Productivity and phenology in a process-driven carbon cycle model
Date
2018
Authors
Cheeseman, Michael J., author
Denning, Scott, advisor
O'Dell, Chris, advisor
Barnes, Libby, committee member
Klein, Julia, committee member
Journal Title
Journal ISSN
Volume Title
Abstract
The carbon cycle is a major source of uncertainty in predicting future climate, especially with regard to changes in the terrestrial biosphere. One obstacle in predicting the sources and sinks of the carbon cycle is accurately predicting phenological transitions of the terrestrial biosphere with a global process-driven model. We hypothesize that the terrestrial biosphere and its phenological transitions can be simulated using a set of universal biological strategies and a simple set of plant functional types in the Simple Biosphere (SiB4) model. In order to test our hypothesis, we compare the SiB4 output to a suite of satellite observations of the terrestrial biosphere including solar induced fluorescence (SIF) from the Orbiting Carbon Observatory (OCO-2), MODIS-based LAI, and AVHRR-based NDVI. Our first analysis compares modeled canopy SIF to aggregated satellite observed SIF over different biomes. We find that the model consistently over predicts pixel-scale SIF. Modeled SIF over evergreen needleleaf forests has an especially high bias during the winter. Our second analysis compares modeled and observed phenology over different regions around the globe. We find that SiB4 is generally successful in simulating growing season onset, but often simulates late senescence, especially in grasslands. We also find that SiB4 simulates crops well in the United States but fails to properly predict the planting and harvesting time of crops in other regions, especially the developing world.
Description
Rights Access
Subject
terrestrial biosphere
fluorescence