Seiche Observation and Response Fitting: HaOn Boreholes and Lake Water Levels

This repository contains source data and the analytical code used to process high-frequency water level data and derive aquifer hydraulic properties based on periodic seiche forcing. The workflow consists of two main scripts: an R Markdown script for spectral analysis of the time-series data, and a Python Jupyter Notebook for the physical inversion of the resulting transfer functions. Source Data File: haon_5sa.csv.gz Description: A compressed CSV file of water pressure data. Water pressures in the boreholes and lake was measured using Keller–Druck PA 36XW pressure transducers (0–100 kPa range) with an accuracy of 50 Pa and a precision of 2 Pa. Barometric pressure was measured using a Vaisala PTB110 sensor (accuracy 30 Pa, precision 60 Pa), and the acquisition system was continuously synchronized using a Garmin GPS16X-HVS receiver. Pressure was sampled at 40 Hz and averaged over 5 s for the seiche analysis. The water pressure records were corrected for barometric pressure. Codes: 1. Spectral Analysis Script: haon_lake_cross_spectrum_analyses.Rmd Description: An R Markdown script that applies an adaptive sine-multitaper cross-spectrum analysis to evaluate the transient responses of borehole water levels to lake seiches. It calculates time-varying transfer functions across moving time windows to extract the dynamic relationship between the lake and the aquifer. Input: The data file haon_5sa.csv.gz, containing continuous, synchronized time-series data of lake water levels and borehole fluid levels (e.g., HaOn1, HaOn2). Output: CSV files (output_csv_H*.csv) containing the calculated time-series metrics at dominant seiche periods, including amplitude ratio (gain), phase lag, and coherence. 2. Hydraulic Parameter Inversion Script: fit_DV1_DH1.ipynb Description: A Python-based Jupyter Notebook that uses the output from the spectral analysis to invert a two-layer leaky-aquifer diffusion model. It employs a bounded L-BFGS-B optimization algorithm to find the best-fitting physical parameters for each measurement window, specifically the vertical hydraulic diffusivity (DV1), horizontal hydraulic diffusivity (DH1), and a coupling parameter (K). Input: The Excel file HaOn2_37_A_lag.xlsx, which contains the observed amplitude ratios (Slope) and phase lags (Lag_min) for specific measurement windows. Output: An Excel file appended with the fitted results (output_HaOn2_37_fitted.xlsx), containing the derived parameters (DV1, K, DH1), model error, and a qualitative fit flag (good, ok, poor) for every time window.