Functional data analysis-based yield modeling in year-round crop cultivation

Area of Science:

• Agricultural Science
• Statistical Modeling
• Horticultural Science

Background:

• Effective agricultural management relies on accurate crop yield prediction.
• Longitudinal crop cultivation, especially in controlled environments like plant factories, presents complex dynamics influenced by environmental factors.
• Understanding these dynamics is crucial for optimizing resource use and ensuring sustainable production.

Purpose of the Study:

• To develop and validate a methodology for modeling the relationship between environmental parameters and crop yield in longitudinal cultivation.
• To assess the impact of environmental fluctuations on crop yield using advanced statistical techniques.
• To provide insights for optimizing growth parameters and enhancing resource efficiency in horticulture.

Main Methods:

• Functional Data Analysis (FDA) was employed to model the complex relationships.
• A varying-coefficient functional regression model (VCFRM) was utilized to analyze time-series data.
• The model visualizes seasonal shifts and the dynamic interplay between environmental factors (solar radiation, temperature) and crop yield.

Main Results:

• The VCFRM successfully modeled the relationship between environmental variables and crop yield in strawberry and tomato production.
• Seasonal shifts and the dynamic influence of environmental factors on yield were visualized.
• The interpretability of the FDA-based model provided actionable insights for optimizing crop production.

Conclusions:

• The feasibility of VCFRM-based yield modeling for longitudinal crop cultivation was demonstrated.
• The methodology offers strategies for stable, efficient crop production.
• This approach is pivotal for addressing climate adaptability challenges in plant factory-based horticulture.