Background

The North Eastern Region (NER) frequently faces landslides, flash floods, road blockages, and slope failures due to heavy rainfall, fragile terrain, and unplanned hill cutting. These incidents often disrupt connectivity, damage infrastructure, delay emergency response, and isolate remote villages for days. Currently, monitoring of vulnerable zones is mostly reactive and dependent on manual reporting. There is limited use of real-time predictive systems for identifying high-risk zones and issuing early warnings to authorities and local communities. With increasing climate vulnerability in the region, there is a need for an AI-enabled real-time monitoring and prediction system that can help authorities take preventive action before disasters occur.

Description

This problem statement proposes the development of an Al-powered early warning and monitoring platform capable of predicting and tracking landslide-prone areas in real time across the North Eastern Region. The solution should

Expected Solution

A scalable Al-based software platform with: Real-time GIS dashboard and risk heatmaps AI/ML-based predictive analytics engine Mobile/web application for field reporting and alerts. Integration with IMD weather APIs, satellite feeds, and sensor data Automated SMS/app-based early warning system Cloud-based architecture with offline sync support for remote regions The solution should improve disaster preparedness, reduce loss of life and infrastructure damage, and strengthen climate-resilient governance in the North Eastern Region.

Background

Urban land administration and cadastral management involve integration of multiple spatial and non-spatial datasets generated from various departments,agencies,and survey mechanisms. Under modern land governance programmes such as the NAKSHA Programme, large volumes of geospatial data are being generated through drone surveys, Orthorectified lmagery (ORl), DSM/DTM datasets, Ground Truthing (GT),GNSS surveys,municipal records,utility databases, and revenue land records. At present,harmonization and integration of these datasets largely depend on manual GIS workflows,which are time-consuming and prone to errors.With increasing availability of Al, GeoAl, and automated spatial processing technologies, there is significant scope for development of an intelligent system capable of automatically integrating and synchronizing multi-source geospatial datasets with feature-extracted cadastral data.

Description

The proposed solution should develop an Al-enabled geospatial integration platform capable of automatically integrating, harmonizing, validating, and synchronizing multiple land-related datasets with Al-generated feature extraction outputs. The system should support integration of

Expected Solution

Suggested Technologies:

Background

Watershed development plays a vital role in sustainable management of land, water, and natural resources, particularly in rural and semi-arid regions of lndia. Effective watershed planning and monitoring require accurate spatial information on land use, drainage patterns, vegetation cover, soil moisture, water bodies, and changes occurring over time. Traditional monitoring approaches often rely on field surveys and manual reporting, which are time-consuming, resource-intensive, and limited in spatial coverage. ln recent years, advancements in Geographic lnformation systems (GlS), Remote Sensing (RS), and geospatial technologies have created new opportunities for scientific watershed assessment and evidence-based decision- making.Geo-coded images, integrated with satellite-based spatial datasets, provide location- specific visual information that can significantly improve watershed monitoring and interpretation. The availability of 30 m spatial resolution satellite data through the SRISHTI-DRISHTI platform offers a valuable opportunity to develop analytical frameworks for visualization, interpretation, and assessment of watershed characteristics. Through thematic mapping and spatial analysis, geo-coded images can support identification of land degradation, water conservation structures, vegetation changes, drainage conditions, and other watershed-related parameters. The proposed study focuses on the application of geospatial techniques for visualization and analysis to interpret geo-coded images for enhancing watershed development outcomes. The study aims to develop a systematic and scalable approach for image-based watershed analysis using GIS and remote sensing tools. By integrating geo-coded imagery with satellite datasets from the SRISHTI-DRISHTI platform, the research seeks to improve planning, monitoring, interpretation, and scientific assessment of watershed interventions in a focused and cost-effective manner.

Description of the Study

Despite significant investments in watershed development programs, effective monitoring and interpretation of watershed activities remain major challenges. Existing assessment methods are often fragmented, dependent on manual observations, and lack spatial integration. Many watershed projects face difficulties in accurately visualizing field conditions, tracking spatial changes, identifying intervention impacts, and generating reliable evidence for decision-making. Although geo-coded images are increasingly being collected during watershed implementation and monitoring, their analytical utilization remains limited. ln many cases, geo-tagged photographs are used only for documentation purposes rather than for integrated spatial analysis and interpretation. There is insufficient use of advanced GIS and remote sensing techniques to systematically visualize, analyse, and interpret these geo-coded datasets in relation to watershed characteristics and satellite-derived information. Furthermore, the absence of standardized visualization frameworks restricts the ability of planners and administrators to derive actionable insights from geo-coded imagery. Challenges also exist in integrating field-level geo-coded images with satellite data, thematic layers, and watershed boundaries for meaningful analysis. Limited technical approaches for image interpretation reduce the effectiveness of watershed monitoring systems and hinder scientific evaluation of land and water resource interventions. The SRISHTI-DRISHTI platform provides an opportunity to address these challenges by offering consistent 30 m spatial resolution satellite data that can support integrated geospatial analysis. However, there is a need to develop specialized methodologies and visualization techniques that can effectively interpret geo-coded images and generate meaningful watershed insights. Therefore, the study aims to bridge this gap through a focused analytical framework combining GlS, remote sensing, thematic mapping, and geo-coded image interpretation for enhanced watershed development outcomes.

