Characterizing causal dependencies through information flow: emerging frontier for understanding and prediction of earth system dynamics
Abstract
Spatially dense and contemporaneous multivariate observations at high frequencies are opening up an unprecedented opportunity to understand and characterize deeply embedded, yet hitherto unresolved, inter-dependencies that govern earth system process dynamics. For example, we may ask: how are water, solutes and gaseous fluxes involving biological, ecological, chemical, and energetic processes codependent through cause-effect relationships? New approaches based on information flow among variables are now becoming available to enable the exploration of full range of interdependencies that may be embedded in the observations which have the potential to substantially improve our process understanding and predictive capabilities. We draw upon recent advances in the theory of information flow for geoscience studies to enable systemic understanding of self-organized dynamics in Earth system dynamics, and their characterization for modeling. Information flow captures the attenuation or amplification of fluctuations through interaction among variables thereby revealing the dynamic connectivity between them. Analysis of such information flow can provide a unique vantage point for understanding watershed functions: that of quantifying multivariate causal interactions that link across space and time scales of the watershed dynamics. We show that patterns of interactions exhibited through propagation of fluctuations among interacting variables and measured as information flow between them, which reflects varying patterns of dynamic connectivity, expose asymmetric and multi-scale causal dependencies of the dynamical system. Further, the amplification or attenuation of the fluctuations, as they are transmitted from one variable to another, is both a reflection of and a response to the emergent system structure and function. We illustrate these concepts using observed data and discuss future potential to address complex system studies.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2019
- Bibcode:
- 2019AGUFM.H13D..05J
- Keywords:
-
- 1899 General or miscellaneous;
- HYDROLOGY