Wavelet analysis as a tool to deduce temporal multi-scale influence of climatic events on a young beech forest CO2 exchanges

Eco-physiological behavior of terrestrial ecosystems is highly dependent on climate variations. Underlying eco-physiological processes take place on multiple temporal scales, which hinders their identification as well as the study of their effects. Although this relationship is well-documented on time frames extending from few days to few years, decades-long studies are relatively sparse. With the growing availability of continuous eddy-covariance (EC) records of greenhouse-gas fluxes, the opportunities to conduct such long-term studies are rapidly expanding. The enlargement of available temporal series widens the possibilities to investigate inter and intra-annual variability and sensitivity of ecosystem eco-physiological responses to medium to large scale events. This study aims to explore the influence of such events on a young beech forest at ICOS-candidate Hesse site, north-eastern France. To account for the variable scale of these events, a time-frequency analysis tool, the continuous wavelet transform (CWT) is applied on 20 years of CO2 fluxes exchanges. The CWT is an increasingly popular tool for the investigation of environmental time series, owing to its compatibility with the investigation of non-stationary phenomena. This work implements both the Mexican Hat and Morlet wavelet and exploits associated wavelet tools, such as partial coherencies, to account for interactions between multiple variables. New approaches to detect anomalies in EC fluxes are suggested based on the frequency analysis of time series large scale components. Ultimately, this study aims to improve the characterization of the effect of large-scale events, extending from months to years, on ecosystem eco-physiological responses variations.