Tropical and Subtropical Agricultural Production and Climate Dataset (2013-2024)

This dataset provides a comprehensive record of agricultural production and agroclimatic conditions accross tropical (Indonesia) and subtropical (United States) regions over the period 2013-2024. It is designed to support research on crop productivity, climate variability and machine learning-based prediction models, particularly in cross-regional comparative studies.

The dataset integrates multi-commodity agricultural production data with key climate variables, enabling analysis of how different agroclimatic drivers influence crop yields across distinct ecological systems. This work supports data-driven approaches to understanding food security, climate resilience, and regional agricultural dynamics.

🇮🇩 Indonesia Tropical Agriculture Feature Dataset (IND-TAFD)

• Covers 34 provinces (harmonized from administrative changes)
• Represents tropical agroecosystems
• The dataset includes 1,669 records, covering the following crops : Rice, Maize, Soybean, Peanut, Mung Bean.
• Agricultural production and harvested area data were obtained through web scraping from Badan Pusat Statistik (BPS) for the period 2013–2017 and from the Ministry of Agriculture for 2018–2024. It should be noted that data for 2016–2017 are unavailable due to a statistical moratorium. Additionally, values recorded as 0 should be interpreted as data below the official reporting threshold rather than true zero production.
• Agroclimatic variables—including temperature, precipitation, specific humidity, and root zone soil wetness—were sourced from NASA POWER, providing consistent and spatially aligned environmental data for each region and observation period.
• The dataset is characterized by:
  • High precipitation variability
  • Relatively stable temperatures
  • High humidity influence on crop productivity

🇺🇸 United States SubTropical Agriculture Feature Dataset (US-SAFD)

• Covers 16 states + Washington DC (Southern region)
• Represents humid subtropical agroecosystems
• The dataset includes 738 records, covering the following crops : Corn-grain, Sorghum, Soybean, Peanut, Wheat.
• Agricultural production and harvested area data were obtained from the National Agricultural Statistics Service (NASS) under the United States Department of Agriculture (USDA).
• Agroclimatic variables—including temperature, precipitation, specific humidity, and root zone soil wetness—were sourced from NASA POWER, ensuring consistency in environmental data across regions and time periods.
• The dataset is characterized by:
  • Greater seasonal temperature variation
  • Lower humidity compared to tropics
  • Strong influence of soil moisture (root zone water)

Dataset Structure

The dataset is provided in CSV format and contains the following colum