Estimating comprehensive growth index for drip-irrigated spring maize in junggar basin via satellite imagery and machine learning

Mingjie Ma, Jinghua Zhao, Tingrui Yang, Feng Liu, Yingying Yuan, Shijiao Ma and Zikang Chang

Agricultural Water Management, 2025, vol. 318, issue C

Abstract: Crop growth indicators reflect the growth status and productivity potential of crops. The development of remote sensing technology provides a new perspective for modern agricultural crop growth monitoring. This study utilized Sentinel-2 satellite remote sensing data combined with ground-measured growth indicators (leaf area index, plant height, chlorophyll content, etc.) of drip-irrigated spring maize. A comprehensive growth index (CGI) was constructed using the coefficient of variation method to holistically assess crop growth status. Key feature variables were selected through correlation analysis and recursive feature elimination. Based on random forest (RF) and its optimized models (Bayesian-optimized RF and sparrow search algorithm-optimized RF), a CGI estimation model was proposed. The results indicated that leaf area index and plant height had stronger correlations with spectral indices and band reflectance values compared to other growth indicators. The CGI effectively represented the comprehensive growth status of spring maize and exhibited stronger correlations with spectral indices and band reflectance values than the single growth indicators. The SSA-RF model achieved the highest prediction accuracy for CGI across different growth stages (from jointing to maturity stage), with R² values ranging from 0.575 to 0.795. The total stage model significantly outperformed the single growth stage models in CGI prediction, yielding R2 values between 0.982 and 0.994. Shapley analysis scientifically demonstrated that RDVI was the most influential factor in the SSA-RF model's CGI predictions in total stage. Additionally, the study generated field-scale CGI spatiotemporal distribution maps, revealing the temporal and spatial variation patterns of crop growth. In summary, this study provides an efficient method for monitoring the growth of drip-irrigated spring maize in Northwest China, offering valuable guidance for precision agriculture management and sustainable development.

Keywords: Spring maize; Sentinel-2; Machine learning; Drip irrigation; Comprehensive growth index; Spatio-temporal variation (search for similar items in EconPapers)
Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
http://www.sciencedirect.com/science/article/pii/S0378377425003658
Full text for ScienceDirect subscribers only

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:eee:agiwat:v:318:y:2025:i:c:s0378377425003658

DOI: 10.1016/j.agwat.2025.109651

Access Statistics for this article

Agricultural Water Management is currently edited by B.E. Clothier, W. Dierickx, J. Oster and D. Wichelns

More articles in Agricultural Water Management from Elsevier
Bibliographic data for series maintained by Catherine Liu ().