Terrestrial ecosystems, such as mountain forests, are crucial components of the global carbon cycle, acting as a net CO2 sink. Whether an ecosystem is a source or a sink of carbon depends strongly on how individual components, like soils or different vegetation components, respond to climatic conditions. Understanding these processes and responses under a changing climate is essential for accurate projections of the global carbon cycle and the climate system. While whole-forest CO2 budgets are typically estimated using eddy covariance (EC) above the canopy, partitioning this exchange among different forest components remains challenging. Understory EC measurements within the canopy offer a potential solution. Combined with above-canopy EC systems or soil chambers, this approach allows for partitioning the ecosystem’s total CO2 budget. However, estimating CO2 exchange from understory EC setups is difficult due to frequent violations of ideal atmospheric conditions within the canopy, which are often assumed for standard data processing. This work revisits various aspects of EC data processing in such challenging environments to provide reliable estimates of individual ecosystem components. Data from an above- and within-canopy EC system in Mieming, Tyrol, Austria, are presented, along with results demonstrating the responses of different components to climatic extremes.