PPGIS Uncertainty

This repository contains the code to adopt the method and reproduce the examples given in the paper support the Review of the papers Decision Making under Uncertainty: Increasing the Impact of Public Participatory GIS (accepted for publication in the International Journal of Geographical Information Science) and Embracing Uncertainty in Participatory GIS: Perceptions of tree planting in the English Lake District, published in the proceedings of GISRUK 2025.

• To reproduce the results, run lr_combination.py. This will output a raster containing six bands:
  1. Belief Trees
  2. Belief No Trees
  3. Plausibility Trees
  4. Plausibility No Trees
  5. Probability Trees
  6. Probability No Trees
• Generic methods for the functions are in ling_rudd2.py, if you run this directly it will reproduce the example given in the method.
• To reproduce the topic analysis, run topics.py

The PPGIS dataset collected using map-me is located in data/blobs.csv. The associated text is in data/map-me_answers_23-10-2023_12-39.csv, and the processed output from this text is in ./data/answers_with_terms.csv.

Also contained in the data/ directory is a Shapefile containing geometries for the lakes, which is © 2024 OpenStreetMap Contributors.

The code used for the version using Focal Area Bias and the sensitivity analyses are located in the supplementaries directory.

Dependencies

• lr_combination.py: geopandas rasterio
• ling_rudd2.py: none
• topics.py: pandas, nltk^*

^* Note that you need to run some manual downloads once this is installed, see here.

To create an environment and install all of the dependencies, I recommend running the following commend using anaconda:

conda create --name ppgis --channel conda-forge --override-channels --yes python=3 geopandas rasterio nltk