Resolve "add the example for JD"

- [portfolio run and analysis](portfoliorun_analysis.md) page demonstrating running on a portfolio of prospective locations, including some plotting examples to illustrate the outputs.

- [customised stochastic simulation](customised_stochastic_simulations.md) page demonstrates how to develop your own stochastic frameworks around the core pythermogis doublet simulation functionality.

- [parallelization](parallelization.md) page describes how to parallelize simulations and determine the optimal chunk size for parallelization for your hardware

- [depth optimization](depth_optimization.md) page describes how to setup a set of stochastic simulations to enable assesement of optimal doublet depth in an aquifer.

- 

!!! info "Plotting, calculations and result analysis"

    pythermogis is designed to enable users to run geothermal doublet simulations. 

    Customized plotting is intentionally limited, as users often have specific preferences 

# Depth Optimization

This is an example of how to use the `calculate_doublet_performance` function from 

the `pythermogis` package to run the simulation for a range of transmissivity values across depth in an aquifer.

The final result then allows for optimal assesment of where to place a Doublet system in depth for a single location.

This relies on the following information:

- an estimate of the vertical extent of the aquifer and the uncertainty of that estimate

- an estimate of the porosity-depth relationship of the aquifer and the uncertainty of the relationship

- an estimate of the porosity-permeability relationship of the aquifer and the uncertainty of the relationship

Combining these variables allows a stochastic model to be made combining the uncertainties in these parameters which highlights the correlations between

transmissivty, depth, unit technical cost and power output:

![img.png](../images/depth_optimization.png)

The code used to make this plot is as follows (and is the same as `test_example9` in `test_doc_examples.py` in the test suite of the repo.)

````python

from pathlib import Path

import numpy as np

import xarray as xr

from matplotlib import pyplot as plt

from pythermogis import calculate_doublet_performance

# output directory to write output to

output_data_path = Path(__file__).parent / "resources" / "test_output" / "example9"

output_data_path.mkdir(parents=True, exist_ok=True)

# define the poro-depth and the poro-perm relationships:

# porosity = a * exp(-b * depth), with depth in km

poro_a = 28.97

poro_b = 0.21

poro_std = 1  # 1% uncertainty, if this goes larger it blows up massively...

# ln(permeability) = a * poro + b

perm_a = 0.334

perm_b = -1.655

# number of depths and number of samples per depth

n_samples = 500

n_depths = 10

depths = np.linspace(1e3, 5e3, n_depths)

thickness = 100

ntg = 1.0

porosity = poro_a * np.exp(-poro_b * (depths * 1e-3))

# expand porosity to have n_samples per depth, with uncertainty in porosity

porosity = np.random.normal(porosity[:, None], poro_std, size=(n_depths, n_samples))

porosity = np.clip(porosity, 0.0, 100.0)

permeability = np.exp(perm_a * porosity + perm_b)

reservoir_properties = xr.Dataset(

    {

        "thickness": thickness,

        "porosity": (["depth", "samples"], porosity / 100),

        "ntg": ntg,

        "permeability": (["depth", "samples"], permeability),

    },

    coords={

        "samples": np.arange(n_samples),

        "depth": depths,

    },

)

# For every sample, run a doublet simulation

simulations = calculate_doublet_performance(

    reservoir_properties,

    chunk_size=100,

)

simulations_stoch = simulations.quantile([0.1,0.5,0.9], dim="samples")

# make a plot showing there is a sweet spot between Doublet power, Cost of energy

# and Transmissivity

fig, ax = plt.subplots(figsize=(8, 6))

ax2 = ax.twiny()

p10 = simulations_stoch.sel(quantile=0.9)

p50 = simulations_stoch.sel(quantile=0.5)

p90 = simulations_stoch.sel(quantile=0.1)

# UTC

ax2.fill_betweenx(p50.depth, p10.utc, p90.utc, color="lightgrey", alpha=0.5)

ax2.plot(p50.utc, p50.depth, label="UTC [€/kWh]", color="lightgrey")

# power

ax2.fill_betweenx(p50.depth, p10.power, p90.power, color="black", alpha=0.5)

ax2.plot(p50.power, p50.power.depth, label="Power [MW]", color="black")

# transmissivity

ax.fill_betweenx(

    p50.depth,

    p10.transmissivity_with_ntg,

    p90.transmissivity_with_ntg,

    color="tab:green",

    alpha=0.5,

)

ax.plot(

    p50.transmissivity_with_ntg,

    p50.transmissivity_with_ntg.depth,

    label="Transmissivity [Dm]",

    color="tab:green",

)

ax.set_ylabel("Depth [m]")

ax2.set_xlabel("Power [MW] & UTC [€/kWh]")

ax.set_xlabel("Transmissivity [Dm]", color="tab:green")

ax.set_ylim([depths[-1], depths[0]])

# Collect handles and labels from both axes

handles1, labels1 = ax.get_legend_handles_labels()

handles2, labels2 = ax2.get_legend_handles_labels()

handles = handles1 + handles2

labels = labels1 + labels2

ax.legend(handles, labels)

plt.savefig(output_data_path / "depth_optimization.png")

````

 No newline at end of file

      - potrfolio run and analysis: usage/portfoliorun_analysis.md

      - customised stochastic simulations: usage/customised_stochastic_simulations.md

      - parallelzation: usage/parallelization.md

      - depth optimization: usage\depth_optimization.md

      #- out maps and locations: usage/pvalues_map_locations.md

  - API Reference:

  timestamp: 1740946648058

- pypi: ./

  name: pythermogis

  version: 1.2.2

  sha256: 2dff903415a0c622cd66d54c9e49503f3dbfc95cfc43bdc51e2d631c652d9136

  version: 1.2.3

  sha256: 6c8828289c71f3180f90af7d940428d8362b440e147bf8a6e533c4d28a8232d2

  requires_dist:

  - jpype1>=1.5.2,<2

  - xarray==2024.9.0.*