Continuous long-term eddy covariance (EC) measurements of CO₂ fluxes (NEE) in a variety of terrestrial ecosystems are critical for investigating the impacts of climate change on ecosystem carbon cycling. However, due to a number of issues, approximately 30–60% of annual flux data obtained at EC flux sites around the world are reported as gaps. Given that the annual total NEE is mostly determined by variations in the NEE data with time scales longer than one day, we propose a novel framework to perform gap filling in NEE data based on machine learning (ML) and time series decomposition (TSD). The novel framework combines the advantages of ML models in predicting NEE with meteorological and environmental inputs and TSD methods in extracting the dominant varying trends in NEE time series. Using the NEE data from 25 AmeriFlux sites, the performance of the proposed framework is evaluated under four different artificial scenarios with gap lengths ranging in length from one hour to two months. The combined approach incorporating random forest and moving average (MA-RF) is observed to exhibit better performance than other approaches at filling NEE gaps in scenarios with different gap lengths. For the scenario with a gap length of seven days, the MA-RF improves the R² by 34% and reduces the root mean square error (RMSE) by 55%, respectively, compared to a traditional RF-based model. The improved performance of MA-RF is most likely due to the reduction in data variability and complexity of the variations in the extracted low-frequency NEE data. Our results indicate that the proposed MA-RF framework can provide improved gap filling for NEE time series. Such improved continuous NEE data can enhance the accuracy of estimations regarding the ecosystem carbon budget.

Dryland ecosystems are critical in regulating the interannual variability of the global terrestrial carbon cycle. The responses of such ecosystems to weather and environmental conditions remain important factors that limit the accurate projections of carbon balance under future climate change. Here, we investigated how shifts in vegetation phenology resulting from changes in weather and environmental conditions influenced ecosystem carbon cycling in one semiarid ecosystem in the Hanford area of central Washington, United States. We examined two years of measurements of the phenology camera, eddy covariance, and soil chamber from an upland semiarid sagebrush ecosystem. Both years had contrasting diel and seasonal patterns of CO${}_2$ fluxes, primarily driven by differences in vegetation phenology. The net ecosystem exchange of CO${}_2$ (NEE) and evapotranspiration (ET) in 2019 were enlarged by shifted vegetation phenology, as a cold and snow-covered winter and warm and dry winter in 2020 resulted in constrained magnitudes of NEE and ET during the summer months. The annual gross primary productivity (GPP) was much higher in 2019 than in 2020 (−211 vs. −112 gC m${}^{-2}$), whereas ecosystem respiration was comparable in these two years (164 vs. 144 gC m${}^{-2}$). Thus, the annual NEE in 2019 was negative (−47 gC m${}^{-2}$) with the sagebrush ecosystem functioning as a carbon sink, while the positive annual NEE in 2020 indicated that the sagebrush ecosystem functioned as a carbon source. Our results demonstrate that winter snowpack can be a critical driver of annual carbon uptake in semiarid sagebrush ecosystems.