Estimation of nitrate nitrogen content in cotton petioles under drip irrigation based on spectral indices and wavelet neural network

Abstract Background: Estimating nitrate nitrogen (NO3--N) content in petioles is one of the key approaches for monitoring nitrogen (N) nutrition in crops. Rapid, non-destructive, and accurate monitoring of great NO3--N content in cotton petioles under drip irrigation is of great significance.Methods: NO3--N content in cotton petioles under drip irrigation and the corresponding canopy spectral reflectance of cotton plants grown in experimental plots under various N application levels were analyzed. The correlations among ‘trilateral parameters’ and six vegetation indices, and NO3--N content in petioles were determined. A traditional regression model of NO3--N content in cotton petioles under drip irrigation was established, and a wavelet neural network (WNN) model with different index numbers was developed. The WNN model was verified using independent data, and compared with the random forest algorithm , radial basis function neural network and back propagation neural network.Results: Based on the analyses of ‘trilateral parameters’ and petiole NO3--N content, blue edge amplitude (Db) and blue edge area (SDb) of the blue edge parameters exhibited a strong positive correlation with petiole NO3--N content, and the correlation coefficients was 0.90. Among the blue edge parameters, the coefficient of determination (R2) of the Db polynomial regression equation and petiole NO3--N content was the highest (R2 = 0.89), while the root mean square error (RMSE) of the linear regression model was the lowest (RMSE = 1.04). R2 value of the traditional regression model developed using blue edge parameters and petiole NO3--N content significantly increased, while RMSE value decreased when compared with those of the red edge and yellow edge parameters. Analyses results of the vegetation index developed using original spectral reflectance data and the vegetation index developed using the first set of derivative spectral reflectance data and petiole NO3--N content, revealed that the first derivative vegetation index, normalized difference spectral index (ND705) exhibited a strong negative correlation, with a correlation coefficient of -0.90. The first derivative vegetation index, ND705 and petiole NO3--N content index regression equation had the highest coefficient of determination (R2 = 0.83), while the first derivative vegetation index, red edge model index (CIred-edge) and petiole NO3--N content linear regression equation had the lowest RMSE = 0.92. R2 value of the traditional regression equation for the first derivative vegetation index and petiole NO3--N content significantly increased, while the RMSE value decreased when compared with the original spectral vegetation index. After conducting correlation analyses and developing traditional regression models, Db and SDb of the blue edge parameters, and the first derivative vegetation index, ND705 and CIred-edge were used to develop a WNN model. The model based on blue edge parameters had R2 of 0.88, RMSE of 0.74g/L and mean absolute error (MAE) of 0.58 g/L, the R2 value was 8.6% higher than the R2 the first derivative vegetation index model, in which RMSE and MAE reduced by 18.7% and 20.5%, respectively. The model was tested using independent verification data, and which revealed that the R2 value of the model was 0.88, RMSE was 0.65g/L, and MAE was 0.47g/L based on the blue edge parameters, predicted value of WNN, and true value of the verification model, which was superior other models. The performance of the WNN model based on the blue edge parameters improved by 7.3%, and RMSE and MAE reduced by 25.2% and 30.9%, respectively when compared with those of the vegetation index model.Conclusion: The present study demonstrated that an inexpensive approach consisting of WNN algorithm and spectrum can be used to enhance the accuracy of NO3--N content estimation in cotton petioles under drip irrigation, which reflects their practical application potential.