Prediction model based on neural network and complex plane analysis: A case study of agricultural carbon emissions in Henan Province

Junjie Hao¹Yisha Pan^2*

• ¹Chongqing Chemical Industry Vocational College, Chongqing, China
• ²School of Public Policy and Management, China University of Mining and Technology, Xuzhou, China

Controlling carbon dioxide emissions and taking the road of green development are inevitable choices for the sustainable development of the earth. The prediction of agricultural carbon emissions is of guiding significance to accelerate the pace of reducing carbon emissions and is beneficial to guide green technology innovation. Based on six influencing factors such as chemical fertilizer, pesticide and agricultural film, this paper uses neural network model and nonlinear surface fitting method to carry out regression and prediction of total carbon emissions, and the conclusions are as follows: 1) In carbon emission regression prediction, the multilayer perceptron model exhibits a smaller absolute residual error and significantly higher accuracy (R²=0.998) compared to the radial basis function model (R²=0.933), making it more suitable for agricultural carbon emission forecasting; 2) The Gaussian multi-modal fitting method was employed to predict the time-varying values of factors influencing agricultural carbon emissions (with all R² values exceeding 0.9), which enables further prediction of agricultural carbon emissions using the model. These prediction results offer data-based scientific support for promoting the "green production" concept, thereby enhancing the persuasiveness of ecological education. 3) The results of the neural network and the nonlinear surface fitting method show that the carbon emission in Henan Province changes quadratically with time. From 2001 to 2030, the total agricultural carbon emission in Henan Province can shows the characteristics of rapid growth, stable growth and fluctuation decline. The predicted results can serve as a theoretical reference for advancing green agricultural development in Henan Province. Simultaneously, they provide empirical support for promoting the green production concept and disseminating low-carbon policies. This contributes to fostering an ecological governance closed loop characterized by "consciousness - voluntary action - effect translation" ultimately aiding in the synergistic enhancement of ecological and social benefits.