Scope of the Study

Table to be Added here

Expected Solutions

The proposed study is expected to provide a structured geospatial framework for visualization and interpretation of geo-coded images in watershed development programs. The key expected solutions include

Background

Accurate forecasting of the Expected Time of Arrival (ETA) for coaching trains is vital for improving passenger satisfaction and operational efficiency in Indian Railways. Currently, ETA is often estimated using static schedules, current delays and in-built recovery times, which may not reflect real-time ground realities such as speed restrictions, congestion, unscheduled stoppages or historical patterns. As a result, passengers, station staff, and downstream logistics services face uncertainty and planning difficulties. With the growing demand for real-time train information and forecast, there is a pressing need to shift towards a data-driven, dynamic ETA prediction system that continuously adapts to actual train running conditions.

Detailed Description

Indian Railways operates a vast network of passenger trains across diverse geographies, weather conditions, and traffic patterns. These coaching trains often face variability in journey times due to multiple real-world factors such as signal halts, congestion on busy routes, delays in preceding trains, temporary speed restrictions, unscheduled maintenance blocks, level crossing gates and operational bottlenecks.Despite this, ETA predictions at intermediate and destination stations are still often based on the train schedule, current delays and in-built recovery times, which lack accuracy and responsiveness.This limitation affects not just passengers but also impacts station planning, crew scheduling, platform allocation, cleaning operations, and feeder transport services. For long-distance trains with multi-day journeys, even a small deviation can cascade and lead to significant uncertainty. In an era where passengers expect real-time updates through mobile apps and station displays, inaccurate or outdated ETA predictions undermine service quality and trust. The challenge is to create a system that can dynamically forecast the ETA of trains at various points in their journey using real-time data feeds. These may include GPS-based location data, signal aspects, average sectional running times, weather conditions, historical delay patterns, and congestion levels on downstream tracks. The system must also be scalable to cover thousands of trains simultaneously and adaptable to the Indian Railways diverse operational zones. It should account for temporal and spatial variability in train performance and continuously refine its predictions using machine learning or statistical models. Such a system can serve as a foundation for better passenger communication, resource planning, and delay management.

Expected Solution

The expected solution is a real-time ETA prediction system for coaching trains using data-driven models.It should integrate live train location data, operational parameters, historical delay trends, and network conditions to forecast arrival times at upcoming stations.The system must dynamically update ETAs in response to real-time events and delays. Machine learning or statistical forecasting techniques should be employed to improve accuracy over time. The solution should feature APIs for integration with mobile apps, station displays, and control room dashboards, ensuring that passengers and staff receive reliable, up-to-date information to support decision-making and planning.

Background

Communities across Jharkhand encounter numerous local challenges related to education,healthcare, agriculture, water management, sanitation, environment, rural livelihoods,accessibility, urban infrastructure, and public service delivery. While citizens are often the first to identify these issues, there is currently no structured mechanism through which they can submit such problems for systematic evaluation and innovation-driven resolution.At the same time, Higher Education lnstitutions (HEIs) possess significant academic expertise,research capabilities, and a large pool of students capable of developing practical solutions.Industries and start-ups also have technical expertise, financial resources, and implementation capabilities that can complement academic research. However, collaboration among citizens,universities, and industry remains largely fragmented and project-specific.The National Education Policy (NEP) 2020 emphasizes experiential learning, multidisciplinary research, innovation, industry collaboration, and community engagement. Establishing a technology-enabled platform that connects societal challenges with academic institutions and industry partners can foster demand-driven innovation while enabling students and researchers to work on real-world problems that create measurable social impact.

