You want to verify before you trust. Good — so do we.
Transparency is built into the site’s operating model. An Ecowitt station in the valley. Readings every five minutes. A forecast model with a public accuracy scorecard, regenerated every five minutes from recorded observations against WXSIM’s prediction. No hand-waving.
What actually helps you
These are the parts of the site built specifically for scepticism — methodology, raw data access, and a public track record.
1. Forecast Accuracy Page
Live comparison — Dayboro Model vs BoM vs Inigo Jones, last 30 days. MAE, RMSE, bias, precipitation categorical accuracy. Regenerated every 5 minutes by cron.
Open the accuracy dashboard2. 2026 Forecast Methodology (paid)
Station specs, calibration, data pipeline, model blend logic, known biases. Full write-up, regularly updated. This page is behind the $3.95 subscription — it’s the one artefact sized specifically for the scepticism-first reader.
Methodology page3. Monthly scorecard blog posts
Every month, a plain-prose post reviewing last month’s forecast accuracy — including the months where we missed. February 2026 Scorecard, March 2026 Review — real records, not cherry-picked.
Blog index4. REST API (free, rate-limited)
Endpoint base: /wp-json/dayboro-weather/v1/ — current conditions, forecast, accuracy, historical aggregates. Used internally by Garden Buddy and the mobile app.
5. Climate data atlas
Historical Dayboro climate aggregates, year by year. 2019 and 2023 pages currently paid-only (subscriber reward). Newer years move to free after their initial window.
Climate Data 2019 (paid)6. The Inigo Jones question
We include the Inigo Jones long-range method alongside standard NWP models — but we mark it distinctly. If you don’t buy astronomical forecasting, ignore those columns. The scorecard tracks each method separately.
Inigo Jones explainerThe numbers that matter
Bias: −0.56°C (model runs cool). Wind MAE: 1.9 kph (RMSE 2.65). Precipitation categorical: 64.5%. All live, all JSON-backed, visible at the accuracy page.
Where trust gets earned
If the scorecard page convinces you the methodology is worth reading, the paid tier is where the full methodology lives. $3.95/month. If the scorecard tells you we’re wrong too often, don’t subscribe — that’s the whole point of publishing it.
Read the methodology (paid) Just the scorecardHonest note — the caveats
Three things to flag up front, because you’ll find them anyway:
- One station. Dayboro township has a single Ecowitt station. Calibrated, but not a network. A farm 6km up the range will have different conditions than what we record — we don’t correct for that yet.
- Inigo Jones is a separate column. It’s an astronomical method, not NWP. We track its accuracy separately so you can judge on the data, not the label.
- No CSV export of historical data yet. On the backlog, not done. If you want something specific for analysis, email webmaster@dayboro.au — we can run a one-off.