Predicting Atmospheric Humidity Using Artificial Neural Networks and BeiDou Satellite Navigation Data

Atmospheric humidity information is very important basic data in scientific research such as weather forecasting and climate change assessment. However, because there are many factors affecting atmospheric humidity, it is difficult to predict it, and there is less work. We found that the satellite data of the Beidou satellite navigation system can have a greater impact on the predicted atmospheric humidity. Therefore, this paper combines the current hot artificial intelligence and neural network algorithms and uses the data obtained from the Beidou satellite navigation system in Wuhan to train the network, so as to achieve the purpose of predicting the actual atmospheric humidity information. The prediction results show that the use of neural networks and BDS satellite data to predict atmospheric humidity can achieve relatively high accuracy, and the mean absolute error (MAE) obtained by comparing with the real data is basically the same. Compared to the previous MAE without BDS satellite data.