Method of Fitting Temperature Changes during Foehn Process in Mountainous Areas of Southern Sichuan Based on BP Neural Network

Using the good function simulation ability of BP neural network, the temperature, time, longitude, latitude, and altitude of the foehn process occurring in the mountainous areas of southern Sichuan are constructed as the suitable training dataset for BP neural network training. After error analysis of the trained BP neural network, the analysis results show that the established BP neural network model can well simulate the temperature changes during the foehn process, and can be used to identify whether the temperature rise in the foehn phenomenon is abnormal or suspect, providing a new idea for temperature quality control. The trained BP neural network can be applied to interpolate and analyze temperature changes, produce high-resolution live grid products, and provide data support for weather forecasting, forecasting, and service businesses.