shuli-gis/electricity-reliability

Scripts for time-series analysis of nighttime lights to examine electricity reliability patterns in South Africa # electricity-reliability Scripts for time-series analysis of nighttime lights to examine electricity reliability patterns in South Africa This repository contains scripts and workflows for analyzing electricity reliability in South Africa using monthly VIIRS nighttime light time series and related spatial datasets. ## Overview This project uses nighttime light time series as a proxy for electricity reliability and instability across South Africa. The analysis focuses on spatial and temporal variation in light intensity to identify patterns associated with electricity disruption, variability, and recovery. The workflow includes raster preprocessing, time-series construction, statistical analysis, clustering, and mapping. Outputs are used to examine reliability patterns at different spatial scales and over time. ## Main Components - preprocessing monthly nighttime light data - building time-series datasets - calculating variability and trend metrics - decomposing time series into trend, seasonal, and residual components - clustering temporal patterns - generating maps and summary figures ## Data Main datasets used in this project include: - monthly VIIRS nighttime lights - South Africa administrative boundaries - supporting environmental or contextual datasets where applicable Large raw datasets are not included in this repository. ## Main workflow 1. preprocess monthly VIIRS nighttime light rasters 2. build a South Africa nighttime light data cube 3. calculate pixel-level temporal metrics, including: - mean - standard deviation - skewness - kurtosis - STL detrended anomalies - z-scores - annual index of dispersion (IoD) 4. preprocess CHIRPS precipitation data 5. summarize reliability patterns at province and national levels 6. validate city-level nighttime light indicators using Cape Town load-shedding data ## Repository Structure ```text sa-electricity-reliability/ ├── README.md │ └── scripts/ │ ├── 01_preprocess_viirs_monthly.py │ ├── 02 …