Replication Materials for “Extreme Rainfall, Antecedent Soil Wetness, and Short-Run Local Economic Activity in Bangladesh”
Description
This repository contains replication materials for the manuscript “Extreme Rainfall, Antecedent Soil Wetness, and Short-Run Local Economic Activity in Bangladesh.” The study examines how extreme rainfall and antecedent soil wetness are associated with short-run local economic activity across Bangladesh. The empirical analysis uses a balanced district-month panel covering all 64 districts over 132 months from 2015 to 2025, comprising 8,448 district-month observations. Monthly nighttime-light intensity from VIIRS is used as a proxy for local economic activity and is combined with precipitation information from CHIRPS, climate and hydrological variables from ERA5-Land, MODIS land-surface temperature, and district-level structural characteristics derived from HIES 2022. The preferred analytical sample contains 7,725 district-month observations satisfying the specified VIIRS spatial-coverage threshold. The replication archive contains the preserved consolidated district-month source panel, analysis-ready data, Stata 14-compatible analysis code, source-processing documentation, statistical tables, diagnostic outputs, figures, and result-verification scripts. The package reproduces the statistical analysis from the preserved district-month panel to the reported results. It does not reconstruct the earlier native raster and Google Earth Engine extraction stage because the original source-specific raster exports and historical extraction script are not included. Third-party source materials remain subject to the terms of their respective data providers.
Files
Steps to reproduce
Download and extract the replication ZIP while preserving the folder structure. Read the separately deposited README.txt before running the analysis. The replication code requires Stata 14.0 or later. Set the Stata working directory to the extracted package root. Run do "code/00_INSTALL_REQUIRED_PACKAGES.do" to install or verify the required community-contributed Stata commands, followed by do "code/00_RUN_FINAL_REPRODUCTION.do" to reproduce the deposited analysis. The master script audits the preserved source panel, validates the frozen analysis data, reproduces the statistical tables and figures, and executes the result-verification routines in code/99_VERIFY_REPRODUCED_RESULTS.do. Successful verification ends with the message ALL CORE CHECKS PASSED. The package reproduces the analysis from the preserved consolidated district-month panel; it does not reconstruct the earlier native-raster/Google Earth Engine extraction stage.
Institutions
- Jagannath UniversityDhaka Division, Dhaka