Description

Every year, citizens across Jharkhand identiff thousands of local issues that require innovative technological or process-based solutions. These challenges often remain unresolved due to the absence of a centralized platform that enables problem collection, categorization, expert evaluation, institutional assignment, and industry collaboration.There is a need to develop a digital platform capable of: Allowing citizens, community organizations, local bodies, and government agencies to submit societal challenges thiough an intuitive web and mobile interface, supported by photographs, videos, location details, and relevant documents. Automatically categorizing submitted problems based on thematic domains such as education, agriculture, healthcare, water resources, environment, energy, urban development, accessibility, public administration, and rural livelihoods using Al-enabled classification techniques. Routing validated problem statements to appropriate universities based on their academic disciplines, research expertise, innovation centres, incubation facilities, and faculty specialization. Enabling universities to evaluate submitted challenges, constitute multidisciplinary student and faculty teams, and prepare solution proposals or research projects. Facilitating collaboration between universities and industry partners, startups, MSMEs,CSR organizations, research laboratories, and innovation ecosystems for mentorship,funding, prototyping, testing and deployment of solutions. Providing workflow management for problem review, institutional allocation, project monitoring, stakeholder communication, milestone tracking, and solution validation. Generating dashboards and analytics for govemment departments to monitor the number of challenges received, domain-wise distribution, institutional participation, industry engagement, project progress, and measurable social outcomes. The platform should support a transparent and scalable innovation ecosystem that transforms community-driven challenges into research, innovation, entrepreneurship, and deployable solutions.

Expected Solution

A comprehensive Societal Innovation Collaboration Portal comprising the following components: A citizen engagement module enabling individuals, community groups, Panchayati Raj Institutions, Urban Local Bodies, and government departments to submit societal challenges with multimedia evidence, geographical location, and supporting information. An Al-enabled problem management module capable of automatically categorizing, prioritizing, deduplication, and routing validated challenges to appropriate universities based on subject expertise and institutional capabilities. A university collaboration module allowing Higher Education Institutions to review assigned challenges, form multidisciplinary project teams, assign faculty mentors, manage project workflows, and submit solution proposals. An industry partnership module facilitating participation by industries, startups, MSMEs,CSR organizations, research institutions, and innovation hubs for mentoring, co-development, funding, prototyping, pilot implementation, and technology transfer. A project lifecycle management system for monitoring milestones, deliverables, approvals, documentation, testing outcomes, intellectual property generation, and implementation status. A visual analytics dashboard providing real-time insights on challenge submissions, university participation, industry collaborations, thematic trends, project completion rates,innovation outcomes, patents, startups created, and community impact across districts and sectors. A notification and communication system enabling seamless interaction among citizens,universities, industry partners, mentors, and government departments throughout the project lifecycle.

Develop a Digital Twin framework for Maitri and Bharati stations integrating infrastructure, energy, logistics and environmental monitoring for efficient remote management.

Background

Weather information is often distributed through multiple portals, bulletins, satellite products, and forecast systems, making it difficult for common users, researchers, disaster managers, and government agencies to quickly obtain actionable insights. There is a need for an intelligent conversational platform that can provide real-time weather information, forecasts, warnings, climate analysis, and decision support in natural language.

Objective

Develop an AI-powered chatbot platform named WeatherGPT that integrates meteorological datasets, forecasting models, and disaster warning systems to provide accurate, contextual, and multilingual weather intelligence through conversational interfaces.

Key Features

Expected Solution

Participants should develop A mobile-based conversational AI platform. Backend integration with meteorological databases, website and APIs. AI/LLM-based query understanding engine. Scalable architecture supporting real-time data ingestion.

Suggested Technology Stack

Python / FastAPI / Node.js MQTT / WIS2.0 / WebSocket LLMs (OpenAI, Llama, Gemini, etc.) GIS tools and weather APIs PostgreSQL / MongoDB Docker / Kubernetes

Expected Outcomes

Faster dissemination of weather information. Improved public accessibility to forecasts. Better disaster preparedness and response. Intelligent weather decision-support system for agriculture, aviation, marine, and urban planning.

Possible Use Cases

Farmers seeking crop-weather advisories. Aviation weather briefing. Flood/cyclone warning dissemination. Smart city weather monitoring. Climate analytics for researchers.

