Time-Frequency Analysis of Soil–Atmosphere Interactions

Published: 6 November 2025| Version 1 | DOI: 10.17632/55swnv5szz.1
Contributors:
Mauricio Felipe Revelo Aristizabal, Daniel Felipe Ruiz Restrepo, Juan Felipe Paniagua-Arroyave

Description

This dataset contains the preprocessed and derived data used in the three case studies analyzed in the research "Time-Frequency Analysis of Soil–Atmosphere Interactions." The data include preprocessed time series of soil and atmospheric variables, as well as the results obtained from the Ensemble Empirical Mode Decomposition (EEMD) analysis. The dataset is organized by study site.

Files

Steps to reproduce

As an example, the code for La Roque-Gageac is included and described. The R script starts by loading the necessary libraries and importing the dataset. It then renames the columns for clarity and converts the date column to a proper time format. After defining the time step and period ranges in months, a univariate wavelet transform is performed on a single variable (e.g., TD1_2m) to examine how its energy varies across time and frequency. The resulting power spectrum is saved as an image. Next, a second dataset is created to perform a wavelet coherency analysis between two variables (e.g., TD2_6m and D2_6m), which reveals how strongly they are related across different time scales. Both the wavelet power and coherency plots are exported as PNG files for further interpretation. For the Hilbert-Huang transform (EEMD) and Spearman correlation analysis, a Python script is included. This script contains routines for computing the HHT spectrum, the power of each IMF, and the Spearman correlation matrix. All related information is documented in the accompanying notebook. For the other datasets, the same procedure is followed, and the accompanying notebook includes the required libraries to run it.

Institutions

  • Universidad EAFIT Escuela de Ciencias y Humanidades

Categories

Spectral Analysis of Signal, Non-Linear Time Series, Coupled Problem in Geotechnics

Licence