assessments may be used by CCFPD as justification for potential future land acquisition, in addition to other ecosystem services such as recreation and soil stabilization. Our second goal is to design a survey methodology for CCFPD, or affiliated team, to estimate sequestered carbon with newly collected field and/or LiDAR data.

Inorganic carbon is sequestered from the atmosphere and stored within the ecosystem’s biomass. Trees store carbon long-term, whereas more volatile reservoirs, like the forest understory and animal communities, store less carbon. By mass, the carbon content of trees is approximately half of their biomass (Aalde et al., 2006; Pearson et al., 2007). Therefore, biomass measurements can be easily converted into stored carbon. Species-specific allometric equations are a common method of calculating a tree’s aboveground biomass by its trunk diameter (Jenkins et al., 2004; Chojnacky et al., 2014). From this, stored carbon can be estimated based on field data related to tree size and species, or state of decay for dead wood. However, collecting this type of data for large areas of forest is time consuming and labor intensive.

Instead of expensive and cumbersome field surveys, remote sensing methods can efficiently survey large areas. Several remote sensing methods, such as optical sensors and LiDAR, are commonly used for biomass estimation. Optical sensor data are a common data source for biomass estimation but are limited in their ability to perceive structure through forest canopies.

LiDAR, as an active remote sensor, can collect elevation and height data over large areas with fine resolution. The active sensor penetrates the forest canopy so both the ground and forest structure are observed in the resulting point clouds. A LiDAR-derived canopy height model represents the height of the uppermost forest canopy as a continuous surface. We collected field data and compiled LiDAR data for CCFPD properties to assess aboveground tree biomass. We performed linear regressions in R and a support vector regression in Python to test two model variations, linear and non-linear power, to relate the LiDAR-derived canopy height model to the aboveground biomass estimates from on field data. The support vector regression analysis was effective, but required implementation with Python rather than R. We elected to consolidate data analysis in R. A least-squares regression proved robust for field data, but could not be implemented effectively to the landscape scale. Finally, a generalized linear regression model with a logarithmic link performed the best, yielding realistic results with simpler implementation.

Our estimates of carbon accumulated in the aboveground, stable biomass of forested areas on CCFPD properties average 116.04 Mg/ha, similar to that of second-growth forests in the Midwest (Burrascano et al., 2013). Stored carbon variability within CCFPD (Table 6) reflects main vegetation types (Table 4). For example, Sangamon River Nature Preserve stores the least carbon and contains a mixture of prairie and savanna ecosystems amidst the wooded areas. In contrast, the two properties with the greatest storage of carbon, Heron View and Riverview Nature Preserves, are densely populated with trees and have well-developed canopies.

Given these findings we recommend: 1) CCFPD could enhance carbon sequestration by targeting acquisitions of structurally mature forests with closed canopies and high hardwood composition. 2) Within existing preserves, selective thinning of low-carbon species (e.g., invasive Norway maple) and suppression of understory competitors (e.g., invasive Amur honeysuckle) may accelerate carbon accumulation. 3) Integrating biennial LiDAR surveys with field validation would enable spatially explicit monitoring, particularly in transitional ecosystems like Sangamon River where targeted restoration could shift carbon trajectories toward those observed in high-performing preserves.