Evaluation Parameters

Accuracy and relevance. Response latency. Multilingual capability. User interface and accessibility. Scalability and innovation. Integration with real-time meteorological systems. Voice-enabled interaction for rural accessibility

Design and develop a scalable National Weather Big Data Analytics Platform capable of collecting and processing real-time weather-related information for India from multiple internet-based sources including social media platforms, public datasets, websites, APIs, and citizen reports. The platform should automatically collect weather related posts and information tagged with #IMD and other relevant weather hashtags, along with metadata such as date & time, city, state, GPS location, photos, videos, and event category, and store the information in a centralized database. The system should leverage big data technologies and open-source tools to support large-scale real-time data ingestion, processing, storage, and visualization. Participants are encouraged to use machine learning and AI-based techniques to identify fake or misleading reports, verify untrusted sources, remove duplicate entries, and automatically categorize weather events such as rainfall, thunderstorms, flooding, heatwaves, fog, dust storms, and strong winds. Develop a web-based dashboard and Admin Panel for monitoring and analysing collected data with features including: Date-wise filtering Event-wise filtering Location-wise filtering Verification status tracking Real-time visualization and analytics

AI/ML-Based Integrated heavy rainfall Early Warning and Inundation Prediction System using Satellite, Radar, observational Weather and numerical weather prediction model data.

AIML based Nowcasting of thunderstorm and lightning using atmospheric observation including multiple radars, satellite, lightning and model data.

Title SkyGuard AI Intelligent Real-Time Anomaly Detection System for Temperature, Pressure, and Humidity Sensors in Automatic Weather Stations

Background

Automatic Weather Stations (AWS) are critical components of modern meteorological observation networks. These stations continuously monitor atmospheric parameters and provide real-time data for weather forecasting, climate monitoring, disaster management, aviation, agriculture, and scientific research.However, AWS observations often contain anomalies caused by sensor malfunction,communication failures, calibration drift, power fluctuations, harsh environmental conditions, and data corruption.Erroneous observations can significantly impact weather forecasting accuracy and decision-making systems. Traditional threshold-based quality control methods are often insufficient for identifying complex or hidden anomalies in meteorological data streams.

Problem Statement

Develop an AI/ML-based intelligent anomaly detection system capable of automatically identifying abnormal, inconsistent, or faulty observations from Automatic Weather Stations in real time using only the following parameters: Temperature (°C) Atmospheric Pressure (hPa) Relative Humidity (%) The system should distinguish between genuine meteorological events and sensor/data anomalies while minimizing false alarms and enabling scalable deployment across large weather observation networks.

Objectives

Detect anomalies in real-time AWS data streams. Identify sensor faults, spikes, frozen values, and communication errors. Learn normal temporal and seasonal patterns of temperature, pressure, and humidity. Perform multivariate consistency analysis among atmospheric parameters. Provide confidence scores and explainable AI-based reasoning for detected anomalies. Predict possible sensor degradation and maintenance requirements. Optionally suggest corrected/imputed values for anomalous observations.

Expected Inputs

Participants may use historical AWS datasets, simulated anomalies, or streaming sensor data containing the following meteorological parameters: Parameter- Unit Temperature °C Atmospheric Pressure hPa Relative Humidity %

Expected Outputs

Real-time anomaly alerts Severity and confidence scores Root-cause classification Visualization dashboard Sensor health status Corrected data estimation (optional) Suggested Technologies Explainable AI (SHAP/LIME) (Preferable) Edge AI for low-power deployment on ESP32

Evaluation Criteria

(To be evaluated in anomaly injected data) Criteria Weightage Innovation & Novelty 25% Detection Accuracy 20% Real-Time Capability 15% Explainability 10% Scalability 10% Practical Deployability 10% Visualization/UI 5% Energy Efficiency 5%

Example Use Case

An AWS suddenly reports a temperature of 55°C with extremely high humidity and abnormal pressure variation while neighboring stations show normal conditions. The AI system should analyze temporal and spatial consistency, identify the reading as a probable sensor anomaly, generate an alert, and suggest corrective action.

Grand Challenge

Can AI build a self-aware and self-healing weather observation network capable of delivering trustworthy atmospheric data under all environmental conditions?

Output

Fully executable code with example usage and a document explaining various use cases

Downscaling of weather forecast from Block level to Panchayat level: Inferring high-resolution plots/ data/ information from low-resolution plot /data /information /variables for agro-meteorological advisory services.

Problem Statement

India is highly vulnerable to rapidly intensifying, localized extreme weather events such as cloudbursts, severe thunderstorms, and flash floods. Traditional physics-based Numerical Weather Prediction (NWP) models often suffer from computational latency and struggle to capture the rapid, small-scale atmospheric changes that preceded these events. There is a critical need for a real-time, hyper-local early warning system capable of 'nowcasting' severe weather 2 to 6 hours before impact, providing actionable lead time for disaster management.

