| Hyperspectral estimation modeling of photosynthetic rate in cotton canopy using chlorophyll data |
|
View Fulltext View/Add Comment Download reader |
| |
| DOI：10.7606/j.issn.1000-7601.2025.01.21 |
| Key Words: cotton photosynthetic rate chlorophyll density feature band selection spectral estimation model |
| Author Name | Affiliation | | HOU Buping | College of Agriculture, Tarim University, Key Laboratory of Genetic Improvement and Efficient Production of Characteristic Crops in Arid Areas of Southern Xinjiang, Research Center of Oasis Agricultural Resources and Environment of Southern Xinjiang, Alar, Xinjiang 843300, China | | WANG Jiaqiang | College of Agriculture, Tarim University, Key Laboratory of Genetic Improvement and Efficient Production of Characteristic Crops in Arid Areas of Southern Xinjiang, Research Center of Oasis Agricultural Resources and Environment of Southern Xinjiang, Alar, Xinjiang 843300, China | | LI Fuqing | College of Agriculture, Tarim University, Key Laboratory of Genetic Improvement and Efficient Production of Characteristic Crops in Arid Areas of Southern Xinjiang, Research Center of Oasis Agricultural Resources and Environment of Southern Xinjiang, Alar, Xinjiang 843300, China | | SHI Jing | College of Agriculture, Tarim University, Key Laboratory of Genetic Improvement and Efficient Production of Characteristic Crops in Arid Areas of Southern Xinjiang, Research Center of Oasis Agricultural Resources and Environment of Southern Xinjiang, Alar, Xinjiang 843300, China | | GAO Ju | College of Agriculture, Tarim University, Key Laboratory of Genetic Improvement and Efficient Production of Characteristic Crops in Arid Areas of Southern Xinjiang, Research Center of Oasis Agricultural Resources and Environment of Southern Xinjiang, Alar, Xinjiang 843300, China | | SHEN Dongyan | College of Agriculture, Tarim University, Key Laboratory of Genetic Improvement and Efficient Production of Characteristic Crops in Arid Areas of Southern Xinjiang, Research Center of Oasis Agricultural Resources and Environment of Southern Xinjiang, Alar, Xinjiang 843300, China | | LI Keyuan | College of Agriculture, Tarim University, Key Laboratory of Genetic Improvement and Efficient Production of Characteristic Crops in Arid Areas of Southern Xinjiang, Research Center of Oasis Agricultural Resources and Environment of Southern Xinjiang, Alar, Xinjiang 843300, China |
|
| Hits: 3 |
| Download times: 3 |
| Abstract: |
| By applying different irrigation gradients, data on spectral reflectance, chlorophyll density, and canopy net photosynthetic rate (Pn) were collected across five growth stages of cotton: budding stage, beginning flower stage, full blossom stage, blossing and boll\|forming stage, and full bloom stage. A canopy photosynthetic rate prediction model was developed using support vector machine (SVM) and random forest (RF) algorithms, incorporating both chlorophyll\|fused and non\|fused data. The results demonstrated a positive correlation between chlorophyll density and the net photosynthetic rate under water stress. The CARS + SPA algorithm was employed to repeatedly perform feature band screening, achieving a remarkable dimension reduction effect and high efficiency in eliminating redundant bands. The feature bands were 332, 347, 416, 466, 672, 695, 711, 733, 752, 848, 954 nm and 1 069 nm in full blossom stage. The monitoring results of the model showed that the model fitting degree of the fused chlorophyll data was better than that of the unfused chlorophyll data. Compared with the estimation ability and model accuracy of different models, the random forest (RF) model was superior to the support vector machine (SVM) model. The R2 of the calibration set of the RF model fusing chlorophyll density in the five growth periods were 0.659, 0.676, 0.808, 0.744 and 0.633, respectively, and the R2 of the validation set were 0.635,0.675,0.786,0.725 and 0.627, respectively. Compared with the model without chlorophyll density data, the R2 of the calibration set increased by 5.59% on average, the RMSE decreased by 2.92% on average, and the RPD increased by 7.26% on average. The average R2 of the validation set was 4.12% higher than that of the unfused chlorophyll data, the RMSE was reduced by 1.64% on average, and the RPD was increased by 5.27% on average. The analysis demonstrated that the spectral estimation model of the cotton canopy photosynthetic rate, integrated with chlorophyll density data, exhibits superior fitting accuracy and stability. |
|
|
|