Harvora™ calibration brief

Calibration and reliability at Harvora

Calibration asks whether a probability behaves like the frequency it represents; Harvora keeps sample size, truth source and proxy status beside every published evidence path.

Direct answer

What this supports

There is no single site-wide accuracy figure: frost, water balance, phenology, disease, pests and weather hazards have different targets and evidence tiers.

Intended for Growers, technical reviewers, partners and insurers evaluating prediction risk.

Evidence and method

How the signal is formed

Evaluation keeps metric, denominator, split, truth source and deployment status together; under-sampled or blocked outcomes remain explicitly limited. Qualifying field outcomes enter reviewed calibration and retraining review; submitting an outcome does not automatically update a serving model.

Evidence grade: ESTIMATE

Weather-benchmark measured today; every grower outcome logged in the product feeds the field benchmark it is measured against next.

Boundary

How to read this evidence

The public catalogue does not turn reanalysis, formula reproduction or historical climatology into independent orchard-outcome validation.

Sources and related evidence

Published 25 August 2026 · Last reviewed 25 August 2026 · Harvora technical review

Apply it to your operation

Work with us

Discuss a grower pilot