Proposed Solution

We propose an advanced AI predictive engine designed for high-precision severe-weather nowcasting. Specifically, the system simultaneously predicts the onset of highly localized, rapidly intensifying events, namely severe thunderstorms, cloudbursts, and the subsequent flash floods,with an actionable lead time of 2 to 6 hours. Instead of relying on computationally intensive thermodynamic simulations, the system utilizes a spatiotemporal deep learning architecture to recognize the complex, multivariate atmospheric signatures that precede these extreme events. A critical component of this methodology is storm nowcasting using variations in integrated water vapor (IWV). By tracking rapid spatial and temporal accumulations of IWV, the model accurately identifies the concentrated moisture pools required for heavy precipitation. To predict multiple extreme events simultaneously, the engine employs a multi-task learning approach. A shared neural network backbone extracts foundational atmospheric features (moisture, instability, and lift) from the input grids. The network then branches into distinct output layers, allowing a single unified model to generate hyper-local probability risk maps for thunderstorms, cloudbursts, and flash floods simultaneously, entirely bypassing the computational latency typical of traditional numerical weather prediction (NWP) models.

Predictive Matrix

Key Atmospheric Variables

Severe convective storms require three primary ingredients: moisture, instability, and lift. Our AI model tracks the critical precursors across all three categories to ensure high accuracy and low false-alarm rates:

Observational Signatures

Rapid cooling of cloud tops, measured as the Cloud Top Temperature(CTT) Drop Rate, provides real-time validation of explosive vertical updrafts within the system.

Topographic Dynamics (The Flood Catalyst):

To accurately predict flash floods, the AI overlays the atmospheric probability maps onto a high-resolution Digital Elevation Model (DEM). This allows the system to calculate how terrain slope, elevation, and natural drainage basins will channel the extreme precipitation generated by a predicted cloudburst.To capture these predictors with hyper-local accuracy, the model fuses multi-modal, high resolution datasets: IMDAA Reanalysis Data (Historical Baseline & Thermodynamics): Multi-level air temperature, specific humidity profiles (for calculating CAPE/CIN), geopotential height, and U/V wind components (for calculating shear and convergence). Satellite Observations (INSAT-3D/3DR via MOSDAC): Water Vapor (WV) Channels. This is essential for deriving real-time Integrated Water Vapor (IWV) fluctuations necessary for our storm nowcasting. Thermal Infrared (TIR) Channels: Utilized to calculate the rapid Cloud Top Temperature (CTT) drop rate. Quantitative Precipitation Estimation (QPE): Satellite-derived precipitation estimates are used to monitor real-time rainfall intensity, serving as a reliable, openly accessible alternative to ground-based radar. Digital Elevation Model (DEM): High-resolution topographical data (such as ISRO's CartoDEM or SRTM) provides a static baseline of elevation, slope, and surface drainage networks, enabling translation of atmospheric cloudburst predictions into actionable flash flood warnings on the ground.

Technical Methodology

Expected Solution

The final deliverable for the Smart India Hackathon will be a fully functional, real-time prototypeof the AI-Driven Hyper-Local Early Warning System. At its core is a deployed multi-task inference engine that continuously ingests live INSAT satellite data and IMDAA thermodynamic baselines to simultaneously generate predictive risk maps for severe thunderstorms, cloudbursts, and flash floods within a 2 to 6-hour predictive window. This backend integrates with an interactive, webbased spatial dashboard designed for disaster management authorities, featuring dynamic risk maps overlaid on a Digital Elevation Model (DEM) and an Explainable AI (XAI) module that transparently displays meteorological triggers. Finally, an automated API will translate these predictive insights into immediate, categorized alerts sent directly to first responders and vulnerable communities the moment critical thresholds are breached.

Problem Statement

Identifying and tracking the exact geographic footprints of extreme weather anomalies (such as severe cyclones, heat domes, or cold waves) within massive global Numerical Weather Prediction (NWP) outputs is computationally intensive and heavily reliant on manual interpretation. In medium-range forecasting (3 to 10 days), atmospheric chaos renders traditional deterministic models highly uncertain. Furthermore, standard deep learning models (like standard CNNs or U-Nets) suffer from spectral smoothing—they tend to 'average out' spatial data, which destroys the extreme amplitudes (the high-intensity peaks of rainfall or wind speed) that forecasters actually need to track. There is a critical gap between broad, coarse 12 km global ensemble datasets and localized, high-fidelity threat tracking.

Proposed Solution

We propose an automated, state-of-the-art AI tracking and downscaling pipeline that shifts the paradigm from manual weather data sorting to automated, physics-informed anomaly tracking.Instead of relying on a single deterministic forecast run, our system directly processes multivariable, 4D Ensemble Prediction Systems (EPS) data.The system uses a two-stage hybrid AI architecture to solve the spectral smoothing problem:First, it utilizes a graph neural network (GNN) to map atmospheric variables onto a spherical mesh,instantly isolating moving anomalies and calculating their trajectory over a 3- to 10-day forecast window. Second, it pipes this isolated region into a generative diffusion model to perform statistical downscaling. This physics-constrained generative model mathematically derives a hyper-local 5km subgrid impact zone without flattening or blurring the severe amplitudes of the extreme weather event.

