New Delhi, Jan. 16 -- India Meteorological Department will be launching Mission Mausam phase II by April this year, M Ravichandran, secretary, ministry of earth sciences said on Thursday, on the occasion of 151st Foundation Day of IMD. This would involve a massive upgradation of observational systems over the land and the ocean, he said. One of the most significant aspects of this would be installation of 200 automatic weather stations in important cities including Delhi NCR, Mumbai and Chennai. There are around 1000 AWS in India. The upgradation will begin with installation of 20 AWS in Delhi this month. Another concern that Mission Mausam phase II will try to counter is ocean observations. Presently 50% of the ocean observations are being provided by US institutions. "Because of the current geopolitical situation, we need to very rapidly upgrade ocean observations because our seasonal predictions especially for monsoon are completely dependent on ocean parameters. We are also asking other countries to upgrade their ocean observations so we are not dependent on one country," said Ravichandran. He also pointed that monsoon forecasts have improved from 7% error margin to 2% mainly because of ocean observations. "In this regard we need deep ocean parameters, ocean parameters up to 100 km from coast, coastal radars etc. For short term forecasts we need AWS whereas for medium and long term forecasts, ocean observations are critical because the ocean has that memory," he said. To improve the forecast of extreme rainfall and other weather events in cities, 200 new AWS will be installed. "There was a time when there would be uniform rainfall over a 100 km stretch of area. Now that has reduced to 10 km and very soon we will see highly localised events say over 1 km. So we need 1 km resolution observations. That is why the number of AWS is being increased," explained Ravichandran. During Mission Mausam I, IMD increased the number of radars to 46 and unveiled Bharat Forecasting System-a high-resolution (6 km) numerical global model for operational forecasts up to the panchayat level, which is expected to help address extreme rainfall events, among other things. Ravichandran said IMD has found mixed results with AI models. "They have definitely improved forecasts but we cannot depend on AI alone." "For the phase II of mission mausam, we need a very huge number of observations in terms of gaps we need to fill wherever the observations are not available. This is not only from the ground but also in the upper air observations. We need to use the aircraft observations. And also we need to develop some mechanism to measure the boundary layer observations by drones. And more importantly it is not only for the land but also we need to cover the ocean. And also we need to go to coastal regions and polar regions. So, the first important task is we need to augment more and more observations," Ravichandran said during the 151st Foundation Day. "The second important task is basically we need to have more challenges in physics. In terms of convection representation in the model. The forecast in the tropical weather is not so easy. For that what we need is observation to understand the process and try to better model it. And the third one is data assimilation," he said....