关键词: 作物生长模型, 算法构建, 生产力预测, 效应评估, 决策支持, 系统平台, 数字农作

Abstract:

Agricultural information technology is formed as the result of integrating information technology and agricultural science, and has further facilitated the rapid development of digital agriculture (DA) and smart agriculture (SA). As one of the core technologies of DA and SA, crop growth model can dynamically simulate crop growth and development processes and their relationships with climate condition, soil characteristics and management strategy, so as to overcome the limitation of the spatial-temporal characteristics of traditional research on agricultural production management. It can provide powerful quantitative tools for crop productivity prediction and early warning and impact evaluation under different conditions. Through over 20-years systematic and profound exploration and practicing in wheat and rice crops, and based on the workflow of “physiological mechanism analysis-model algorithm development-dynamic productivity prediction-quantitative effect assessment-simulation platform development”, our research team has been devoted to the development and application of crop simulation model CropGrow, by integrating the technologies of system analysis, dynamic modeling, virtual reality, scenario simulation, and decision support. Firstly, based on the system analysis method and dynamic modeling technology, the comprehensive and mechanistic crop growth model CropGrow has been developed, including the submodels of phasic development and phenology, organ development and population establishment, photosynthetic production and biomass accumulation, assimilate partitioning and yield/quality formation, nutrient dynamics, and water balance, along with three-dimensional morphological and visual submodels, which could digitalize and visualize the processes of crop growth and productivity formation under different conditions. Further, by coupling geographic information system (GIS) and remote sensing (RS), the model-based regional crop productivity prediction technology has been established. Then, based on the scenario analysis, the contributions of climate change, soil improvement, variety updating, and strategy optimization to regional crop production have been quantified, and applications extended to generation of suitable management plan, design of ideal cultivar, assessment of climate impact, evaluation of land use and decision-making of agricultural policy. Finally, based on the component-based programming technology, a model-based digital and visual crop growth simulation system and decision support platform has been developed by integrating the crop production database and crop model components, further realizing the comprehensive functions of data management, parameter optimization, growth simulation, remote sensing coupling, regional prediction, management strategy design, effect evaluation, safety early warning and product release. In the future, based on the improvement of agro-information database, additional efforts in crop modeling will be made toward enhancing prediction ability, quantifying gene effects, developing intelligent decision-making, and coupling multiple models, which will provide digital support for the prediction and early warning of food production, quantitative evaluation of scenario effects, decision-making on management strategy, and optimal design of new crop cultivars, thus facilitating the security of national food and development of digital agriculture.

Key words: crop growth model, algorithm development, productivity prediction, impact assessment, decision support, system platform, digital farming