Spatio-temporal patterns in repeated spontaneous potential measurements in the crater of Teide volcano (Tenerife).

The analysis of multiparametric geophysical and geochemical datasets presents significant challenges due to the diverse nature of measurements and their potential interactions. The development of advanced statistical and data mining techniques has enabled researchers to identify and characterise complex patterns within such datasets. This work builds upon a previous study that applied Independent Vector Analysis (IVA) to analyse multiparametric measurements collected at Teide volcano crater (Tenerife, Canary Islands) between 2020 and 2024. Our main goal is to extend the initial findings by identifying endogenous and exogenous factors influencing the observed patterns and characterising their temporal behaviour.

The dataset includes spontaneous potential, CO₂ and H₂S fluxes, and thermal gradient measurements taken within the crater of Teide volcano on 38 fixed points. The application of IVA, which is an extension of Independent Component Analysis (ICA), allows for a multivariate approach that leverages vectorial data instead of scalar quantities. This method has proven effective for disentangling spatio-temporal interactions and isolating independent processes that govern the observed geophysical and geochemical variations.

Based on previous preliminary results, this study incorporates new data collected during 2023 and 2024, allowing a better definition of the spatio-temporal patterns. Using the IVA, we identify and quantify evolving endogenous patterns potentially related to magmatic processes. Simultaneously, we assess the influence of exogenous factors such as seasonal temperature fluctuations and hydrological changes.

Our results highlight the robustness of IVA in separating and characterising independent processes contributing to spatio-temporal multivariate datasets, such as the specific case of the Teide volcano. The results reveal a strong correlation between spontaneous potential anomalies and localised gas emissions, validating this methodology in volcanic environments. Moreover, this extended study underscores the importance of integrating temporal dynamics into multivariate analyses to improve the understanding of volcanic systems.

This work demonstrates the potential of IVA as a powerful tool for analysing repeated geophysical and geochemical surveys. It offers significant advantages for monitoring active volcanic systems. Future applications could include adding more datasets, such as remote sensing and/or other geophysical or geochemical parameters, to understand volcanic processes comprehensively.