Quantifying the impact of 2015-2016 ENSO over global soil and vegetation water content by causality models

Abstract:

Establishing causal relations from observational data is perhaps the most important challenge for today’s advances in Earth system science. Here we are interested in uncovering spatio-temporal causal relations at different time scales between two relevant Earth observation variables -soil moisture (SM) and vegetation optical depth (VOD)- and their coupled relation with the global ocean-atmosphere extreme conditions induced by the 2015-2016 ENSO event. SM and VOD variables are estimated from the natural microwave emission of the land surface, which is mostly driven by the water content in soils and vegetation. The cause-effect analysis of their interactions allows us to shed light into important aspects of ecosystem functioning and health, such as quantification of water movements in the soil-vegetation continuum and their relation to global atmospheric extremes and ENSO in particular. To deal with the complexity data problem, we first extract the most relevant and expressive feature components from 9 years of global SM and VOD data with the nonlinear kernel-based dimensional reduction method ROCK-PCA [1]. To infer causality relations, we use the cross-information kernel Granger causality (XKGC) method introduced in [2], which accounts for nonlinear cross-relations between the involved variables and generalizes nonlinear GC methods. Our results show that the 2015-2016 ENSO event induced anomalous dry/wet global teleconnection patterns at interannual time scales mostly represented over the tropics with distinct region-based delays over soils and vegetation. The developed causality framework allows revisiting ENSO global footprints on continental lands with EO data as well as to uncover the impact of ENSO on yet unreported areas.

[1] D. Bueso, M. Piles and G. Camps-Valls, "Nonlinear PCA for Spatio-Temporal Analysis of Earth Observation Data" in IEEE Transactions on Geoscience and Remote Sensing, vol. 58, no. 8, pp. 5752-5763, Aug. 2020

[2] D. Bueso, M. Piles and G. Camps-Valls, "Cross-information Kernel Causality" (2019). Proceedings of the 9th International Workshop on Climate Informatics: CI 2019