文献类型： 外文期刊

作者： Wu, Zongjun ¹ ; Cui, Ningbo ¹ ; Zhang, Wenjiang ¹ ; Yang, Yenan ¹ ; Gong, Daozhi ³ ; Liu, Quanshan ¹ ; Zhao, Lu ¹ ; Xing, Liwen ¹ ; He, Qingyan ¹ ; Zhu, Shidan ¹ ; Zheng, Shunsheng ¹ ; Wen, Shenglin ¹ ; Zhu, Bin ¹ ;

作者机构： 1.Sichuan Univ, State Key Lab Hydraul & Mt River Engn, Chengdu 610065, Peoples R China

2.Sichuan Univ, Coll Water Resource & Hydropower, Chengdu 610065, Peoples R China

3.Chinese Acad Agr Sci, Inst Environm & Sustainable Dev Agr, State Engn Lab Efficient Water Use Crops & Disaste, Key Lab Dryland Agr, Beijing 100083, Peoples R China

4.Sichuan Acad Agr Machinery Sci, Chengdu 610066, Peoples R China

关键词： Soil moisture; Thermal infrared data; Multi-spectral data; Unmanned aerial vehicle (UAV); Multi-modality data fusion; Remote sensing

期刊名称：AGRICULTURAL WATER MANAGEMENT （ 影响因子：5.9； 五年影响因子：6.2 ）

ISSN： 0378-3774

年卷期： 2024 年 302 卷

页码：

收录情况： SCI

摘要： Accurate and timely prediction of soil moisture in orchards is crucial for making informed irrigation decisions at a regional scale. Conventional methods for monitoring soil moisture are often limited by high cost and disruption of soil structure, etc. However, unmanned aerial vehicle (UAV) remote sensing, with high spatial and temporal resolutions, offers an effective alternative for monitoring regional soil moisture. In this study, multi-modal UAV remote sensing data, including RGB, thermal infrared (TIR), and multi-spectral (Mul) data, were acquired in citrus orchards. The correlations between different sensor data and soil moisture were analyzed to construct seven input combinations. Convolutional neural network (CNN), long short-term memory (LSTM) models and a new hybrid model (CNN-LSTM), were employed to predict soil moisture at depths of 5 cm, 10 cm, 20 cm and 40 cm. Additionally, the impact of standalone sensor, texture features and multi-sensor data fusion on the accuracy of soil moisture prediction was explored. The results indicated that the model with RGB + Mul + TIR achieved the highest prediction accuracy, followed by those with Mul + TIR and RGB + Mul, with the coefficient of determination (R2) ranging 0.80-0.88, 0.64-0.84, and 0.60-0.81, and root mean square error (RMSE) ranging 2.46-2.99 m3 center dot m-3, 2.86-3.89 m3 center dot m- 3 and 3.15-4.25 m3 center dot m- 3, respectively. Among single sensor inputs, the Mul sensor data has the highest prediction accuracy, followed by TIR and RGB sensor, with the coefficient of determination (R2) ranging 0.54-0.72, 0.36-0.52 and 0.14-0.26, and root mean square error (RMSE) ranging 3.72-4.58 %, 3.81-5.04 % and 4.27-6.21 %, respectively. The hybrid CNN-LSTM model exhibited the highest prediction accuracy, followed by CNN and LSTM models, with the coefficient of determination (R2) ranging 0.20-0.88, 0.16-0.83, and 0.14-0.81, and root mean square error (RMSE) ranging 2.46-5.01 m3 center dot m- 3, 2.68-5.35 m3 center dot m-3 and 2.81-6.21 m3 center dot m-3, respectively. The prediction accuracy of the models was the highest at the depth of 5 cm, followed by 10 cm, 20 cm and 40 cm, with the coefficient of determination (R2) average of 0.63, 0.62, 0.59, and 0.55, and root mean square error (RMSE) average of 3.70 m3 center dot m- 3, 3.79 m3 center dot m- 3, 3.85 m3 center dot m- 3 and 4.21 m3 center dot m- 3, respectively. Therefore, the hybrid CNN-LSTM model with RGB + Mul + TIR is recommended to predict soil moisture in citrus orchard. It provides method and data support for regional precision irrigation decision-making.