Technical Methodology

& Architecture Spherical Anomaly Tracking (Stage 1 GNN): To eliminate the geographic distortions caused by processing the spherical Earth on flat 2D pixel grids, the system maps the 12 km NCMRWF Global Ensemble (NEPS-G) grids directly onto an icosahedral mesh. The message-passing GNN calculates the Extreme Forecast Index (EFI) against a 30-year historical ERA5 baseline distribution to isolate standard deviations and draw a macro-scale temporal bounding box around the anomaly's trajectory. Amplitude-Preserving Downscaling (Stage 2 Diffusion): The system passes the cropped, macroscale bounding box into a conditional denoising diffusion probabilistic model. Rather than optimizing for mean errors (which blurs peaks), the diffusion model learns the physical relationships between synoptic-scale

features

Expected Outcome

& Key Deliverables The Tracking Core: A production-ready Spatio-Temporal GNN module that continuously processes global NWP streams to

output dynamic, automated 4D bounding boxes around evolving weather threats. The Downscaling Core A generative diffusion module capable of ingesting a 12 km resolution anomaly slice and outputting a probabilistically sound, 5 km resolution sub-grid array that retains extreme value amplitudes. The Visualization & Alert Dashboard: An automated system that translates the mathematical 5 km centroid arrays into clean, geographic visual layers. The Alerting API A lightweight, production-ready REST API that programmatically drops a pinpoint coordinate at the core of the severe anomaly and triggers categorized spatial alerts (low, moderate, and severe) across a precise 5 km geographical impact radius. Use Cases & Societal Impact Eliminating Alert Fatigue for the NDRF: Current weather alerts are often too broad, covering entire states or districts, which leads to public complacency. This solution allows meteorologists to issue hyper-localized, highly targeted warnings. It changes a generic'heavy rain in the district' alert into a precise 'high risk of flash flooding within your specific 5 km radius in the next 12 hours' alert, empowering first responders to deploy assets perfectly. Protecting Rural Economies Grants farming communities a highly accurate, 3- to 10- day lead time regarding localized catastrophic anomalies like sudden frost, hail, or heat domes. This structural foresight lets farmers alter harvesting schedules or apply cropprotection covers, shielding rural livelihoods from sudden climate shocks. Democratizing Supercomputing Power Once this hybrid AI pipeline is trained, it processes live inference data on a standard cloud GPU node in seconds, making high-fidelity climate forecasting highly affordable and easily accessible.

Problem Statement

Medium-range weather forecasts sometimes show large errors during rapidly evolving systems such as monsoon depressions, heavy rainfall events, western disturbances, cyclones, heat waves and break/active monsoon phases. Such forecast failures, or 'forecast busts', can affect operational decision-making.

Challenge

The challenge is to develop an AI/ML-based system that can identify regions and lead times where the forecast is likely to have high uncertainty or large error. The system should compare current NWP forecast patterns with historical forecast error behaviour and provide a forecast confidence indicator.

Expected Outcome

Problem Statement

Rainfall forecast errors over India vary with weather regimes such as active monsoon, break monsoon, monsoon lows/depressions, orographic rainfall, coastal rainfall and western disturbances. A single bias-correction method may not work equally well in all situations.The challenge is to build an AI/ML-based rainfall post-processing system that first identifies the prevailing weather regime and then applies suitable correction to the raw NWP rainfall forecast.The aim is to improve district/grid-level rainfall forecasts, especially for heavy and very heavy rainfall events.

Expected Outcome

Traditional Air Quality Index (AQI) forecasting models typically treat meteorology and pollution dispersion as separate entities. However, in highly polluted urban landscapes like Delhi NCR, there is a critical, dynamic feedback loop between the weather and pollutants. During peak pollution seasons (such as the winter stubble-burning period), atmospheric inversion layers trap particulate matter close to the ground. Conversely, dense concentrations of aerosols (PM2.5) block sunlight,altering local temperatures, wind patterns, and planetary boundary layer (PBL) heights. Ignoring these coupled meteorological-chemical feedback loops leads to significant inaccuracies in standard AQI predictions. To achieve high-accuracy, actionable insights, there is an urgent need for an integrated system that simulates real-time interactions between atmospheric physics and chemical transport. The challenge is to build a high-resolution, coupled forecasting system specifically tailored for Delhi NCR that predicts AQI for the next 72 hours. The solution must leverage advanced weather-chemistry models (such as WRF-Chem or similar open-source coupled frameworks) to dynamically interlink meteorology with pollution dispersion(specifically PM2.5 and Ground-level Ozone). A core focus should be accurately modeling the impact of atmospheric inversion on external pollution spikes, such as regional stubble burning,and how those trapped pollutants subsequently alter local weather conditions.Implement a workflow that handles two-way feedback between meteorology (temperature, wind,PBL height) and chemistry (PM2.5, PM10, O3,NOx). A user-friendly, real-time dashboard displaying high-resolution AQI forecasts for Delhi NCR with a 72-hour outlook.

