This study is testing whether satellite technology can be used to support a new system that pays farmers for adopting sustainable rice-growing methods. We are working with 800+ farmers in nine villages in Telangana and Karnataka. Farmers are randomly divided into two groups. One group receives financial payments (about ₹4,500 per acre) for using water-saving methods called Alternate Wetting and Drying (AWD) and Direct Seeded Rice (DSR). The other group does not receive payments. All farmers are offered training in AWD. Satellites are used to check whether farmers are using the practices, and this is confirmed with farmer surveys and in-person visits. We will compare results on water use, crop yields, income, and farmer satisfaction, as well as environmental impacts like methane emissions.
Intervention
This study provides training on sustainable rice-growing practices, specifically Alternate Wetting and Drying (AWD) and Direct Seeded Rice (DSR), to all participating farmers. Farmers in the treatment group also receive financial incentives of about ₹4,500 per acre (up to 2 acres per farmer) for successfully implementing AWD. Adoption is verified primarily through satellite imagery, supported by farmer reports and occasional field checks. Farmers in the control group receive training but no payments.
Experimental design
We are working with 800+ smallholder rice farmers across nine villages in Telangana and Karnataka. Farmers are randomly assigned to two groups: a treatment group that receives financial incentives (about ₹4,500 per acre, up to 2 acres per farmer) for adopting water-saving methods, and a control group that does not. Both groups receive training on Alternate Wetting and Drying (AWD) and Direct Seeded Rice (DSR).
Satellite imagery is the primary way to verify whether farmers are implementing AWD practices, with additional confirmation from farmer reports and occasional field checks. To strengthen results, we will also use a comparison group of non-enrolled plots, monitored by satellite, to help build a “dynamic baseline” that reflects changing weather and environmental conditions.
Design details
Randomization method
The design is a matched pair design within strata defined by village and land size rankings. Farmers were sorted by village, then by total land size, and assigned to pairs. Within each pair, one farmer was randomly assigned to treatment and the other to control. For strata with an odd number of farmers, simple randomization with 0.5 probability was used. All randomization was done by computer.
Randomization unit
The unit of randomization is the individual farmer. Randomization was stratified by village and landholding size, but assignment was made at the individual level. There is no higher-level (e.g., village-level) randomization.
Clusters
855 individual farmers.
Observations
855 farmers (same as clusters, since randomization is at the individual level).
Environmental:
- Number of AWD drying events (satellite-derived)
- Duration of AWD drying events (satellite-derived)
- Estimated methane emissions flux based on drying event variables using v2.1 of the Source-selective and Emission-adjusted GHG Calculator (SECTOR) tool from the International Rice Research Institute (IRRI).
Agricultural and economic:
- Paddy yields (quintals/acre) - from endline survey
- Net profits (revenue minus input and labor costs) - calculated from output revenue data and cost modules during high-frequency visits
Secondary outcomes
Adoption and Practice:
- Farmer-reported continuation intention - will continue AWD/DSR next year (endline)
- Method consistency - final method used matches intended method (endline)
- Training attendance rate - from training session records
Environmental:
- Stubble burning incidence - residue burning done (yes/no) from endline + satellite detection
- Soil moisture indicators - from high-frequency monitoring (dry/moderate/wet classifications) [for a subsample of farmers with moisture measurements]
- Estimated water use - We will attempt to extrapolate the relationship between drying events and water use based on pumping hour metres on a sub-sample of plots.
Economic:
- Input cost variations - collected through periodic cost modules during high-frequency visits
- Labor allocation changes - from cost and labor tracking modules
Farmer Experience:
Perceived crop health - visual assessment (good/moderate/poor) from endline
Cited by
No citation data available for this registration yet.