Soil‑Weather Multivariate Crop Disease Dataset
Year: 2026
Type: dataset
Abstract: This dataset contains 250,000 real observations of crop, soil, and weather measurements collected from agricultural fields. It is designed for multi‑class disease forecasting in precision agriculture. The data can also be used for general supervised learning, feature engineering, and early warning system development. Key Characteristics Rows: 250,000 Features: 23 (raw + engineered) Disease classes: 31 Crops: 6 (Maize, Potato, Rice, Sugarcane, Tomato, Wheat) Temporal ordering: By an implicit time step (daily or near‑daily observations) Feature Composition Category Variables Soil properties Nitrogen, Phosphorus, Potassium, pH Value, Soil Temperature Weather measurements Humidity, Rainfall, Weather Temperature (°F), Weather Humidity, Cloud Cover (%), Wind Speed (km/h), Rolling Rainfall (mm), Temperature Variance Derived indicators Weather Favorability Score, Low/High Temperature Warnings, High Humidity/Wind Warnings, Cloudy/Wet Spell Warnings Engineered features Thermal Gap, Moisture Gap, Stress Index, Crop‑Weather interaction, Thermal/Moisture/Wind‑Rain products, Crop‑Temperature
Authors: Md. Abid Hasan, Rafi, Md. Dulal, Haque
Language: English
Topic: Smart Agriculture and AI
Subfield: Plant Science
Domain: Life Sciences
SDG: Zero hunger
Open Access status: green
OpenAlex ID: W7158284591