aibrahim-art/kainji-jebba-water-quality-temporal-dynamics

R scripts for variance partitioning, Granger causality, and changepoint detection of water-quality dynamics in the Kainji–Jebba reservoir cascade, Nigeria # kainji-jebba-water-quality-temporal-dynamics R scripts for variance partitioning, Granger causality, and changepoint detection of water-quality dynamics in the Kainji–Jebba reservoir cascade, Nigeria. ## Reference Ibrahim, A., Sadiku, S. O. E., Robertson-Andersson, D. V., & Okpeku, M. (2026). Temporal dynamics outweigh spatial gradients in shaping water quality across a regulated Afrotropical reservoir cascade. *Environmental Monitoring and Assessment*, 198, 906. Temporal dynamics outweigh… This repository contains the R scripts used for data analysis in the above manuscript. ## Scripts | Script | Purpose | Required local input | |---|---|---| | `01_land_use_extraction.R` | Land-cover extraction (forest, agriculture, urban cover) from ESA WorldCover 2021 within 5-km buffers around sampling sites, via Google Earth Engine (`rgee`); includes Moran's I spatial autocorrelation check | `sampling_site.gpkg`, `buffers_5km.gpkg` | | `02_WQI_computation.R` | Computation of the modified NSF Water Quality Index (WQI) and summary statistics/plots by site, habitat, season, and month | `WQI_data.csv` | | `03_variance_partitioning.R` | Partial redundancy analysis (RDA) and variance partitioning of water quality among spatial, temporal, and land-use predictor sets; PERMANOVA; NMDS ordination | `water_quality_data.csv` | | `04a_granger_causality_full.R` | Full stratified Granger causality testing (480 tests: 5 climate drivers × 8 water-quality parameters × 12 strata), with stationarity checks, AIC-based lag selection, and Benjamini–Hochberg FDR correction | `Grouped_Dataset_Kainji_Jebba.xlsx` | | `04b_granger_heatmap.R` | Generates the seasonally split Granger causality heatmaps (Fig. 5a–b) from the near-significant results of `04a_granger_causality_full.R` | *(values embedded from Table 3 of the manuscript)* | | `05_changepoint_detection.R` | Multi-method changepoint detection (PELT, binary segmentation, CUSUM, Bayesian, variance-based, and moving-average meth …