The 2015 drought in Maharashtra’s Marathwada remains one of India’s most devastating humanitarian and agrarian crises. Triggered by consecutive years of deficit rainfall and extreme heat, the region recorded little to no-rainfall in 2014 and 2015, starving rainfed crops in the critical sowing period. Unusually high temperatures worsened the crisis, while the extensive cultivation of water-intensive crops like sugarcane severely depleted local groundwater reserves.
As a result, Kharif crops were devastated, with agricultural yields dropping by more than 50% across all affected districts. This environmental calamity led to wide-scale financial distress, tragically resulting in over 1000+ farmer suicides and forcing mass migrations in search of food, employment, and drinking water.
To prevent such catastrophic surprises, the scientific community has been working to build better early-warning systems. In recognition of his breakthrough work in predicting these exact climate emergencies, Prof. Karthikeyan Lanka has been awarded the National Academy of Sciences, India (NASI) Young Scientist Award 2026. As an Associate Professor at the Center of Studies in Resources Engineering, IIT Bombay, Prof. Lanka leads the Ecohydrology Group. His team studies land-surface water processes and how the ground interacts with the atmosphere across both local and macro scales.
Strong Academic Roots
Prof. Lanka completed his Bachelor’s Degree in Civil Engineering from Shanmugha Arts, Science, Technology and Research Academy (SASTRA) University, Thanjavur, in 2011, where he was a merit scholarship holder throughout his tenure. Notably, he scored the highest marks in the GATE 2010 exam among all final-year and pre-final-year Civil Engineering students at the university.
He went on to earn a Master’s Degree in Civil Engineering in 2013 from the Indian Institute of Science (IISc), where he received the prestigious Prof. P.S. Narayana Medal for the Best Thesis. After completing his Ph.D. at IISc, Prof. Lanka worked as a Postdoctoral Researcher at Clemson University before joining IIT Bombay in 2019 as an Assistant Professor.
About His Work
In a major collaborative breakthrough, a research team including Prof. Lanka, Saurabh Verma, and Prof. Subimal Ghosh developed a highly accurate AI framework that predicts farming droughts one to three months before they strike, giving communities vital time to prepare.
To achieve this, the team developed a specialized AI architecture known as a Stacked Long Short-Term Memory (SLSTM) model, a deep learning system specifically designed to track complex, changing data over time. Their framework measures drought severity using the Soil Moisture Percentile (SMP), an indicator that tracks exactly how dry the dirt is compared to normal historical levels. To make accurate predictions, the model connects local conditions like recent ground temperature and humidity, known as local antecedent land-atmosphere variables, with global climate forces like Sea Surface Temperature anomalies, which are unusual temperature shifts in distant oceans that alter global weather.
Because major droughts are rare, standard AI systems often struggle to predict them. To solve this, the team invented a specialized mathematical coaching system called an extreme-aware custom loss function. This function heavily penalizes the AI if it gets the timing wrong or misses a severe weather shift, forcing the model to adapt specifically to extreme dry spells.
The scientists trained this system using historical weather records from 1981 to 2015 and tested it against real-world events between 2015 and 2022. It easily outperformed traditional forecasting tools, showing exceptional accuracy in spotting major droughts across Northern, Southern, and Eastern India a month in advance. While looking further ahead is naturally more challenging, the system remains incredibly reliable and cautious, successfully minimizing false alarms without overpredicting disasters that might never happen.
How Important It Has Become
Prof. Lanka regularly posts these climate forecasts on his LinkedIn page, reflecting the deep dedication and enthusiasm he brings to his research.
As climate change accelerates, extreme weather has become increasingly unpredictable and dangerous across the subcontinent, evidenced by both India’s volatile monsoon shifts and recent devastating flash floods in neighboring Nepal. In this era of climate uncertainty, society must not only adopt more responsible habits but actively work toward a sustainable future. The breakthrough achieved by Prof. Lanka and his collaborators stands as a perfect example of how rigorous education and scientific curiosity can directly protect human lives and livelihoods.
Muskan Singh
