| 陈铭杰,何昊,柯浩楠,曹琰梅,史家宾,屈珂樨,赵婧,李琪,胡正华.CO2浓度和温度升高下不同类型水稻高光谱特征及叶绿素含量反演对比分析[J].农业环境科学学报,2025,44(5):1148-1159. |
| CO2浓度和温度升高下不同类型水稻高光谱特征及叶绿素含量反演对比分析 |
| Hyperspectral characteristics and chlorophyll content inversion of different rice types under elevated CO2 concentrations and temperature |
| 投稿时间：2024-07-18 |
| DOI：10.11654/jaes.2024-0613 |
| 中文关键词: CO2浓度 温度 水稻 叶绿素含量 光谱参数 |
| 英文关键词: CO2 concentration temperature rice chlorophyll content spectral parameter |
| 基金项目:江苏省碳达峰碳中和专项（BE2022312）；国家自然科学基金项目（42375114，42071023） |
|
| 摘要点击次数: 197 |
| 全文下载次数: 240 |
| 中文摘要: |
| 为了探究大气CO2浓度和温度升高对粳稻和籼稻的高光谱特征及叶绿素含量的影响，本研究以粳稻“金香玉1号”和籼稻“扬稻6号”为试验品种，通过开顶式气室（OTC）进行田间试验，设置四种处理：对照（环境CO2浓度和气温，CK）、CO2浓度比CK升高200 μmol·mol-1（C+）、温度比CK升高2℃（T+）、CO2浓度和温度共同升高（C+T+），测定各处理下水稻叶片的叶绿素相对含量（SPAD）和高光谱反射率，对原始光谱进行一阶导数变换，比较粳稻和籼稻的高光谱特征，并采用一元和多元方法构建叶绿素含量的最佳估算模型。结果表明：C+处理使SPAD值增加2.8%~8.8%，而T+处理使其降低4.4%~11.1%，且T+处理下粳稻“金香玉1号”的叶绿素含量下降更为明显。CO2浓度和温度处理虽未改变水稻冠层原始光谱曲线的波形，但显著影响了反射率。水稻一阶导数光谱反射率呈现“单峰”或“双峰”特征，各处理的粳稻在整个生长季中主峰位置均出现蓝移，而籼稻在T+和C+T+处理下呈现先红移后蓝移。粳稻的最优叶绿素反演模型为支持向量回归（SVR），其决定系数R2为0.63，均方根误差（RMSE）为2.94；籼稻的最优叶绿素反演模型为BP神经网络，其R2为0.66，RMSE为2.73。综上，在CO2浓度和温度升高条件下，粳稻和籼稻的高光谱特征存在显著差异，且多元模型在叶绿素含量反演中表现出更优的估算效果。 |
| 英文摘要: |
| To explore the effects of elevated atmospheric CO2 concentration and increased temperature on the hyperspectral characteristics and chlorophyll content of different types of rice, the Japonica variety ‘Jinxiangyu No. 1’ and the Indica variety ‘Yangdao No. 6’ were selected as test subjects. Field experiments were conducted using open-top chambers（OTC）with four treatments：control group（CK）, elevated CO2 concentration group（C+）, elevated temperature group（T+）, and combined elevated CO2 concentration and temperature group （C+T+）. The relative chlorophyll content（SPAD）and hyperspectral reflectance of rice leaves were measured under each treatment. The original spectra were subjected to first derivative transformation to compare the hyperspectral characteristics of Japonica and Indica rice, and the optimal estimation models for chlorophyll content were constructed using univariate and multivariate methods. The results showed that：The C + treatment increased the SPAD value by 2.8% - 8.8%, while the T + treatment decreased it by 4.4% - 11.1%, with the chlorophyll content of Japonica rice“Jinxiangyu No. 1”decreasing more significantly under T + treatment. The CO2 concentration and temperature treatments did not change the waveform of the raw spectral curves of the rice canopy but significantly affected the reflectance magnitude. The first-order differential spectral reflectance of rice showed "single-peak" or "double-peak" characteristics. The main peak position of Japonica rice in all treatments exhibited a blue shift throughout the growing season, while Indica rice showed a red shift followed by a blue shift under T+ and C+T+ treatments. The optimal chlorophyll inversion model for Japonica rice was Support Vector Regression （SVR）, with a coefficient of determination（R2）of 0.63 and a root mean square error（RMSE）of 2.94. The optimal chlorophyll inversion model for Indica rice was the BP neural network, with an R2 of 0.66 and an RMSE of 2.73. In summary, under elevated CO2 concentration and temperature conditions, the hyperspectral characteristics of different rice varieties show significant differences, and the multivariate model demonstrats superior estimation performance in chlorophyll content inversion. |
| HTML
查看全文
查看/发表评论 下载PDF阅读器 |
|
|
|
