Machine learning is turning out to be a powerful learning technique across various domains today. This is a small attempt in applying machine learning in the field of meteorology. I have applied different methods to forecast various meterological variables. I hope some time in future machine learning techniques would be fully employed in weather forecasting.
It currently includes,
- Idntify the Madden-Julian Oscillation's (MJO) phase by using OLR and wind data.
- Detecting Cyclone using Convolution Neural Network (CNN).
- Time series prediction of Monsoon by RNN and SARIMA model.
- Classify the month by looking at the Outgoing Longwave Radiation (OLR) map.
- Retrieving winds from the vorticity and divergence fields
Some of the problems currently under development include,
- Time series forecasting of MJO indexes.
- Detecting the droughts by various indices.
- Modelling of 2D and 3D trubulent flow evolution. and many more to follow...