Estimating causal networks in biosphere--atmosphere interaction with the PCMCI approach Biogeosciences

Kurzfassung

The dynamics of biochemical processes in terres- trial ecosystems are tightly coupled to local meteorological conditions. Understanding these interactions is an essential prerequisite for predicting, e.g. the response of the terres- trial carbon cycle to climate change. However, many em- pirical studies in this field rely on correlative approaches and only very few studies apply causal discovery methods. Here we explore the potential for a recently proposed causal graph discovery algorithm to reconstruct the causal depen- dency structure underlying biosphere-atmosphere interac- tions. Using artificial time series with known dependencies that mimic real-world biosphere-atmosphere interactions we address the influence of non-stationarities, i.e. periodicity and heteroscedasticity, on the estimation of causal networks. We then investigate the interpretability of the method in two case studies. Firstly, we analyse three replicated eddy covariance datasets from a Mediterranean ecosystem. Sec- ondly, we explore global Normalised Difference Vegeta- tion Index time series (GIMMS 3g), along with gridded cli- mate data to study large-scale climatic drivers of vegetation greenness. We compare the retrieved causal graphs to sim- ple cross-correlation-based approaches to test whether causal graphs are considerably more informative. Overall, the re- sults confirm the capacity of the causal discovery method to extract time-lagged linear dependencies under realistic settings. For example, we find a complete decoupling of the net ecosystem exchange from meteorological variability during summer in the Mediterranean ecosystem. However, cautious interpretations are needed, as the violation of the methods assumptions due to non-stationarities increases the likelihood to detect false links. Overall, estimating directed biosphere-atmosphere networks helps unravel complex mul- tidirectional process interactions. Other than classical correl- ative approaches, our findings are constrained to a few mean- ingful sets of relations, which can be powerful insights for the evaluation of terrestrial ecosystem models.