calibration of a winter saturated well with apparently strong evaporation effect in summer (France)

[VeloVolant]

Hi ! I am facing difficulties to calibrate a head time serie from a well located in Auvergne (link), France, within the volcanic range (about 1000 masl). The shaft well is there/ggle street view. From my point of view gwl reach a recurring maximum during winter and decline in summer due to evaporation from trees around, direclty "pumping" the shallow gw. The local geology is considered as poorly aquifer. I believe this looks like a wetland gw signature.

It is within a country side, without any pumping wells. This observation well is monitored and frequently visited by the French Geological Survey (BRGM).

I have tried various softwares and solutions (e.g. Pastas with FlexModel including an interception reservoir, and Gardenia with a double reservoir), however none of them allows to reproduce the summer declines. Winter fluctuations are quite easy to calculate. I am using ERA5-Land precipitations and PET. Maybe such PET under-estimates the potential evaporation from trees.

Any suggestions would be welcome ! Regards

[mbakker7]

Thanks for the question @VeloVolant. Do you have a figure with the time series data and the fit produced with Pastas and Flexmodel? Maybe a results figure that also shows the fitted parameters?
BTW, the link doesn't work (for me).

[dbrakenhoff]

I figured out how to see the graph yesterday. If you follow this link: https://ades.eaufrance.fr/Fiche/PtEau?Code=BSS001WHRD then click > "Mesures de niveau d'eau" -button at the top of the page and then the "Chronique piézométrique" button below that, you get to this graph:

[VeloVolant]

You are a French master :)

[VeloVolant]

Here is the dataset
all_data.xlsx

[VeloVolant]

import pandas as pd
import pastas as ps
import numpy as np
import matplotlib.pyplot as plt

oseries = pd.read_excel("all_data.xlsx", sheet_name="Niv", index_col=0, parse_dates=True).squeeze()
rain = pd.read_excel("all_data.xlsx", sheet_name="Era5_L_PP", index_col=0, parse_dates=True).squeeze()
evap = pd.read_excel("all_data.xlsx", sheet_name="Era5_L_PET", index_col=0, parse_dates=True).squeeze()
well = pd.read_excel("all_data.xlsx", sheet_name="Era5_L_PET", index_col=0, parse_dates=True).squeeze()

ps.plots.series(oseries, [evap, rain])

ml = ps.Model(oseries.dropna().squeeze(), name="gwl")

rch = ps.rch.FlexModel(interception=True)
sm = ps.RechargeModel(rain, evap, recharge=rch, rfunc=ps.FourParam(), name="recharge")
ml.add_stressmodel(sm)

ml.solve(tmin="2000", tmax="2023")
ml.plots.results(figsize=(17, 9))
ml.plot(figsize=(17, 9), oseries=0, tmin="2000", tmax="2023")