Global soil respiration: interaction with macroscale environmental variables and response to climate change

The response of global soil respiration (Rs) to climate change determines how long the land can continue acting as a carbon sink in the future. This dissertation research identifies how temporal and spatial variation in environmental factors affects global scale Rs modeling and predictions of future Rs under global warming. Chapter 1 describes the recommend time range for measuring Rs across differing climates, biomes, and seasons and found that the best time for measuring the daily mean Rs is 10:00 am in almost all climates and biomes. Chapter 2 describes commonly used surrogates in Rs modeling and shows that air temperature and soil temperature are highly correlated and that they explain similar amounts of Rs variation; however, average monthly precipitation between 1961 and 2014, rather than monthly precipitation for a specific year, is a better predictor in global Rs modeling. Chapter 3 quantifies the uncertainty generated by four different assumptions of global Rs models. Results demonstrate that the time-scale of the data, among other sources, creates a substantial difference in global estimates, where the estimate of global annual Rs based on monthly Rs data (70.85 to 80.99 Pg C yr-1) is substantially lower than the current benchmark for land models (98 Pg C yr-1). Chapter 4 simulates future global Rs rates based on two temperature scenarios and demonstrates that temperature sensitivity of Rs will decline in warm climates where the level of global warming will reach 3°C by 2100 relative to current air temperature; however, these regional decelerations will be offset by large Rs accelerations in the boreal and polar regions. Chapter 5 compares CO2 fluxes from turfgrass and wooded areas of five parks in Blacksburg, VA and tests the ability of the Denitrification-Decomposition model to estimate soil temperature, moisture and CO2 flux across the seasons.

Cumulatively, this work provides new insights into the current and future spatial and temporal heterogeneity of Rs and its relationship with environmental factors, as well as key insights in upscaling methodology that will help to constrain global Rs estimates and predict how global Rs will respond to global warming in the future.