More than a century after becoming the country’s diviner of clouds to forecast the annual monsoon, the India Meteorological Department (IMD) is looking to transfer its massive troves of data into the cloud. The purpose, according to a top official, is to have more centralised control with its regional centres across the country and to be able to make it easier for research arms to access weather-related data.
“Functional efficiency is the biggest reason,” IMD Director-General K.J. Ramesh said. “We’ve already had discussions with a firm and expect to be ready within a year.”
Mr. Ramesh said the IMD would require data storage “in the order of 5 petabytes” that works out to about 500,000 GB or, according to a 2015 study in the Public Library of Sciences, roughly the amount of data that Twitter is expected to consume in 2025.
Cloud-based storage usually means that organisations store their data on third-party servers or access applications without physically storing them on their premises. However, the IMD will locate its cloud on its Delhi campus, a kind of “private” cloud.
Were the IMD’s plans to materialise, it would be among the largest data storage centres in the country. It has already announced that it will be investing Rs. 400 crore in a supercomputer facility, which will process about 10 peta flops per second, and be used to make a dynamical monsoon forecast. These forecasts factor in evolving, global weather conditions to improve the accuracy of monsoon forecasts.
Though companies such as Amazon and Microsoft lease out storage services through the cloud and have massive data storage facilities, these aren’t publicly disclosed.
