Every monsoon season, millions of Indian farmers make decisions that will define the next several months of their lives. What to plant. When to irrigate. Whether to take a loan. They also face devastating hazards like floods and droughts, leading to severe stress and the loss of livelihoods. These are not small choices; they are bets made largely in the dark, against a climate that is becoming harder to predict with each passing year.
The numbers tell a sobering story. Crop yields in India are projected to drop by 9% by 2040 compared to 2010 levels. If climate risks continue to go unaddressed, rural incomes could fall by as much as 20-25% in the medium term. Heatwaves, floods, and coastal hazards are no longer periodic disruptions; they are accelerating, intensifying, and converging in ways that existing systems were never designed to absorb. The challenge is looking for teams that can build the tools India needs to get ahead of these events.
The gap nobody talks about
India's disaster response systems, though extensive, were not designed to manage the scale and speed of today's climate shocks. For example, crop insurance under PMFBY has processed over 78 crore farmer applications, and yet delays and hazard-trigger mismatches highlight systemic constraints. The protection gap is widening as climate variability intensifies.
What makes this harder to fix is a critical missing piece in the data ecosystem itself. Climate datasets are scattered across public agencies, private institutions, and research bodies, each speaking a different technical language, structured differently, and largely inaccessible to the people who need them the most. And it is not that climate data in India is scarce; it is just that its integration is broken. Predictive models that do exist are often proprietary, expensive, and held in silos. And there is a shortage of credible, India-specific ground truth data to train and validate the kind of AI-driven models that could actually make a difference at the village level.
The result is a blind spot right in the middle of the timeline that matters most: the next 10-15 years. Long enough for farmers, banks, and governments to plan meaningfully. Short enough that the projections can be actionable and specific. This is exactly the window for which reliable forecasting is most needed, and most absent.
A platform already in the works
NABARD, with technical support from the Gates Foundation and Dalberg Advisors, believes that innovators and startups can change this by developing a national climate stack. And the National Climate Stack Innovation Challenge announced by NABARD aims to do just that.
Climate data has been an agenda for NABARD and it has been working toward a solution through DiCRA, (Data in Climate Resilient Agriculture), a smart web based platform collating and analyzing climate data. . Originally conceived by UNDP under its’ UNDP Labs initiative and now hosted and expanded by NABARD, DiCRA is a standalone climate data portal that makes climate analytics freely available across the ecosystem. It is a strong foundation. But then, it’s a foundation, not a building.
The next step for DiCRA is embedding credible near-term hazard forecasting directly into this infrastructure. That vision translates into a National Climate Stack, a system that integrates climate datasets across institutions through interoperable APIs, layers in AI-driven analytics, and translates raw data into decision-ready projections for farmers, banks, researchers, and policymakers. In doing so, DiCRA can structurally and potentially transition from being a digital public good, essentially a knowledge repository, into a digital public infrastructure that actively powers climate resilience on the ground.
The potential impact will be significant. Farmers could access near-term projections to guide cropping and irrigation decisions in real time. Financial institutions could build more responsive credit and insurance products. Government departments could use the stack to shape smarter, more targeted policy. The same platform could serve as a tool for researchers and climate-focused startups building solutions across agriculture, water management, and rural finance.
What the challenge is asking for
To accelerate this transition, NABARD, the Gates Foundation, and Dalberg Advisors launched the National Climate Stack Innovation Challenge at the Bharat Climate Forum in January 2026, in the presence of the Vice President of India. It is a six-month guided sprint combining scientific rigor with applied experimentation.
"The National Climate Stack Innovation Challenge is an effort to bring the best minds forward and help develop a solution that brings all these data streams together in a seamless manner. The larger objective is to develop a tech solution that truly democratizes climate data in a way that has not been done till now, " says Dr. Shaji Krishnan V., the Chairman of NABARD.
The challenge is not looking for a finished product. Participants must address two focused objectives. First, they must propose a robust methodology for generating localized near-term (10 to 15 year) hazard forecasts by leveraging relevant public datasets alongside DiCRA. Second, they must demonstrate how these forecasts translate into tools through illustrative dashboards and interoperable system architecture that enables real-world deployment.
The challenge follows a three-phase process: initial screening (March to April) to review and shortlist applications; model development (April to May), offering a six to eight week mentored development window; and technical validation and jury selection (May to June), where a Technical Advisory Group will assess scientific rigor before final presentations determine the winners. The top three teams will be felicitated on July 12.
The prizes are meaningful: Rs 15 lakh for first place, Rs 10 lakh for second, and Rs 5 lakh for third, all funded by NABARD. For teams with the ambition to go further, post-challenge pilot opportunities could open through programs like TAF and other relevant schemes.
Who should apply
The challenge is looking for dedicated teams of 3-5 members representing a single organisation or a cross-category consortium. To ensure effective long-term scaling, teams must be affiliated with organizations in the private sector, such as startups in climate-tech, agritech, and fintech, or technology and modeling corporations. We also welcome research institutions like IMD, SAC, NRSC, and ICAR institutes, alongside academic bodies like IITs, IISCs, IIMs, and various Central and State universities.
If you are building in the climate data space or developing AI-based forecasting tools for rural India, this challenge was designed for you. This problem requires depth: technical credibility, an understanding of India's institutional landscape, and the ability to design for real-world deployment. If you have been looking for the right problem to point your expertise at, this may be it. Climate risk in India is not a future threat; it is a present reality. The tools to respond to it need to be built now, and they need to be built well.
For more information and to apply, please visit our sitebefore March 30.
Original Article
(Disclaimer – This post is auto-fetched from publicly available RSS feeds. Original source: Yourstory. All rights belong to the respective publisher.)
