Soils of boreal forests are crucial carbon reserves. The response of soil carbon emission to climate change significantly affects the concentration of carbon dioxide in the atmosphere. Soil carbon emission models frequently show a nonlinear response to temperature, but soil moisture is an important limiting factor, often overlooked in energy limited ecosystems. We suggest a statistical model of soil CO2 emission constrained by soil moisture and temperature for different ecosystems in the boreal zone. We tested this modelling strategy using direct measurements of seasonal soil CO2 emission near the research observatory ZOTTO near the Bor settlement, Central Siberia, Russia, in 2012-2017. Soil moisture explained a significant amount of variability of soil emission: the adjusted R2 was twice higher than in the baseline model. Although the temperature-only model describes the annual variability of carbon dioxide emissions quite well, the addition of moisture measurement significantly refines the quality of prediction of the seasonal component dynamics. Models including both temperature and soil moisture could serve as a promising tool to analyze the carbon cycle in boreal forest ecosystems.
Read full abstract