Spatial analysis exam project 2021
Read the report»
View markdown
·
Report bug
·
Request feature
Table of Contents
The quality of Denmark’s groundwater supplies is under threat due to the use of nitrate in agriculture. In the last few years there has been a cultural shift in public opinion, where conventional agriculture is progressively being seen as more damaging to the environment, compared to its organic counterpart. We wish to assess the scientific validity of this belief. Using kriging, linear modeling and coefficient t-tests, we carry out a spatial analysis comparing nitrate concentration levels in Denmark for conventional- and organic agricultural farming. Here, we find significantly higher nitrate concentrations for organic fields, with organic- and conventional fields having mean nitrate concentrations of 24.51 mg/L and 23.48 mg/L, respectively. A discussion of results, confounding factors, and prospects for future research can be found in the written report on the project.
Example visualization from the project, showing nitrate measurement placements and kriged nitrate concentrations in the groundwater
See Groundwater_Pollution_in_Denmark.pdf for the written report. The written report contains the motivation as well as an in-depth introduction to the topic, specifics to running the analysis, results and finally also a discussion of the project.
To get a local copy up and running follow these simple example steps.
NOTE: There may be slight variations depending on the terminal and operating system you use. The following example is designed for Git Bash on Windows 10. You also need to have pip installed:
The spatial analysis requires data layers that exceed the maximum filesize on GitHub. To reproduce the analysis the script data/data_download.sh has been provided. It will automatically download the files needed for replicating our analysis.
For rerunning the analysis we therefore recommend cloning the repository, as well as using the provided script for downloading the data. This can be done using the following lines in an unix-based bash:
git clone https://github.com/emiltj/groundwater_pollution_dk.git
cd groundwater_pollution_dk
bash data_download.shYou should now be ready to run the code.
This repository has the following structure:
| File | Description |
|---|---|
Groundwater_Pollution_in_Denmark.pdf |
Written report containing our exam hand-in for Spatial Analytics 2021 |
groundwater_pollution_dk.md |
Markdown of the spatial analysis |
groundwater_pollution_dk.Rmd |
Script used for the spatial analysis |
groundwater_pollution_dk_files |
Directory containing output images from the script |
README_images/*.png |
Images used for the README |
README.md |
Readme with instructions |
metadata.md |
Metadata document specifying data structure and contents |
data_download.sh |
Script which downloads the data required for the analysis |
.gitignore |
.gitignore file for avoiding the large data files |
LICENSE |
MIT License which specifies the permitted usage of the repository |
Upon running the data_download.sh bash script, the directory will have an additional data directory, with the following content:
| File | Description |
|---|---|
data/nitrat.csv |
Groundwater samples containing information on nitrate and geographical coordinates of samples in the period 1900-2021. |
data/denmark_administrative_outline_boundary.* |
Shapefiles containing a map of Denmark. |
data/Markblok.* |
Shapefiles containing polygons of fields in Denmark in the period 1990-2021. |
data/Oekologiske_arealer.* |
Shapefiles containing polygons of organic field in 2018-2021 |
The data used in the analysis is all publicly available and has been acquired through the online portals of the respective ministries and software companies that are in possession of the data. All data sources are either official Danish governmental agencies or highly respected organisations. As such, we deem the data to be highly trustworthy and fit for scientific research. For ease of accessibility the data can be downloaded through our script data_download.sh located in our Github repository. The four datasets that were used in our analysis are:
nitrate.csvPoint data with samples of nitrate levels in Denmark. This dataset contains 14,350 measurements of nitrate concentrations at different geographic locations in Denmark from 1900 to March 2021. The dataset was provided by courtesy of De Nationale Geologiske Undersøgelser for Danmark og Grønland (GEUS). The included variables used in our analysis are: coordinates, measurement date, nitrate concentration and measurement depth. The data was retrieved from their webpage.Markblok.*A shapefile containing polygons of all current agricultural fields in Denmark as of April 2021. This dataset is provided by the Ministry of Food, Agriculture and Fisheries of Denmark (Danish Agricultural Agency). The dataset was retrieved from the following webpage.Oekologiske_arealer_(2012 - 2020]).*Shapefiles containing polygons of all organic agricultural fields in Denmark, registered each year between the period of 2012 to 2020. This dataset is provided by the Danish Agricultural Agency. The dataset was supposed to be publicly available through the Danish Agricultural Agency’s website but, due to server errors, we were sent the following link to the dataset in a mail correspondence with the agency.denmark_administrative_outline_boundary.*A shapefile containing a polygon in the shape of Denmark, courtesy of the software company IGIS MAP. The shapefile was retrieved from the their webpage.
For metadata information, see metadata.md.
After cloning the repository and downloading the required data, open up the groundwater_pollution_dk.Rmd R-markdown in Rstudio (see written report for version specifics). Running the code should give you direct access to the analysis which lays the foundation for our written report.
As an alternative to rerunning our script, you may merely want to inspect the code and the output.
For simply examining the analysis and its results, we recommend inspecting the markdown version of the analysis groundwater_pollution_dk.md.
Contributions are what make the open source community such an amazing place to be learn, inspire, and create. Any contributions you make are greatly appreciated.
- Fork the Project
- Create your Feature Branch (
git checkout -b analysis/extended_analysis) - Commit your Changes (
git commit -m 'Add some extended_analysis') - Push to the Branch (
git push origin analysis/extended_analysis) - Open a Pull Request
Distributed under the MIT License. See LICENSE for more information.
Feel free to write the authors, Emil Jessen or Johan Horsmans for any questions regarding the scripts. You may do so through our emails (Emil, Johan)
We want to give a special thanks to our instructor for the Spatial Analytics course, Adéla Sobotkova fornot only providing guidance for our project but also generally providing an informative and useful course. We also want to express our gratitude towards Lars West Andersen and the Danish Agricultural Agency for providing explanations of the data on agricultural fields.
Furthermore, we would like to extend our gratitude towards the following:
- RStudio - Software used for conducting the analysis
- GEUS-DATA - Data portal for geological investigations in Denmark/Greenland
- Danish Agricultural Agency - The Danish Agricultural Agency portal, containing geographical information on land use
- README template - README template by othneildrew
- Rstudio - Software used for conducting the analysis
- Overleaf - Software used to format and write the report