Upendo1990/Simanjiro_Drought_Reproducibility

Reproducible data and code accompanying the manuscript: Temporal drought characteristics in semi-arid Northern Tanzania based on multi-scale SPI and SPEI (1994–2024) # Simanjiro drought analysis reproducibility package This repository reproduces the SPI and SPEI analyses for Simanjiro District, Tanzania, for January 1994–December 2024. ## Contents - `R/`: ordered R scripts for data preparation, PET, SPI/SPEI, classification, distribution diagnostics, TFPW trend analysis, FDR correction, post-whitening diagnostics, SPI–SPEI comparison, residual drought, PET sensitivity, and exploratory Pettitt tests. - `data/`: location for the restricted TMA input workbook. - `outputs/`: generated processed datasets, tables, and figures. - `metadata/`: package versions, analysis settings, and complete R session information. - `run_all.R`: runs the complete workflow. ## Input data Place the TMA workbook in: `data/SIMANJIRO_DATA.xlsx` The workbook must contain a worksheet named `Monthly_Data` with: - rainfall in `A2:M33`, - monthly Tmax in `A35:M66`, - monthly Tmin in `A68:M99`, - the first column named `Year`, - monthly columns named `Jan` to `Dec`. The workflow checks that the series contains 372 continuous monthly observations and no missing rainfall or temperature values. ## Run Open R in the repository root and run: ```r source("run_all.R") ``` ## Main methods - SPI: Gamma distribution, unbiased probability-weighted moments. - SPEI: log-logistic distribution, unbiased probability-weighted moments. - Primary PET: Thornthwaite. - PET sensitivity: Hargreaves–Samani using Tmin and Tmax. - Accumulation periods: 3, 6, 9 and 12 months. - Trend analysis: Trend-Free Pre-Whitening Mann–Kendall with Benjamini–Hochberg FDR correction. - Drought threshold: index ≤ −1.00. - Non-finite index values are retained in the audit and converted to missing only for downstream finite-value analyses. ## Data restrictions --- # Citation This repository is archived in Zenodo and can be cited using the following DOI: doi.org GitHub repository: github.com Raw meteoro …