Data for: District-specific accounting and sampling uncertainty in official rice statistics: implications for satellite-based crop validation in Sri Lanka.

This repository contains the full set of derived data used in the analysis. dcs_panel.csv is the primary parsed dataset of Department of Census and Statistics (DCS) paddy statistics by district, season, and year (1979–2026), from which all results in the paper ultimately trace back. yield_standard_errors.csv provides the 405 district‑season yield standard errors extracted from 18 DCS seasonal bulletins (2015–2025), all within the 0–25% relative‑error window and used in Sections 2.4 and 3.4 and Fig. 3. panel_vs_dcs_qc.csv contains a row‑by‑row comparison of the working panel against the published DCS series, identifying the Maha discrepancy documented in Supplementary S1. final_gross_to_net.csv reports the district‑specific gross‑to‑net harvested‑extent conversion factors inferred from the production identity over 1979–2025 (Section 3.1). final_gross_net_published.csv provides the gross and net harvested extents extracted directly from the seasonal bulletin PDFs, forming an independent second route for estimating the conversion factor. final_gross_to_net_crosscheck.csv compares the inferred and published factors district‑by‑district, supplying the direct evidence for the agreement reported in Section 3.1.