A Multi-Decadal Curated Geospatial and Hydro-Climatic Benchmark Dataset for Desert Locust Occurrence Analysis in Egypt

Overview This repository contains the model-ready benchmark dataset associated with the manuscript: “A Multi-Decadal Curated Geospatial and Hydro-Climatic Benchmark Dataset for Desert Locust Occurrence Analysis in Egypt.” - The dataset provides a survey-level benchmark for desert locust occurrence analysis in Egypt covering the period 1981–2024. It integrates source-recorded FAO desert locust survey observations with spatially and temporally aligned environmental information derived from NASA POWER, Open-Meteo, MODIS, and ERA5-Land products. - The primary benchmark contains 19,572 survey records, including 4,196 PRESENT and 15,376 ABSENT observations. The modelling specification used in the associated study contains 15 conceptual predictors: 13 numeric predictors and two categorical predictors (Vegetation State and Season). - The dataset preserves the original survey observation as the unit of analysis. Exact full-row duplicates were removed, and records belonging to exact location/date keys with contradictory PRESENT and ABSENT source labels were excluded from the primary benchmark. - Access to the row-level benchmark dataset is restricted through Zenodo. Researchers may request access through the Zenodo access-request mechanism, subject to the terms and conditions applicable to the underlying third-party data sources. Repository Contents 1. Desert Locust Occurrence Benchmark Dataset for Egypt (1981-2024).csvThe final model-ready benchmark dataset containing 19,572 survey records used for the primary analyses reported in the associated manuscript. Access to this file is restricted. 2. README.mdDocumentation describing the dataset structure, variables, data sources, processing conventions, missing-value handling, target definition, and recommended use. Data Sources The benchmark integrates information derived from: - FAO Desert Locust Information Service (Locust Hub)- NASA POWER- Open-Meteo historical archive- MODIS vegetation products- ECMWF ERA5-Land Important Note - The benchmark represents source-recorded survey PRESENT and ABSENT observations. ABSENT indicates that locusts were not recorded during the corresponding survey and should not be interpreted as proof of ecological absence over a broader area or time period. - Differences between historical and later observations should be interpreted as sample-level covariate shift because survey geography and survey effort also vary through time; they should not be interpreted as direct evidence of national climate trends.