LuaZero

Food Systems & Sustainable Agriculture · Net Zero Challenge 2026

Field evidence for low-emission rice.

We are building an evidence layer for rice cooperatives: farmer water logs, satellite cross-checks and traceable emissions calculations in one review workflow. Explore the working calculation and synthetic evidence prototype below.

Season calculator

Illustrative estimate · not verified

Explore how water management and accounting assumptions affect a rice season. This prototype estimates a scenario; it does not issue carbon credits.

Recent = incorporated shortly before planting; earlier = well before planting. The 5 t/ha straw assumption materially increases the estimate. Select what the field records support.
Enter assessed land-preparation or other relevant emissions. Zero is an unverified demo assumption.

Estimated reduction per hectare, per season

tCO₂e · after scenario deductions

Continuous flooding → selected project regime

Gross methane reduction
N₂O deducted in full
Other project emissions
IPCC-default uncertainty · 15%
SOC risk adjustment
Area total · one season
Annual scenario total
Gross value illustration

Gross value excludes verification, registry, operating and transaction costs, and any benefit-sharing arrangements. It is neither a buyer offer nor a farmer payout.

Where the gross reduction goesEvery bar below is a deduction we apply before anything is called a result. Updates with the inputs above.

Illustrative scenario, not a verified monitoring result.

Assumptions and calculation boundaries

CH₄ reduction = EF × SF(pre-season) × SF(organic) × [1 − SF(project water)] × hectares × days × 28 / 1,000.

We subtract estimated N₂O and user-entered other emissions, then apply a 15% uncertainty deduction to a positive balance. The selected SOC discount applies afterward. N₂O uses 0.00314 kg N₂O/kg project N when N does not increase, or 0.00786 when it does; the prototype retains the entire estimate rather than automatically excluding small amounts. GWP values are AR5 scenario assumptions: CH₄ 28 and N₂O 265. Final project accounting must confirm the applicable vintage and rules.

Baseline and project share the same pre-season and organic-amendment assumptions. This calculator does not model changed residues, measure methane, establish additionality, validate a baseline or prove programme eligibility. Negative balances remain visible and produce zero gross value. Annual multiplication assumes identical seasons; a real report must calculate each season and stratum separately.

Reference: Gold Standard rice methodology and 2025 SOC rule update, relevant to monitoring periods starting in 2026. No endorsement or certification is claimed.

Evidence review

Synthetic demonstration data

Test the review logic on a 12-week sample. Missing observations, low confidence and conflicting logs stay visible. Editing a farmer log preserves the original synthetic observation and records the change in the downloadable report.

The season at a glanceBoth records across all twelve weeks. Anything that is not a plain agreement stays visible rather than being smoothed away. Edit a week in the panel below and this updates with it.
  • Water state
  • Flooded
  • Drained
  • No log / no pass
  • Review
  • Aligned
  • Low confidence
  • Conflict or missing

Synthetic sample. Agreement between two sample records is not proof of a drainage event.

Sample plot DEMO-001 · no real farmer or satellite records
WeekFarmer logObservationConfidenceReview
What a Sentinel-1 pass can and cannot catchA drainage event is only visible if a satellite passes while the field is dry. For a repeating orbit, an event of length D covers the fraction D ÷ repeat-interval of each cycle, so that fraction is the chance of catching it. Planned AWD dry-downs of 7–10 days are usually catchable. Short or partial drainage is not.
Acquisition strategyRepeatPer 95-day seasonCatch an 8–9 d dry-downCatch a 3 d drainage

What the satellite layer still needs to prove

No Sentinel-1 data is retrieved by this demo. A real pipeline must account for acquisition gaps, crop stage, canopy effects, mixed pixels and local calibration. Satellite observations can guide review; they cannot independently measure methane or confirm every short drainage event between passes. An unresolved conflict triggers investigation, not an automatic reduction in credit count.

Why Vietnam

1 million ha

Vietnam's programme creates a clear setting for testing low-emission rice evidence workflows. The World Bank reports a TCAF commitment of up to USD 40 million for results-based carbon purchases.

World Bank programme context. This is national programme finance, not funding awarded to LuaZero. Access depends on the programme's own eligibility and payment rules.

Who would pay

Our proposed buyers are cooperatives, exporters and carbon project developers who need to assemble field evidence. Farmers contribute data and should share in benefits; the intended model does not charge individual farmers.

We will test willingness to pay USD 3–6 per hectare per season, alongside the actual cost of field support. No paying customers, funding or signed pilot partners are claimed.

A pilot built to test the unknowns

Proposed scope: 200–300 unique hectares, two crop seasons, one or two Mekong Delta cooperatives. Nine months from partner and crop-calendar readiness. Proposed budget: USD 55,000.

Months 1–2 · Set up the evidence

Secure a cooperative and research collaborator, agree data consent and accounting approach, map plots and install 20 water-level loggers.

Months 3–5 · Learn in season one

Collect paired water records, calibrate satellite features and commission reference measurements under a research-approved sampling plan.

Months 6–8 · Test on held-out fields

Freeze the model and evaluate season-two observations independently. Track farmer effort, yield, missing data and field support cost.

Month 9 · Review and decide

Request a verifier pre-assessment, publish measured unit costs and decide whether the evidence supports expansion.

Proposed decision targets

  • At least 85% balanced accuracy against held-out logger observations; report coverage and confidence intervals separately.
  • At least 70% of participating plots log in 10 of 12 weeks, with assisted reporting available.
  • Assess emissions agreement against reference measurements; investigate deviations over 20%.
  • Monitor yield and safe water-management practices; agronomists define stop rules.
  • Test a path to less than USD 5/ha/season recurring delivery cost at scale. The research pilot itself costs much more.

These are proposed targets, not achieved results. Partners, sampling design, verifier availability and quotations must be secured.

Supporting notes

Review the assumptions, calculation walkthrough and proposed field validation protocol.

Who is building LuaZero

Two Information Systems and Technology undergraduates at Institut Teknologi Bandung, Indonesia. We are developing the software prototype and seeking Vietnamese agronomy, remote-sensing and field-operation collaborators.

Muhammad Ilham Saripul MilahProduct and partnerships
Dzaky Zahy RabbaniEngineering and carbon accounting

Working today: scenario calculator, synthetic evidence review and report export.
Planned: Zalo integration, real Sentinel-1 processing and Vietnamese field validation.