UCLA

Carbonyl sulfide (COS) is a trace gas participating in key processes of the terrestrial carbon cycle. Despite its low mixing ratio in the troposphere (400–550 pmol mol^–1), the amplitude of seasonal variability of COS greatly exceeds that of CO₂ and is in phase with the gross photosynthesis of the terrestrial biosphere. Over the recent decade, COS has emerged as a promising tracer for quantifying terrestrial gross primary productivity (GPP) independently from respiration across the ecosystem to the global scales, because of the parallel uptake of COS and CO₂ through leaf stomata. While leaf uptake of COS dominates surface COS flux on land in the absence of industrial and biomass burning emissions, soil COS flux is another smaller but significant component. Neglecting the soil component in ecosystem COS budget may bias GPP estimates derived from COS measurements. Soil may also vary from a sink to a source of COS depending on temperature and microbial sulfur metabolism. Due to the presence of potential interference from soil COS activities, using COS as a photosynthetic tracer requires soil COS flux to be separated from the net ecosystem COS exchange. This dissertation is dedicated to the mechanistic understanding of the soil–atmosphere exchange of COS using process-oriented modeling and field observations.

A reactive transport model for soil COS processes is constructed to simulate soil–atmosphere COS flux from environmental variables. This model takes into account the dual-phase diffusive transport and the microbial sources and sinks of COS in the soil column. COS uptake and production rates are parameterized with enzyme kinetics and thermodynamics, consistent with lab incubation data. Leaf litter layer is explicitly resolved to account for litter COS uptake, whenever a litter layer is present. The model is evaluated against published field data of COS flux and demonstrates good skill in predicting both soil uptake and emission of COS. Model simulations further confirm that COS flux dependence on soil moisture is a result of two rivaling controls—the diffusive limitation on COS supply and the water limitation on microbial activity.

Field observations on soil COS exchange have been conducted at an oak woodland in southern California and a boreal pine forest in southern Finland using automated soil chambers and mid-infrared quantum cascade laser spectrometer. Soils at both sites show consistent uptake behavior related to soil moisture and respiration. At the semi-arid oak woodland in California, microbial COS uptake is strongly limited by water availability in the dry season. The intact leaf litter layer contributes a significant portion to the overall soil COS uptake. Litter COS uptake increases with moisture content and shows a strong pulse immediately after the rain event, indicating a rapid reactivation of litter microbial activity following alleviated water stress. In the Finnish pine forest, soil COS uptake is limited by the diffusional supply of COS to soil microbes, according to the negative correlation with soil moisture. The contrasting responses of soil COS uptake to moisture in semi-arid and humid ecosystems reflect the coupling of diffusion and microbial uptake controls on COS flux. At both sites, soil COS uptake correlates well with respiration and the COS : CO₂ flux ratio varies with temperature. The temperature dependence of COS : CO₂ flux ratio may be a common feature of soils and indicate underlying shifts in active microbial groups.

This dissertation advances knowledge of the physical and biological drivers of soil–atmosphere exchange of COS. Anticipated applications of the findings will be to better constrain global soil COS flux and derive COS-based estimates of GPP, which will be useful in understanding the responses of photosynthesis to climate variability.