NRSC/ISRO and IISc join hands for research to assess CO2 source / sinks and predicting extreme rainfall events
Home / NRSC/ISRO and IISc join hands for research to assess CO2 source / sinks and predicting extreme rainfall events

June 10, 2025

On May 13, 2025, the NRSC / ISRO and IISc representatives signed the Joint Project Implementation Plan (JPIP) documents for the two projects at the IISc campus. These projects are ‘Assimilation of Ground & Space-based CO2 data to Assess CO2 Source/Sinks over India-Earth System model’ and ‘Development of a Machine Learning (ML) tool for predicting city-scale Extreme Rainfall Events using a hybrid Physics-AI approach’.

NRSC/ISRO and IISc join hands for research to assess CO2 source / sinks and predicting extreme rainfall events

JPIP signing by the ISRO / NRSC and IISc representative at IISc Bengaluru on 13 May 2025

Following are the research details of these projects:

  1. Assimilation of Ground- and Space-Based CO2 Observations to Evaluate Carbon Sources and Sinks Over India Using an Earth System Model:Terrestrial ecosystems, particularly vegetation and soils, act as major carbon reservoirs by absorbing atmospheric CO2. Accurately quantifying the carbon sequestration potential of these ecosystems is essential in the context of the Paris Agreement and the global goal of limiting temperature rise below 2°C. NRSC has deployed CO2 eddy-covariance flux towers across various Indian biomes and climatic regions to monitor CO2 exchanges. This project focuses on estimating Net Ecosystem Exchange (NEE) through data-driven modelling approaches, aiming to deepen scientific insight into carbon dynamics and to aid policymakers in identifying carbon source and sink hotspots across India.
  2. Development of a Hybrid Physics-AI Model for Predicting Urban-Scale Extreme Rainfall Events: A notable surge in extreme precipitation events has been observed in recent years, sometimes exceeding the seasonal average rainfall, causing significant climatic and societal impacts. Predicting such extremes remains a complex challenge. Conventional forecasting methods based on Numerical Weather Prediction (NWP) models face inherent limitations. This study proposes to develop a hybrid modelling framework that integrates Machine Learning (ML) with physical models to improve the prediction accuracy of extreme rainfall events at the city scale. The hybrid approach complements traditional NWP techniques, enhancing their precision and operational utility in forecasting.

Director NRSC-ISRO, Director FSID-IISc, Director EDPO-ISRO-HQ, Dy. Director, ECSA-NRSC, Dy. Director MSA-NRSC, Dy. Director BGWAA-NRSC, Head Project Management Division-MSA, and the project Principal Investigators from the centres were present during the event.