Python-based PRISMA workflow for bibliometric screening, systematic review, and meta-regression of Land Use and Land Cover mapping studies in Sub-Saharan Africa.
# LULC Sub-Saharan Africa — Systematic Review & Meta-Regression Pipeline
**Authors:** Francisco José Noris · Valéria Cristina Rodrigues Sarnighausen
**Institution:** UNESP — Faculdade de Ciências Agronômicas, Botucatu, Brazil
**Paper:** *Methodologies of land use mapping in Sub-Saharan Africa: a systematic review*
**Target journal:** Remote Sensing of Environment (Elsevier)
---
## Overview
This repository contains the complete computational pipeline for the systematic review
of LULC classification methodologies in Sub-Saharan Africa (1991–2026).
**The pipeline covers four stages:**
```
Stage 1: BibTeX deduplication → 01_deduplication.py
Stage 2: Automated REGEX screening → 02_regex_screening.py
Stage 3: REGEX audit sampling → 03_validation_sampling.py
Stage 4: OLS meta-regression → 04_meta_regression.py ← main results
Stage 5: Figure generation → 05_figures.py
```
**Key result (fully reproducible):**
- N = 155 studies with quantitative Overall Accuracy
- is_MLC: β = −0.369 (SE = 0.162; t = −2.273; p = 0.024)
- Confirmed robust under HC3, bootstrap (B=5,000) and Cook's D sensitivity analysis
---
## Quickstart
```bash
# 1. Clone repository
git clone
github.com]/LULC-SubSahara-MetaRegression.git
cd LULC-SubSahara-MetaRegression
# 2. Install dependencies
pip install -r requirements.txt
# 3. Download data (from Zenodo — NOT included in this repository)
# Place dados_LIMPOS_metaregressao.xlsx in the data/ folder
# Data DOI:
doi.org
# 4. Run the full pipeline
python src/04_meta_regression.py # Main results (Table 2)
python src/05_figures.py # All manuscript figures
# Expected output:
# outputs/Tabela2_OLS_Resultados.csv ← R²=0.052, is_MLC p=0.024
# outputs/TB_VIF_Diagnosticos.csv ← all VIF Noris, F.J. (2026). *LULC Sub-Saharan Africa — Systematic Review Pipeline* (v1.0.0).
> GitHub.
github.com]/LULC-SubSahara-MetaRegression. …