Real time wave and wind forecasting system for the indian coastline

Real time forecasts of wind and waves are useful in taking many operation oriented decisions in the ocean. When site specific forecasts are required time series modeling can be viewed as advantageous over elaborate numerical methods. This paper discusses time series models based on the soft computing approaches of genetic programming (GP) artificial neural network (ANN) to obtain real time forecasts of wind speed, its direction as well as the significant wave height at different locations along the Indian coastline where continuous wave buoy observations get collected. All the developed models over various locations have been integrated into a graphical user interface (GUI) to facilitate web based real time implementation of these models. The GUI starts with a figure showing locations of buoys. The user has to click on the concerned location where it is desired to have the forecasts for next 24 hours. The appropriate and calibrated ANN and GP models will then come into background, linking them with the most recently observed data and will accordingly yield the forecasts immediately