Operation Dronagiri: predicting irrigation needs from SAR soil moisture
A grant-funded project with IIT Tirupati, delivering SAR-based soil moisture estimation and an irrigation decision support tool for Vizianagaram district, Andhra Pradesh.
- Research
- Soil moisture
- SAR
This project is associated with the Technology Innovation Hub (TIH) on Positioning and Precision Technologies at IIT Tirupati Navavishkar I-Hub Foundation (IITTNiF) through Operation Dronagiri, a flagship initiative of the Department of Science & Technology (DST) Geospatial Data Promotion and Development Committee and the Geospatial Innovation Cell under the National Geospatial Policy.
Operation Dronagiri demonstrates the power of geospatial technology in transforming agriculture, transportation and infrastructure, and livelihood and skilling, aligning with the national vision of Viksit Bharat 2047.
The initiative is powered by the Geo-Intel Lab, the Geospatial Intelligence and Applications Laboratory of IITTNiF, which develops applications and platforms for public good and digital governance. The lab integrates GIS, remote sensing, GNSS and data analytics to deliver scalable decision-support systems, and hosts India’s first Federated Geospatial Data Interface (GDI) node.
All products and platforms developed under this collaboration are featured on the National Geo-Intel Portal.
Project at a glance
- Accelerator: Geospatial Innovation Accelerator, IIT Tirupati
- Pilot district: Vizianagaram, Andhra Pradesh
- Sector: Agriculture
- Status: Completed
The problem
Agricultural water management faces significant challenges from unpredictable weather patterns and inefficient irrigation practices. Farmers need accurate, timely data on soil moisture to optimise water usage and improve crop yields while conserving a scarce resource.
What we built
The solution combines SAR (Synthetic Aperture Radar) satellite imagery, machine learning and ground-truth validation to predict soil moisture. By integrating multi-temporal satellite data with field measurements, the models help farmers make informed irrigation decisions: soil conditions, weather patterns and crop requirements become actionable insight delivered through a decision support tool.
Geospatial datasets (GDI)
- SAR satellite imagery for soil moisture estimation
- Weather and climate data integration
- Multi-temporal analysis of agricultural fields
- Field segmentation and crop classification
Alignment with Operation Dronagiri
- Enhancing agricultural productivity through precision irrigation management
- Promoting sustainable water resource management and conservation
- Empowering farmers with data-driven decision-making tools
Milestones
-
Week 1–8
Data collection and processing
Field campaigns finished with comprehensive data collection. Completed
-
Week 9–12
Algorithm development
ML field segmentation model completed and tested. Completed
-
Week 13–18
Model validation
- SAR soil moisture model development completed
- Validation of SAR soil moisture against ground truth completed
- Decision support tool prototype completed
Completed
From the field
The decision support tool
Partnerships
- IIT Tirupati Navavishkar I-Hub Foundation (IITTNiF)
- Department of Science & Technology (DST)
- Geo-Intel Lab
- Local agricultural departments and farmer cooperatives
Impact and outcomes
- Water conservation Expected reduction in water usage through optimised irrigation scheduling, conserving water while maintaining crop health.
- Increased productivity Improved yields through data-driven irrigation decisions at critical growth stages.
- Farmer empowerment Accessible, actionable insight on irrigation timing and water allocation.
- Sustainable agriculture Efficient resource management with lower environmental impact over the long term.
Metrics being tracked
- Accuracy of soil moisture predictions
- Water usage efficiency improvements
- Crop yield variations across irrigation strategies
- Farmer adoption rates and satisfaction scores
- Cost savings through optimised water usage
What comes next
- Model validation Complete validation with new data from diverse agricultural regions and crop types.
- Decision support system Finalise the mobile and web interfaces for farmers.
- Final report and scaling Deliver the final report and prepare to scale to further districts and states.
- Integration with GDI Full integration with the National Geospatial Data Interface for broader access.
Sustainable Development Goals
This project contributes to the following UN Sustainable Development Goals:
- SDG 2 — Zero Hunger
- SDG 6 — Clean Water & Sanitation
- SDG 12 — Responsible Consumption
- SDG 13 — Climate Action
The team
Dr. Mohamed Musthafa
Specialises in machine learning applications for remote sensing and geospatial data analysis. Led the development of the soil moisture prediction models and decision support system.
Prem Mugundhan
Expert in forest and agriculture field work, data collection, data collation and logistics planning. Supported ground truth collection and data labelling.