Ice borehole thermometry: Sensor placement using greedy optimal sampling

This repository provides the data and code supporting the research article "Ice Borehole Thermometry: Sensor Placement Using Greedy Optimal Sampling" by K. Shaju, T. Laepple, N. Hrisch, and P. Zaspel.

Description

This repository contains the following directories and files:

• Forward_Model/ – For data generation.

• Sensor placements/ – Code, data, and results.

• Plots.ipynb – Steps to reproduce article plots.

• requirements.txt – Required Python libraries.

• README.md – Project overview and instructions (that's me!).

Data generation

The data used for sensor placement methods consist of borehole temperature–depth profiles simulated using a heat transfer model based on various possible surface temperature time series. The code for the heat transfer model, along with the surface temperature time series data, is organized in the Forward_Model/ directory.

We consider two sites to evaluate our results: EDML and GRIP. The Forward_Model/ directory includes the following subdirectories and files:

• data/ contains the possible surface temperature time series for both sites.

• src/ holds the code and relevant parameters for the heat transfer model.

• EDML_borehole_simulation.ipynb demonstrates how to simulate the borehole temperature profile for EDML.

• GRIP_borehole_simulation.ipynb demonstrates the borehole profile simulation for GRIP.

The borehole temperature–depth profiles thus simulated are stored under Sensor placements/data_borehole_simulations/, as described in the Code and Data section below.

Code, data and results

The data generated and the code used to produce the results presented in the article are organized in the Sensor placements/ directory. This directory includes the following subdirectories and files:

• code/ contains Python scripts (.py) used to generate the results (i.e., the plots shown in the article). These scripts utilize the data in data_borehole_simulations/ and store output files in the output/ directory.

• data_borehole_simulations/ holds the necessary data, including borehole temperature–depth profiles and other relevant files.

• output/ contains the generated results, which are used to create the figures in the article.

• Sensor placement using greedy optimal sampling.ipynb provides a step-by-step explanation of the greedy optimal sensor placement method, which is the core focus of the research article.

Plots

The procedures for reproducing the article's plots and locating the corresponding results are detailed in Plots.ipynb.

Requirements

The requirements.txt lists the Python libraries required to run the code.

Instructions

README.md - That’s me! 🙂 Your guide to the repository, with an overview and usage instructions.

Software dependencies and installation

This project requires Python 3 (tested with Python 3.10) and the packages listed in requirements.txt.

To be able to execute the notebooks, install the Jupyter notebook for example:

`pip install notebook==6.5.2`