Skip to dashboard
ClimateChain BD
IEEE ClimateChain Global HackathonClimate Data & Environmental Monitoring

Climate risk you can prove.

ClimateChain BD turns open ECMWF/ERA5 rainfall data for the Surma–Kushiyara basin into a transparent risk index — then commits each assessment to a tamper-evident registry, so a single altered byte becomes visible to anyone.

Explore basin riskRun the tamper test

04

Monitoring stations

Surma–Kushiyara basin

1d · 3d · 7d

Rainfall windows

Strictly past-only

05

Verification states

Including graceful RPC failure

SHA-256

Dataset integrity

Manifested & re-checked

Basin pulse

Surma–Kushiyara aggregate

Connecting to stations…

Unknown risk
—/100

mean station SRRI

no trend data

station spread

Severe stations

0 / 4

Registry

Checking…

Mode

Connecting

Built for relief audit, parametric insurance & civic planning

Research prototype · not an official warning system

Sylhet Division scope

SRRI is an empirical meteorological rainfall-risk indicator calibrated to BMD thresholds — not a validated hydrological flood probability. Machine-readable flood ground truth was unavailable, so no flood labels were fabricated.

Scenario
API connectedAwaiting data

Risk intelligence

Station risk assessment

A transparent 24h / 3-day / 7-day rainfall index with weighted contribution attribution.

Past-only features
Station
24.8949°N, 91.8687°E · Surma RiverSource: Open-Meteo / ECMWF ERA5 · SRRI v1

Station network

Four sentinels on the Surma–Kushiyara

Every card is computed from the same past-only feature pipeline and re-ranked on each scenario change.

Surma–Kushiyara basin

Geospatial station network

Real coordinates, river corridors, and live SRRI color coding. Click a marker to reassess a station.

4 nodes
SURMA RIVER →KUSHIYARA RIVER →MEGHALAYA PLATEAUHAOR WETLANDSINDIABANGLADESHN

SRRI legend

LowModerateHighSevere
Rainfall data is real ERA5/ECMWF via Open-Meteo. River gauge levels and flood-extent labels remain unmonitored/null where no verified open source exists — we do not fabricate hydrology.

Observations

Precipitation record · Sylhet Sadar

No observation records are available for this station yet.

No observation records available.

Try the live scenario or a historical monsoon date.

7-day outlook

Forecast rainfall · Sylhet Sadar

Numerical weather prediction is kept explicitly separate from observed risk and is never registered as historical provenance.

Open-Meteo NWP

Trust layer

Blockchain provenance & tamper evidence

Canonical JSON → domain-separated SHA-256 → EVM-ready commitment. Change one byte and the verdict flips.

Checking registry…

01Canonicalize

Sorted keys · normalized floats

02Hash

Domain-separated SHA-256

03Commit

bytes32 commitment

04Verify

Recompute & compare every byte

Report commitment

Deterministic identifiers for station · —

Commitment ID (bytes32)

Derived at registration

Report hash (committed)

Not registered in this session

Data batch hash

Computed at registration

Model version

—

Canonical schema

climatechain-provenance-v1

Registry

Local simulator

Contract

Simulated commitment

Integrity console

Register the commitment, then verify the exact payload.

API console

Registration target: labelled local simulator — no transaction or gas is claimed.

Run verification to compare the local canonical hash against the registry.

No verification in this session

Registry audit trail

Commitments registered through this API instance (local simulator or RPC).

loading…
CLIMATECHAIN_BD_PREDICTION_REPORT_V1: domain separationPENDING / UNAVAILABLE are first-class states — no false VERIFIED verdicts.

Why it matters

From open rainfall data to trusted climate action

The same verifiable primitive can support insurance, relief accountability, and long-horizon adaptation planning.

Payout integrity

Parametric climate insurance

A rainfall threshold can trigger a payout while the canonical event report remains independently auditable — reducing dispute friction after a disaster.

Audit trail

Relief & grant accountability

Donors and auditors can verify that the rainfall snapshot behind a needs assessment was not quietly rewritten before funds were released.

Open baseline

Civic planning & research

Planners get a transparent, reproducible rainfall-risk baseline for the haor belt, with limitations and data lineage exposed instead of hidden.

Built to be judged — and re-run

Climate relevance

Surma–Kushiyara flash-flood corridor

Station-level risk for Sylhet Sadar, Sunamganj, Kanaighat and Zakiganj.

Data credibility

Real ECMWF / ERA5 rainfall

Open-Meteo ingestion with committed SHA-256 manifests and integrity re-checks.

Technical novelty

Verifiable provenance layer

Canonical JSON + domain-separated hashing + EVM commitment + five-state verification.

Demo quality

One-click tamper simulation

Judges can see a report flip from VERIFIED to MISMATCH in seconds.

Designed for non-experts

Plain-language summaries, explicit limitations, and no black-box claims. A journalist, relief coordinator, or judge can follow the evidence chain without reading Solidity.

Scalable next steps

FFWC / BWDB gauge adapterQ1
IPFS-pinned report payloadsQ2
Parametric trigger contractR&D

Scientific methodology

Transparent by construction

The data lineage, formula, assumptions, and ethical boundaries behind every SRRI score.

Open, reproducible rainfall data

Daily precipitation is fetched from Open-Meteo's ECMWF/ERA5 reanalysis archive and high-resolution forecast API for four Sylhet stations. No API key, no scraping, no synthetic flood masks.
ERA5 reanalysisECMWF NWPCC BY 4.0Open-Meteo

Committed dataset manifests

0/0 integrity checks

Manifest metadata unavailable. The API will surface these when reachable.

Pipeline guarantees

  • Chronological sorting + duplicate-date removal
  • Range checks: negative and >500 mm flagged
  • Trailing 3d/7d windows reference past records only
  • Seed ingest writes a sample + SHA-256 manifest; API reads never rewrite it
  • Nullable river levels and flood labels — never fabricated

What this is not

No flood extent labels, no river gauge values, and no official warning status are invented. Where a verified open source is missing, the field stays null and the limitation is disclosed at the API and UI layers.
Open source · MIT licence · reproducible by designBMD-calibrated rainfall semantics · SRRI v1
Built for the IEEE ClimateChain Global Hackathon · Climate Data & Environmental MonitoringNext.js 14 · FastAPI · Solidity 0.8.24 · Open-Meteo / ECMWF