The weather which is affected by many factors has changeable and uncertain,Aerosol content, moisture content,cloud cover, temperature, and other factors have a keen effect on the atmospheric electric field. Under different weather conditions,atmospheric electric field exhibit different characteristics. A weather phenomenon recognition algorithm is put forward based on the characteristics of the atmospheric electric field. Atmospheric electric field amplitude domain, frequency domain characteristics are extracted by the use of statistical methods and wavelet energy spectrum analysis and then normalized,and finally are trained by using BP neural network technology features,weather phenomena recognition model is established. Experimental results show that characteristics of atmospheric electric field are helpful to understand the relationship between climate change and atmospheric electric field. The algorithm can achieve the recognition of sunny,cloudy,rainy and thunderstorms weather phenomena. These works are of great significance to promote the automatic ground meteorological observation of all elements.