Komanawa Groundwater Age Tools#
a small repo to hold groundwater age tools developed by Komanawa Solutions Ltd.
Dependencies#
pandas>=2.0.3
numpy>=1.25.2
scipy>=1.11.2
tables>=3.8.0
Installation#
This package is currently held as a simple github repo, but the intention is to make it available on PyPI in the future.
Install from PyPI#
pip install komanawa-gw-age-tools
Install from Github#
pip install git+https://github.com/Komanawa-Solutions-Ltd/komanawa-gw-age-tools
conda create -c conda-forge --name gw_detect python=3.11 pandas=2.0.3 numpy=1.25.2 scipy=1.11.2 pytables=3.8.0
conda activate gw_detect
pip install git+https://github.com/Komanawa-Solutions-Ltd/komanawa-gw-age-tools
Usage#
Detailed documentation is available in the docstrings of the functions and classes. The following is a brief overview of the package.
Exponential Piston Flow and Binary Piston flow models#
key functions:
exponential_piston_flow_cdf / binary_exp_piston_flow_cdf: calculate the cumulative distribution function for the exponential piston flow model
exponential_piston_flow / binary_exp_piston_flow: calculate the probability density function for the exponential piston flow model
check_age_inputs: check the inputs to the binary exponential piston flow model (convince function)
make_age_dist function#
make_age_dist is a convenience function that uses the binary exponential piston flow model to calculate the age distribution and age fraction of water in a groundwater system.
mrt = 15 # mean residence time
mrt_p1 = 10 # mean residence time of the first piston
mrt_p2 = None # will be calculated from mrt and mrt_p1
frac_p1 = 0.5 # fraction of the flow that is in the first piston
precision = 2 # the precision of the age distribution (0.01 years per age step)
f_p1 = 0.7 # fraction of the flow in the first piston which is exponential
f_p2 = 0.7 # fraction of the flow in the second piston which is exponential
mrt, mrt_p2 = check_age_inputs(mrt, mrt_p1, mrt_p2, frac_p1, precision, f_p1, f_p2)
age_step, ages, age_fractions = make_age_dist(mrt, mrt_p1, mrt_p2, frac_p1, precision, f_p1, f_p2, start=np.nan)
# ages are the age of water in years (from 0 to historical)
# age_fractions are the fractions of water in each age step (from 0 to historical)
# age_step is the age step in years (0.01 years in this case)
Source and Receptor analysis tools#
predict_source_future_past_conc_bepm#
mrt = 20
mrt_p1 = 10
frac_p1 = 0.7
f_p1 = 0.8
f_p2 = 0.75
initial_conc = 10
prev_slope = 0.5
max_conc = 20
min_conc = 1.
age_range = (-20, 50)
fut_slope = -0.1
total_source_conc, receptor_conc = predict_source_future_past_conc_bepm(initial_conc, mrt, mrt_p1, frac_p1, f_p1,
f_p2,
prev_slope, fut_slope, age_range,
max_conc, min_conc, max_fut_conc=20,
min_fut_conc=1)
import matplotlib.pyplot as plt
fig, ax = plt.subplots()
ax.plot(total_source_conc.index, total_source_conc.values, label='source_conc', color='b')
ax.plot(receptor_conc.index, receptor_conc.values, label='receptor_conc', color='r')
ax.axvline(0, color='k', ls='--', label='initial time')
ax.set_xlabel('time (years)')
ax.set_ylabel('concentration')
ax.set_title('predict_source_future_past_conc_bepm')
ax.set_xlim(age_range)
ax.legend()
fig.tight_layout()
plt.show()
predict_future_conc_bepm#
input_series = pd.Series(index=[-35, 0., 20, 27, 35, 40, 50, 100, 200],
data=[1, 1, 3, 5, 15, 21, 18, 2.4, 2.4])
mrt = 20
mrt_p1 = 5
frac_p1 = 0.2
f_p1 = 0.8
f_p2 = 0.75
data = predict_future_conc_bepm(once_and_future_source_conc=input_series,
predict_start=20,
predict_stop=200,
mrt_p1=mrt_p1, frac_p1=frac_p1, f_p1=f_p1, f_p2=f_p2, mrt=mrt, mrt_p2=None,
fill_value=1,
fill_threshold=0.05,
pred_step=0.5)
import matplotlib.pyplot as plt
fig, ax = plt.subplots()
ax.plot(input_series.index, input_series.values, label='Source Concentration (input)')
ax.plot(data.index, data.values, label='Receptor Concentration')
ax.set_xlabel('time (years)')
ax.set_ylabel('concentration')
ax.set_title('predict_future_conc_bepm')
ax.legend()
plt.show()
predict_historical_source_conc#
mrt = 20
mrt_p1 = 10
frac_p1 = 0.7
f_p1 = 0.8
f_p2 = 0.75
init_conc = 10
prev_slope = 0.5
max_conc = 20
min_conc = 1.
mrt_p2 = (mrt - (mrt_p1 * frac_p1)) / (1 - frac_p1)
precision = 2
source_conc = predict_historical_source_conc(init_conc=init_conc,
mrt=mrt, mrt_p1=mrt_p1, mrt_p2=mrt_p2,
frac_p1=frac_p1, f_p1=f_p1,
f_p2=f_p2, prev_slope=prev_slope, max_conc=max_conc,
min_conc=min_conc, start_age=np.nan, precision=precision)
import matplotlib.pyplot as plt
fig, ax = plt.subplots()
temp = pd.Series(index=source_conc.index)
temp[:] = init_conc + (prev_slope * source_conc.index[:])
temp[temp<min_conc] = np.nan
ax.plot(temp.index, temp.values, label='receptor concentration (input)', color='orange')
ax.plot(source_conc.index, source_conc.values, label='source concentration (predicted)', color='b')
ax.axvline(0, color='k', linestyle='--', label='initial time')
ax.set_xlabel('time (years)')
ax.set_xlim(-100,10)
ax.set_ylabel('concentration')
ax.set_title('predict_historical_source_conc')
ax.legend()
fig.tight_layout()
plt.show()
