Agricultural system modeling: current achievements, innovations, and future roadmap

• Published in: Arabian journal of geosciences Vol. 15; no. 4
• Main Authors: Ahmed, Zeeshan, Gui, Dongwei, Qi, Zhiming, Liu, Yi, Liu, Yunfei, Azmat, Muhammad
• Format: Journal Article
• Language: English
• Published: Cham Springer International Publishing 01.02.2022; Springer Nature B.V
• Subjects: Agricultural production; Algorithms; Applications programs; Big Data; Cloud computing; Coding standards; Computer programs; Earth and Environmental Science; Earth science; Earth Sciences; Farming systems; Farms; Food security; Learning algorithms; Machine learning; Mobile computing; Model accuracy; Modelling; New technology; Parallel processing; Programming languages; R&D; Remote sensing; Research & development; Review Paper; Simulators; Software; Software testing; Soil water; Sustainability; Sustainable agriculture; Climate change; Crop improvement; Crop models; Agro-ecosystems; Software development; Food security
• ISSN: 1866-7511; 1866-7538
• DOI: 10.1007/s12517-022-09654-7

Agricultural system models are tools that provide a synthesis and quantification to evaluate the effects of water, soil, crops, management practices, and climate on the sustainability of agricultural production and to ensure food security. Present-day agricultural models are the outcomes of research initiatives started 3–4 decades ago. However, existing models are not fully equipped with the important advancements achieved in the field of data and information and computer technology (ICT). The majority of the existing models are still using the old programming languages and legacy codes, software testing is uncommon while maintenance of documentation and software/codes is also a neglected avenue. These deficiencies could be rectified through better data harmonization and interlinking of models by developing different frameworks such as BioMA (Biophysical Model Applications) and APSIM (the Agricultural Production Systems Simulator). These developments assist in data compatibility by creating a common vocabulary and datasets for model ensembling. For next-generation modeling, gaps in the existing data should be minimized, a transition from supply-driven approach to demand-driven approach is needed to develop models according to the demands of end-users. Finally, focus on the software design and development should be encouraged in the modeling community as ICT has opened new horizons in the form of parallel processing or cloud computing methods, software languages and coding standards, and the development of user-friendly community-driven mobile applications that will enable the use of models to a more divergent group of stakeholders. Overall, agricultural systems modeling needs to rapidly adopt new technologies such as ICT, big data, remote sensing, and machine learning algorithms that will help enhance crop models’ accuracy and efficiency in designing sustainable agricultural systems at different farms, landscape, regional, and continental scales to meet the future demands of end-users.