Dr LS Rathore, Director General, India Meteorological Department, talks about how they managed to get it right in Cyclone Phailin.
In this Idea Exchange, Dr LS Rathore, Director General, India Meteorological Department, talks about how they managed to get it right in Cyclone Phailin and why it couldn’t have been the same in Uttarakhand, why predicting weather phenomenon remains a difficult task, and what has changed over the years. This session was moderated by Senior Editor Anil Sasi
Dr LS Rathore:The process of prediction begins with defining the initial conditions. Over time, in both defining the initial conditions and the way we give our predictions, there has been a paradigm shift. To define the initial conditions, we need observations and we don’t have observations for large parts of the globe. Mainly, we have 2/3rd (of the Earth as) oceans, we have mountains, we have deserts, we have poor countries which are not able to operate or manage observation systems… In the process we have some amount of error. Often we are asked as to why in tropics the predictions are not as accurate as they are in extra-tropical regions. It is primarily because the intrinsic property in tropics is that weather is driven more by energy forces and, therefore, it can fall all of a sudden, it could be very intense in a limited period of time, it could move in any direction. In an extra-tropical region, the weather is very stable. Once it forms, it will last for a week and it is always uni-directional, west to east, and it is over a huge spatial domain… In tropics the predictability in summer is five days and in winter seven days. In extra-tropics, it is for two weeks, at times three weeks….
The whole emphasis in the gamut of weather prediction is two-fold one is to enhance the skill of forecast, i.e. to minimise error, and second to extend the temporal range. That means you need to have a weather forecast in nowcast (for the next six hours); short range; medium range, that is three to 10 days; extended range, a couple of weeks or a month; and the season as a whole. And now we are attempting long-time climate prediction, for five, 10, 20 years….