Features that explicitly track atmospheric inversion strength and predict how stubble-burning plumes will disperse under prevailing weather conditions.

In recent years, the frequency, duration, and intensity of heatwaves across India have escalated sharply due to climate change. However, traditional meteorological warnings rely almost exclusively on ambient dry-bulb temperature thresholds. This creates a critical vulnerability:standard forecasts ignore the deadly compounding effects of relative humidity, wind speed, and solar radiation on the human body. A temperature of 40°C at 20% humidity feels vastly different from 40°C at 70% humidity—the latter can be fatal. Current public health infrastructure lacks localized, impact-based forecasting that translates raw weather data into actual physiological risk, human thermal stress levels, and projected mortality rates. The challenge is to design an intelligent, localized early warning system that shifts heatwave forecasting from 'what the weather will be' to 'what the weather will do' to human health. Participants need to build a predictive platform that computes a comprehensive Human Thermal Stress Index (integrating temperature, humidity, wind, and radiation) and links it directly to an automated Mortality Risk Index. The system should offer high-resolution forecasts to help municipal corporations, healthcare systems, and disaster management authorities deploy targeted,preemptive interventions. Develop algorithms to calculate advanced heat stress metrics such as the Wet-Bulb Globe Temperature (WBGT), Universal Thermal Climate Index (UTCI), or Heat Index (HI) rather than relying on temperature alone. Integrate historical public health, demographic (e.g., elderly or outdoor worker density), and localized weather data to predict heat-induced mortality and hospitalization spikes 3 to 5 days in advance. A dynamic GIS-mapped dashboard providing colorcoded, hyper-local alerts (Zone/Ward level) paired with actionable, automated public health advisories. An API capable of pushing automated SMS/WhatsApp regional alerts or localized triggers for city administration to initiate heat action plans (e.g., opening cooling centers, adjusting power grids, shifting outdoor work hours).

Convective storms, such as severe thunderstorms, hail, downburst winds, and cloudbursts are among India’s deadliest natural hazards, especially during the pre-monsoon and monsoon seasons.Despite advancements in Numerical Weather Prediction (NWP) models, traditional systems often fail to accurately capture these mesoscale extreme weather events.The primary limitation stems from spatial and temporal constraints: these violent storms develop rapidly within a window of minutes and occur at localized scales that slip through coarse grid resolutions. Current early warning infrastructures struggle to provide high-resolution, short-term forecasts (0–6 hours), leaving local administrations, aviation sectors, and rural farming communities vulnerable to sudden, devastating impacts. The challenge is to build a real-time, convective-scale Nowcasting System (0–6 hour lead time) operating at a hyper-local 1–3 km spatial resolution. Because traditional physics-based models are too computationally slow to simulate these rapid developments in real-time, participants must design a system rooted in Multi-Source Data Fusion architectures. The core

objective is to ingest high-frequency, heterogeneous meteorological streams, automatically detect early convective initiation, and dynamically forecast severe storm parameters (including lightning strike density, hail probability, downburst velocity, and cloudburst thresholds).Design a robust, real-time ingestion engine that fuses data streams from multiple sources: Doppler Weather Radars (DWR reflectivity and velocity fields), geostationary satellite imagery (INSAT-3D/3DR thermal/infrared bands), and ground-based lightning detection networks. A real-time,interactive GIS-mapped dashboard showcasing high-resolution (1–3 km) hazard zones with live countdown clocks for storm arrivals.

