Role of Micrometeorological Memory in Modulating Sub-Daily Scale Variability of Net Ecosystem Exchange

Net Ecosystem Exchange (NEE) is crucial for understanding the carbon balance in ecosystems, indicating whether they act as carbon sinks or sources. While the impact of hydrometeorological factors on NEE at daily and monthly scales has been well-researched, the significance of sub-daily variability and the influence of memory in micrometeorological variables remain understudied. This study addresses this gap by analyzing the temporal dynamics of NEE using half-hourly data from 29 FLUXNET sites over at least 6 years. We found that sub-daily variability of NEE contributes 10%–55% of 13-day NEE variability, depending on seasonal cycles and biome characteristics. Using an information theory based transfer entropy (TE) approach, we identified the causal drivers of NEE variability at sub-daily scales within a 6-hr memory. Our results show that the memory of micrometeorological variables significantly impacts NEE, surpassing their instantaneous effects. Temperature (TA), vapor pressure deficit (VPD), and soil water content (SWC_Mean) consistently affect NEE within this 6-hr memory, whereas the influence of sensible heat (H) and incoming shortwave radiation (SW_IN) diminishes at higher lags. While the magnitude of average TE from micrometeorological variables to NEE exhibits notable seasonal variations, the temporal structure of how information is transferred does not significantly differ across seasons, as reflected by the shape of TE values over various time lags. SWC_Mean, VPD, and TA impact NEE jointly, while H and SW_IN have overlapping effects. Additionally, precipitation influences NEE indirectly through SWC_Mean. Our findings highlight the importance of accounting for high-frequency NEE variability and its underlying drivers when investigating the ecohydrological interactions, shedding light on the role of memory in carbon-water interactions.