At the global scale, the greatest uncertainty in forest biomass estimates (in units of carbon mass per land area) is in the highest biomass systems. Because of the strong relationship of tree height…

Northern Arizona University received a $580,649 research grant award from a civilian federal agency (Shared Services Center) on January 15, 2021, with an ultimate completion date of November 30, 2024. This non-set-aside grant supports research to evaluate and improve forest biomass estimates derived from the Global Ecosystem Dynamics Investigation (GEDI) satellite instrument. The project addresses a critical gap in global carbon cycle modeling by focusing on exceptionally tall forest ecosystems, where biomass estimation uncertainty is greatest. The research leverages NAU's extensive ground-based studies and detailed datasets from three primary forest regions: coastal temperate coniferous forests in western North America (particularly Sequoia sempervirens ecosystems), tropical forests across Amazonia, the Congo Basin, and Borneo, where the university maintains over 180 one-hectare research plots with comprehensive tree height, biomass, and wood density measurements.

The research will utilize robust allometric equations developed from intensive field measurements and a network of fixed-area monitoring plots with five-year re-measurements to validate GEDI biomass estimates and identify sources of error in satellite-derived predictions. By combining detailed ground-based data with satellite observations, the project aims to improve allometric equations and biomass prediction models that are essential for understanding the biotic carbon cycle and enhancing climate prediction accuracy at both tropical and temperate scales. This work directly supports the advancement of forest carbon science and addresses a significant source of uncertainty in global climate modeling efforts.

At the global scale, the greatest uncertainty in forest biomass estimates (in units of carbon mass per land area) is in the highest biomass systems. Because of the strong relationship of tree height to biomass, these high biomass systems coincide with tall forest canopies. Tropical forest biomass estimates represent the largest uncertainty in the biotic carbon cycle, which is making future climate prediction difficult. The biggest error in amazonian biomass estimates is in tree height and wood density. New technologies are enabling the detection of canopies of exceptional height in amazonia and in malaysia, but the relation of these tall forests to biomass is still uncertain. In the temperate zone, tall forests are found in coastal environments where temperatures are not extreme, rainfall plentiful, and soils relatively fertile. Prime examples include the coniferous forests of western north america. We propose to evaluate gedi biomass estimates in exceptionally tall forests where we have worked extensively and have data that can be used to improve these estimates. Robust allometric equations for sequoia sempervirens forests have been developed for all vascular plant species in these ecosystems based on intensive measurements of standing trees and limited destructive sampling in both tall oldgrowth and shorter regenerating forests. A network of fixed area plots has established plots that includes 5-YR RE-measurements of all vegetation and annual mortality checks to quantify aboveground biomass, leaf area, and growth increments. Here we will leverage detailed ground-based studies to evaluate gedi biomass estimates and identify sources of error. In tropical forests, we have detailed estimates of tree height, biomass, wood density from 180, 1 HA plots (35 with traits and spectral properties) across the key tropical biomes of amazonia, the congo basin and borneo. We will use gedi, our ground-based data to improve allometric equations and biomass predictions at these sites. Further, we will use desis hyperspectral data to create an algorithm predicting wood density at these sites. In previous work, we have demonstrated the ability of hyperspectral leaf level remote sensing to accurately predict wood density in borneo and peru. Overall, our work will greatly improve gedi plot-level biomass predictions in the regions of the greatest uncertainty. We will further explore an exciting novel hypothesis that canopy height may have changed in different forested regions in recent decades because of the balance of influences by rising CO2 and VPD. With rising CO2, water use efficiency (WUE, CO2 assimilation per unit water loss) has increased over the past 20 years mainly through reductions in ET by ~80 g H2O M2 HA2. In theory, increased WUE may allow the tallest trees to grow taller. In regions where vapor pressure deficit (VPD) has increased strongly in recent decades, water stress may have increased in treetops and caused a reduction of canopy height. We propose to identify forested regions where increased WUE and VPD should have driven increases or decreases in tree height based on theoretical considerations, and then we will compare our estimates to observations of change in canopy structure using glas and gedi data. We recognize that both gedi and glas have large vertical uncertainties in tree height. However, with millions of canopy heights available, a power analysis indicates that we should be able to see a statistical difference given moderate height growth (~20CM according to the power analysis), and thus test whether the growth matches our theoretical expectations. Although height change alone is insufficient to infer short-term biomass change, this element of our proposed research will establish a foundation for linking remotely sensed canopy height to environmental drivers of forest structure that are expected to alter forest biomass in the future.