Forecasting, with precision, the location of landfall and the height of surge of cyclonic storms prevailing over any ocean basin is very important to cope with the associated disasters. The main objective of the present research is to develop models to forecast the exact location of landfall and the surge heights of tropical cyclones over North Indian Ocean. Artificial neural network (ANN) model is developed for forecasting the location (latitude–longitude) of landfall, and neuro-fuzzy coupled (NFC) model is developed for forecasting the surge heights. The sea surface temperature, vertical wind, minimum pressure at the centre of the cyclones, maximum wind speed and pressure drop at the centre are taken to form the input matrix of both the models. The result shows that position of landfall can be predicted with high accuracy with ANN model. The result further reveals that the prediction error with NFC model in forecasting the surge height is 2.448 % with 6-h lead time, whereas the error is observed to increase with the increase in lead time. However, for forecasting the surge height with 18-h lead time, the accuracy is observed to be 11.4 %. The skill of both the models is verified through validation with the observations of cyclones Jal (2010) and Thane (2011). © 2014, Springer Science+Business Media Dordrecht.
|Publisher||Kluwer Academic Publishers|