A review of Weather Forecasting Using CNN and Cloud

Computer Engineering

Weather conditions play a major role in our daily life of a farmer. It affects the productivity of the work and results in the loss. It plays important in getting good productivity from the crops he grows by getting early warning and insights of what will be the weather conditions so that he can be ready and save his crops. There are lot of websites which keep daily weather data which if used properly can give correct insight of the weather behaviour in a specific month. So, the current data with the historical data can prove to be very helpful in developing a weather forecast system for a specific area using latest technologies such as Artificial Intelligence, Cloud computing and Mobile computing together and give a better prediction to a user. So, we are developing a system where a mobile app will be provided to the user side from where he can register and authenticate himself first. Then he will enter a place name and address and send the weather forecast prediction request. This request is then forwarded to the admin which will be a desktop application. The admin will then view all the requests and then select a request. The address will then be converted to latitude and longitude using reverse geocoding and the historical data for this location will be downloaded using worldweatheronline.com API. Then this data which is in JSON format will then be converted to a training dataset. This training dataset will then be used to train the deep learning algorithm CNN which will have various layers in its architecture. Then the CNN will predict the weather forecast which will be sent to the app user automatically. All the communication between an admin and user will be handled using Google cloud platform.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.