Simulating Complex CO2 Background Conditions for Indianapolis, IN, with a Simple Ecosystem CO2 Flux Model
Open Access
- Author:
- Murphy, Sam
- Graduate Program:
- Meteorology and Atmospheric Science
- Degree:
- Master of Science
- Document Type:
- Master Thesis
- Date of Defense:
- December 08, 2023
- Committee Members:
- Paul Markowski, Program Head/Chair
Kenneth James Davis, Thesis Advisor/Co-Advisor
Natasha Lynn Miles, Thesis Advisor/Co-Advisor
Armen R. Kemanian, Committee Member - Keywords:
- land-atmosphere interactions
ecosystem modeling
atmospheric boundary layer
eddy covariance
carbon cycle
urban greenhouse gas
carbon dioxide - Abstract:
- In this study, we evaluated the ability of a simple ecosystem carbon dioxide (CO2) flux model, the Vegetation Photosynthesis and Respiration Model (VPRM), to capture the complex CO2 background conditions observed in Indianapolis, IN. Using simulated biogenic CO2 fluxes in conjunction with mole fraction tower influence functions, we estimated biogenic CO2 mole fractions at three background towers in the Indianapolis Flux Experiment (INFLUX) network for three years (April 2017 to March 2020). From the simulated biogenic CO2 mole fractions, we estimated CO2 differences between two of the background towers compared to a third tower, which we call CO2 enhancements. We compared modeled and observed afternoon average CO2 enhancements at daily and monthly time scales for both towers. We evaluated the random errors introduced by the model for daily, monthly, seasonal, and yearly averaging periods. Additionally, we compared modeled and observed average daily cycles of CO2 fluxes during the growing season at agricultural eddy covariance flux sites surrounding Indianapolis. Monthly mean model-observation residuals rarely differed significantly from zero (only 7 of 72 site-months), indicating that the model can capture afternoon average CO2 enhancements at a monthly time scale with no significant bias. We found that the random error was smaller than 1 ppm for monthly, seasonal, and yearly averaging periods. When compared to the average observed daily cycles of CO2 fluxes during the growing season at corn and soybean sites, the modeled CO2 fluxes captured the site-to-site differences well. For 9 out of 14 site-months, the modeled maximum afternoon CO2 drawdown was within 25% of the observed peaks despite the observed maximum drawdowns ranging from -8 µmol m-2s-1 to -66 µmol m-2s-1. The model had a harder time capturing the peak nighttime respiration, with the modeled peaks differing from the observed peaks by between 22% to 81%. Although not central to our intended application, the model-observation residuals for CO2 enhancements at a daily time scale were on the same order of magnitude as the observed enhancements themselves; therefore, the model could not capture the observed day-to-day variations of afternoon average CO2 enhancements. The results of this study indicate that the simple and computationally inexpensive VPRM can be effectively used in urban CO2 inversions to represent complex seasonal variations in background conditions observed in Indianapolis. Indianapolis, a modest-size city surrounded by strong ecosystem fluxes, represents a rigorous test for the VPRM system; these results are thus encouraging for the use of VPRM in other urban settings.