A UAV-based hybrid approach for improving aboveground dry biomass estimation of winter wheat

文献类型： 外文期刊

第一作者： Zhao, Yu

作者： Zhao, Yu;Wang, Chao;Feng, Meichen;Xiao, Lujie;Yang, Wude;Zhao, Yu;Feng, Haikuan;Han, Shaoyu;Li, Zhenhai;Yang, Guijun;Han, Shaoyu

作者机构：

关键词： Unmanned aerial vehicle; Aboveground dry biomass estimation; PROSAIL; Hybrid model; Growing degree day

期刊名称：EUROPEAN JOURNAL OF AGRONOMY （ 影响因子：5.5； 五年影响因子：5.9 ）

ISSN： 1161-0301

年卷期： 2025 年 168 卷

页码：

收录情况： SCI

摘要： Aboveground dry biomass (AGB) is one of the main manifestations of the farmland carbon sink, and its determination is essential in yield estimation, farmland management, and understanding of the farmland carbon cycle. The PROSAIL model, based on the physical understanding of vegetation optical and biophysical processes, serves as a crucial tool and method for retrieving the physiochemical parameters of vegetation. The PROSAIL model, by explaining the radiation transfer processes within the canopy, helps establish a clear physical relationship between biomass and remote sensing observations. AGB retrieval using the PROSAIL model typically relies on transforming leaf characteristics and leaf area index (LAI). However, the use of fixed parameters may not adequately capture the variations in AGB across different growth stages. Thus, this study suggested a hybrid approach to retrieve AGB by combining the PROSAIL model with growing degree days (GDDs) to address the limitation of biomass models not extending across various growth stages. Results indicated the following: (1) the relation between LAI and AGB varied considerably across different growth stages. (2) The PROSAIL model effectively predicted LAI, achieving an R-2 of 0.76 and a root mean square error (RMSE) of 0.41. (3) Compared to support vector machines and random forests, the constructed hybrid method performs better in biomass monitoring across multiple growth stages, with R-2 and RMSE values of 0.86 and 1.40 t ha(-1) for calibration datasets and 0.76 and 1.83 t ha(-1) for validation datasets, respectively. The proposed method extends the application potential of the PROSAIL model, demonstrating strong robustness and universality in AGB estimation across different growth stages, providing technical support for AGB estimation of other crops.

分类号：

  • 相关文献
作者其他论文 更多>>

微信小程序