Urban flooding in major Indian metros like Mumbai, Delhi, and Chennai has become an annual crisis. Traditional Numerical Weather Prediction (NWP) models fall short because knowing how much rain will fall does not automatically translate into knowing where the streets will flood.Urban flooding is a hyper-local phenomenon dictated by micro-topography, concrete imperviousness, and heavily strained, invisible drainage networks. Currently, municipal bodies lack real-time, street-level predictive systems. Consequently, cities are caught off guard by rapid water accumulation, leading to severe traffic gridlocks, economic disruption, and loss of life. The challenge is to design a high-resolution, real-time Urban Flood Nowcasting System (0–3 hour lead time) capable of predicting street-level inundation before it happens. Participants must move away from isolated weather models and instead build a coupled framework. This system must fuse real-time rainfall nowcasts with high-resolution Digital Elevation Models (DEM) and a graph-based mathematical model of the city’s underground drainage network. By mapping how water flows, accumulates, and surcharges across concrete surfaces and drainage nodes, the solution should pinpoint exactly which streets or intersections will flood.Develop a pipeline that takes high-resolution rainfall nowcasts (from Doppler Weather Radars) and instantly routes that volume across a 2D surface terrain model. Represent the city's stormwater drain network as a directed graph (nodes as manholes/inlets, edges as pipes/canals). The model must calculate hydraulic capacity and predict where blockages or overcapacity will cause backflow onto the streets. A dynamic, web-based GIS dashboard showing real-time, street-by-street flooding projections (e.g., water depth estimations in centimeters) with a 0–3 hour forward-looking window.An API utility that can interface with navigation maps to suggest flood-safe alternative routes for emergency services, public transit, and commuters during heavy downpours.

Background

In India, due to natural disaster various natural dam / lake formations were observed which can be a major reason of flash flood in the lower catchment, for example, natural lake formed over the Rishi Ganga river of Uttarakhand in Feb 2021, Wapriyang river in Nov 2021, Phuktal river near Sumdo, J&K in Mar 15, Kosi river in 2008 etc. Devastating flood happened in the Kashmir valley, Assam in 2014 and many other places over a period of time. Therefore, simulation modelling for flash flood and scenario generation is important from Humanitarian Assistance and Disaster Relief (HADR) point of view. Another important aspect is water release issues from the Dam of major rivers. In the crisis situation, if the dam brakes, how much water will flow into the river and what are the area it will inundated / impacted need to be estimated. In order to carry out this work simulation modelling needs to be done for the same. In other way if any dam break situation happened then what will be the impacted area. Development of a modelling framework is required which will carry out simulation modelling for the same.

Description

The above problem statement envisages that a software tool need to be developed which should automatically carry out the simulation modelling for Dam break analysis and identify the inundated area due to flash flood in the lower catchment. The modelling framework should be developed using hydrological data, DEM and satellite imagery of any river. The software/ tools should be capable of carrying out the simulation modelling of water flow in case of dam break or water release through ‘Smooth Particle Hydrodynamics’ and ‘Delf3D’ model and compare the scenario.

Expected Solution/Deliverables

The proposed study aims to illustrate the current problems regarding framework generation of Humanitarian Assistance and Disaster Relief using simulation modelling related to flood management as follows:

Background

India's disaster-prone regions face recurring hazards such as landslides, floods, coastal erosion, and cloudbursts. Vulnerable habitations often remain in unsafe zones, leading to repeated loss of lives and property. Current relocation efforts are largely reactive, initiated after disasters strike, rather than proactively planned.

Description

The initiative seeks to develop an intelligent, GIS-enabled decision support platform. This platform will dynamically identify and update multi-hazard Red Zones (areas unsuitable for permanent habitation), assess the carrying capacity of safer alternative sites, and prioritize vulnerable habitations for relocation. The system will integrate hazard intensity, population vulnerability, and disaster history to guide evidence-based decisions.

Expected Solution

A robust, AI-driven GIS platform that Maps and updates hazard-based Red Zones in real time, assesses suitability and carrying capacity of safer relocation sites, prioritizes vulnerable habitations for immediate, short-term, and medium-term relocation and provides actionable insights to State Disaster Management Authorities for proactive planning.

Background

Hilly states in India are highly vulnerable to landslides and flash floods, which often occur with very short warning times. These sudden events result in significant loss of lives and property, and current early warning mechanisms are inadequate for hyper-local prediction and timely evacuation.

Description

The proposed initiative aims to develop a predictive system that integrates multiple data sources rainfall data, soil moisture sensors, slope stability models, historical landslide inventories, and real-time IoT inputs. By combining these datasets, the system will generate hyper-local forecasts at the village or ward level, providing sufficient lead time for evacuation and risk mitigation.

Expected Solution

A comprehensive flash flood prediction system that Integrates rainfall, soil moisture, slope stability, and historical disaster data, utilizes IoT sensors for real-time monitoring, issues hyper-local early warnings at village/ward level, and provides actionable lead time for evacuation and disaster preparedness.

Disaster management includes ideas related to risk mitigation, Planning and management before, after or during a disaster.