Understanding the complex relationship between climate change and traffic thus becomes mandatory for effective urban planning and transport management. This paper aims at providing a new method based on deep learning for predicting the traffic impact on the basis of weather condition. The deep regression-based analysis aims at combining the weather data with the imagery of traffic from several cameras in a bid to predict the effect of climate change on the volume of traffic. The proposed model is validated over a total of two datasets, which include the weather data and real-time traffic imagery for training and validation purposes. The deep learning algorithm will be used to determine the spatio-temporal complexity between climate change and traffic congestion, hence ensuring accurate estimation in climatic conditions of the dynamics of the traffic condition. Results obtained experimentally demonstrate that the proposed method is more efficient and predictable than all previous ones up to mean squared error (MSE) 0.074. By integrating the weather and camera network, the model can sub-data record between the weather conditions and traffic flow, which in turn gives an accurate forecast.