New Statistical Model to Predict Summer Temperatures and Climate Anomalies

Research groups from all over the world are involved in developing models for climate prediction in a different time (months/years) and spatial (regional/global) scales using known data from past and scientific understanding about the physical and dynamic processes of climate evolution. Recently, researchers at the Indian Institute of Science Education and Research (IISER), Bhopal have developed a statistical model that can predict summer temperatures and extended climate anomalies in India, using the weather data from the preceding winter.

Funded by the Department of Science and Technology (DST), Government of India, the IISER Bhopal team’s model can predict the temperature of an Indian summer season (March-April-May or MAM) using weather data from the previous winter (Dec-Jan-Feb).

The model, apart from predicting the summer temperatures, has also helped in understanding the relationships among various weather parameters and how they have dynamically co-evolved over the past 69 years, said IISER, Bhopal statement.

The research was led by Dr. Pankaj Kumar, Assistant Professor, Department of Earth and Environmental Sciences, IISER Bhopal. The model development and results of the prediction studies have been recently published in the International Journal of Climatology, in a paper co-authored by Dr. Pankaj Kumar and his research scholar Aditya Kumar Dubey.

Highlighting the need for this research, Dr. Pankaj Kumar said, “With climate change and global warming being increasingly recognized as a threat to the ecosystem, socio-economy and, perhaps, life itself, it is important to understand and be able to predict seasonal patterns for better preparedness.”

Research groups from all over the world are involved in developing models for climate prediction in a different time (months/years) and spatial (regional/global) scales using known data from past and scientific understanding about the physical and dynamic processes of climate evolution. Recently, researchers at the Indian Institute of Science Education and Research (IISER), Bhopal have developed a statistical model that can predict summer temperatures and extended climate anomalies in India, using the weather data from the preceding winter.

“We have used a multi-linear statistical technique called Canonical Correlation Analysis to predict summer temperatures and understand the relationships among the various weather parameters”, explained Mr. Aditya Kumar Dubey, Research Scholar.

The researchers have used parameters such as the sea surface temperature, sea level pressure, zonal wind, precipitation, and maximum, minimum, and average air temperatures from the previous winter, to predict the summer temperatures throughout India.

“We have found that the summer temperature has seen a significant increase, especially in North India, during recent decades,” said the lead researcher.

The researchers have also shown that the summer temperature predictability is better for South India than North, due to the former’s proximity to the ocean and the greater impact of the sea surface temperature on summer heat in the subcontinent.

Because of the effect of the sea surface temperature, South India has been found to be warmer during El-Niño years and cooler during La-Nina. The North Indian summer, on the other hand, is affected by the high pressure and circulation systems at upper levels (~5.5-12.5 Km), which form a heat dome and lead to adiabatic heating, thereby pushing up the summer temperature irrespective of the El-Niño or La-Nina effect.

The model by the IISER team has been able to predict MAM temperatures a season ahead. The scientists have considered the role of all possible parameters in developing the predictive model and plan to elucidate the mechanisms behind their interplay.

“Timely, actionable and reliable climate prediction is crucial for policymaking to help manage development opportunities and risks, and for adaptation and mitigation activities,” said Dr. Kumar on the importance of work such as theirs on predictive climate modeling.