nitinmagima/KenyaDroughtAnalysis

This repository contains a set of scripts for analyzing drought conditions and trends in Kenya using Google Earth Engine (GEE) and clustering drought metrics to identify patterns across administrative units. # README: Drought Identification, Trend Analysis, and Clustering Across Kenyan Administrative Units ## Project Overview This repository contains a set of scripts for analyzing drought conditions and trends in Kenya using Google Earth Engine (GEE) and clustering drought metrics to identify patterns across administrative units. The analysis is based on methods described in the paper and aims to replicate the findings for drought identification, trend analysis, and clustering at both the admin 1 and admin 2 levels. The repository contains four files: ### Files Overview 1. **kenyadrought_admin1_withgee.ipynb**: This Jupyter notebook performs the analysis at the admin 1 level by downloading the required data using Google Earth Engine (GEE). 2. **kenyadrought_admin2_withgee.ipynb**: This notebook performs the analysis at the admin 2 level, also using GEE to download the necessary data. 3. **kenyadrought_admin1.ipynb**: This notebook performs the analysis at the admin 1 level using pre-downloaded data from GEE. 4. **kenyadrought_admin2.ipynb**: This notebook performs the analysis at the admin 2 level using pre-downloaded data from GEE. 5. **kenya_admin2_droughtmetrics.csv**: Includes administrative information, drought metrics such as drought duration, severity, intensity, and other categorical features (e.g., region, season). 6. **kenya_admin1_precip.csv**: Pre-downloaded precipitation data for admin 1 from GEE. 7. **kenya_admin2_precip.csv**: Pre-downloaded precipitation data for admin 2 from GEE. 8. **public_emdat_custom_request_2024-10-01_36a9efba-9545-41bc-a272-3b49c638962b.xlsx**: EM-DAT data for Kenya. 9. **Geospatial Data**: Administrative boundaries are stored in a GeoDataFrame (admin_boundaries). We use the FAO GAUL: Global Administrative Unit Layers for the administrative boundaries. ## Analysis Steps ### Step 1: Data Collection Obtain long-term satellite-derived precipitation data using the CHIRPS dataset available in GEE. This data will be used to an …