Large-scale adoption of alternative agricultural practices is needed to manage farm productivity and resources under global changes. Ecosystem models are a tool for developing and evaluating potential solutions at management and policy scales. However, data collection and compilation often lack documentation that hinders model calibration and testing. The objectives of this study were to describe the current soil datasets and to combine soil property data from various sources that are sufficient to support the use of agricultural models under Korean conditions. Numerous sources provide soil data that describe the farm environment by land cover, such as rice and other field crops. The interquartile range of soil organic carbon was 9.10–14.50 g kg-1 across paddy and non-rice field systems. The results suggest that it is premature to perform site-specific simulations of farming systems due to no direct carbon accounting and hydrological conditions of the soil. Global gridded soil information, such as SoilGrids, may supplement the local data, especially soil organic carbon (the interquartile range of 23.43–36.92 g kg-1). However, the descriptive statistics of soil properties extracted from the datasets were not in agreement. Because the Korean agricultural land is characterized by small size (less than 1 ha) and family operation as a management unit, it is crucial to build datasets more representative of small-size farms and relevant to the new generation of models.

Abstract
Introduction
Materials and Methods
    1. Ecosystem models
    2. Soil data and information
    3. Administrative divisions
    4. Data processing and analyses
Results and Discussion
Acknowledgments
References

• Ju-Hwan Lee(Assistant Professor, Department of Smart Agro-industry, Gyeongsang National University) Corresponding author
• Young-Son Cho(Professor, Department of Smart Agro-industry, Gyeongsang National University)