Agroup of scientists working on mathematical models to forecast the surge of coronavirus cases said Sunday they could not predict the exact trajectory of the devastating second wave as the virus dynamics and its transmissibility changed substantially over time

. A statement issued by the Department of Science and Technology, and signed by IIT Kanpur Professor Manindra Agrawal, Integrated Defense Staff Deputy Chief Madhuri Kanitkar and IIT Hyderabad Professor M Vidyasagar, said mathematical models had predicted a second wave of coronavirus and its peak in the third week of April with daily cases of around I lakh.

They also rejected reports that the scientists working on the SUTRA model cautioned in March about the second wave, but that their warning was ignored. ''This is incorrect,'' they said in the statement.

Last year, the government had formed a group of scientists, mathematicians and experts to forecast the surge of cases. They said the nature of the virus has been changing very rapidly, and in such a scenario, any prediction must be continually readjusted, sometimes almost daily.

''We are working closely with the government and our inputs have always been received positively. While we could not predict the exact nature of the second wave earlier, we continue our efforts to better estimate its future trajectory,'' they said.

They added that the government sought its inputs on April 2 when they predicted a peak would come around the third week of that month.

''A meeting was called on 2nd April to seek our inputs by one of the very senior officers of the government coordinating the national pandemic response. We indicated that the SUTRA model predicted the second wave to peak by the third week of April and to stay most likely around 1 lakh daily cases,'' the scientists said in the statement.

However, the daily figures actually reported were thrice the predicted numbers. On April 15, India reported 2,00,739 cases, while it recorded 3,14,835 cases on April 22. Breaching its all previous records, the country hit a record daily high with over 4 lakh new infections on May 1.

''Clearly, the model predictions in this instance were incorrect because of the reasons below,'' they said, listing reasons why. A mathematical model, they said, can only predict the future with some certainty so long as virus dynamics and its transmissibility do not change substantially over time.

Mathematical models can also provide a mechanism to predicting alternate scenarios corresponding to various policy decisions such as non-pharmaceutical interventions, they added.