Rice carbon programmes rarely stall for want of a practice. They stall on choosing one that farmers will keep up and that the evidence can carry. DSR vs AWD is the first form of that choice most project developers meet, and the two practices answer it very differently.
Both go after the same cause. Conventional rice is transplanted into puddled fields and kept under standing water for weeks, and that prolonged flooding is what feeds the methane-producing bacteria in waterlogged soil. Direct seeded rice and alternate wetting and drying each shorten it. They act at different points in the season, ask different things of the farmer, and leave different evidence behind. Those differences should decide which practice a programme leads with, more than the headline percentages do.
DSR vs AWD: one cause, two points in the season
DSR acts at establishment. Seed is sown directly into the field instead of being raised in a nursery and transplanted into puddled, flooded ground, so the prolonged flooding of the transplanting phase never happens. The decision is taken once, at sowing, and it shapes the whole season.
AWD acts during the growing season. Rather than keeping the paddy continuously flooded, the field is allowed to dry down and re-flood on a cycle. Farmers typically judge the timing with a perforated pipe set into the soil, which shows the water table falling below the surface. The decision is not taken once. It is taken again at every dry-down, for as long as the crop is in the field.
DSR is one decision made at sowing. AWD is a decision made again every time the field is allowed to dry.
What each practice asks of the farmer
Adoption is the binding constraint on every agricultural carbon programme, so the farmer’s side comes first.
- DSR takes work away. No nursery, no transplanting, and much less of the labour that transplanting demands, which lowers the cost of growing the crop. That is often the reason a farmer adopts it at all. The carbon benefit follows the economic one rather than leading it.
- AWD changes irrigation timing. Crop, seed and livelihood stay the same. It reduces irrigation cost, which matters where water is paid for or rationed. It is a practice for irrigated lowland rice, and its mechanism assumes the field can be re-flooded when the cycle calls for it.
Neither asks the farmer to accept a smaller harvest. AWD maintains yields and DSR keeps them comparable. That matters more than any emissions figure: a practice that costs yield will not be sustained, however the credits are structured.
The numbers side by side, and why they are not like for like
| Direct seeded rice (DSR) | Alternate wetting and drying (AWD) | |
|---|---|---|
| Water use | 30–40% less | Up to 30% less |
| Emissions | ~30% less methane | Up to 50% less greenhouse gas |
| Labour and cost | Lower: no nursery or transplanting | Lower irrigation cost |
| Yield | Comparable | Maintained |
| Baseline | Puddled transplanting | Continuous flooding |
| When it acts | At sowing | Through the growing season |
Read quickly, the table says DSR saves more water and AWD cuts more emissions. Read carefully, it says less than that. The AWD figures are ceilings (“up to”), while the DSR methane figure is a typical value. The AWD emissions figure covers greenhouse gas as a whole, while the DSR figure covers methane alone. Each is also measured against its own baseline.
A verifier will not accept either headline as the number for your project, and a feasibility model should not either. The credited reduction is modelled per plot, from what was actually observed on that plot in that season. Use the ranges to decide which practice is worth piloting. Do not use them to forecast issuance.
What a verifier needs to see for each
This is where the two differ most, and it is the difference the agronomy hides.
AWD: many events, each one evidenced
An AWD claim is built from a sequence. Pipe installation is recorded with accuracy checks and a geo-located, time-stamped photograph. Every aeration event through the season is then photographed inside the plot’s geo-fence, with date, time and accuracy verification. Satellite mapping of flooding and drainage extends that field record across every plot. The strength is a dense trail at the level of individual events. The weakness is that each missed event leaves a gap, and gaps add up over a season.
DSR: one establishment event, observed from orbit
A DSR claim turns on how the crop was established. Sowing windows are identified from spectral signatures in time-series satellite imagery. Reduced flooded area is then measured against the puddled-transplant baseline, savings are modelled per plot, and the result is cross-checked against field enrolment data. The evidence comes early in the season and is mostly remote. Its weakness is the opposite of AWD’s. With few events to capture, each plot’s establishment has to be classified correctly, because everything after it depends on that call.
Neither is simply easier to evidence. They put the burden in different places. AWD relies on field capture that runs all season. DSR relies on the satellite record at sowing and the enrolment data it is checked against. Either way, the traps described in why agricultural carbon data fails audit apply: boundaries that do not match the ground, and records that cannot be traced back to one farmer and one plot.
Which methodology each is credited under
AWD has an established registry route. On the compliance side, most registered AWD projects are quantified under AMS-III.AU, the UNFCCC small-scale CDM methodology for methane emission reduction by adjusted water management practice in rice cultivation. On the voluntary market, Verra’s VM0051 covers improved water and crop management in rice. VM0042 can also apply, depending on project design. The Gold Standard DREAM methodology is the newer route. The AWD carbon credits article covers each one.
For DSR, the route we document is DREAM, where it is quantified digitally in a single project alongside AWD and residue management. If you plan to credit DSR on its own under another methodology, check that methodology’s applicability conditions before designing the programme around it. A practice is only worth what the methodology will count.
Why DSR vs AWD is often the wrong question
For many programmes the two are not alternatives, and a single project can run both. Historically the choice was forced by methodology rather than agronomy. Methodologies credited one practice at a time, and double counting was a genuine risk wherever practices overlapped.
Under the Gold Standard DREAM methodology, a single project can stack AWD, DSR and residue management, each quantified digitally with attribution kept separate. Up to 48% methane avoided per season is quantifiable under it. The stacking rules are set out in the Gold Standard DREAM methodology explained. The practical point is that leading with one practice does not mean giving up the other.
How to decide which to lead with
Four questions settle most programmes. All four are cheaper to answer before enrolment than after it:
Where is the labour pressure?
Where transplanting labour is scarce or costly, DSR can be sold to farmers on its own merits, before any carbon payment. That is the strongest adoption case either practice has.
How reliable is control over water?
AWD is built for irrigated lowland rice and depends on re-flooding when the cycle calls for it. Where that control is weak, the dry-down events the claim is built on will be irregular.
Where can the field team be, and when?
AWD needs geo-fenced event capture for the whole season. DSR puts the effort around sowing and enrolment. Match the practice to the field capacity you actually have.
Which methodology will you register under?
Settle it first, because it decides which practices count and what evidence is worth collecting.
Answer those, and DSR vs AWD becomes a question of sequencing rather than a bet on one practice. For the wider picture, see how rice carbon credits are measured and verified.
Related reading
The DSR dMRV page covers how direct seeding is mapped from satellite sowing signatures. The AWD dMRV platform covers the mobile app and dashboard behind aeration-event capture, and the DREAM platform shows how the two stack in one project.
- Rice carbon credits: how methane reduction is measured
- AWD carbon credits: how alternate wetting and drying generates them
- The Gold Standard DREAM methodology, and what it changes for digital rice
- Why agricultural carbon data fails audit
Talk to us about your programme
If you are weighing DSR, AWD or both for a rice programme and want to know what the data side would look like, we are happy to walk through it. We usually reply within one business day.
Book a demo →