The AI/ML Model At IIT Delhi Says That 2023 Will Be A Normal Rainfall Year

A new machine learning model made by an international team of researchers led by the Indian Institutes of Technology (IIT) Delhi says that India will have a normal monsoon in 2023.

The model said that the next monsoon season would have an All India Summer Monsoon Rainfall (AISMR) of about 790 mm, which means a normal monsoon.

It made very accurate predictions 61.9% of the time during the test period, which was from 2002 to 2022. This depends on whether or not the model can predict the AISMR to within +/- 5% of the real values that are seen each year.

Prof. Saroj K. Mishra, of the DST Centre of Excellence in Climate Modelling and Professor at the Centre for Atmospheric Sciences, IIT Delhi, said, “This study is very important for the whole country because an accurate monsoon forecast well in advance is crucial for making important decisions in many socioeconomic sectors, such as agriculture, energy, water resources, disaster management, and health.”

Prof. Mishra also said that the data-driven methods will be improved so that monsoon rainfall predictions can be made for each state. This will make them more useful for regional applications. The AI/ML model does a better job of predicting monsoons than the physical models that are used in the country right now.

The AI/ML model can make an estimate up to a few months in advance, depending on how soon the Nino3.4 index and Indian Ocean Dipole forecast are available, and it can be changed based on how they change.

So, the data-driven models are more adaptable to inputs and can better describe the nonlinear relationships between the factors that affect monsoons. They also take less time to run.

The expert said that a small group of people using these models on a personal computer in a short amount of time can make a more accurate prediction of monsoon rains than traditional physical models, which require a lot of time and resources.