Coverage metrics
Coverage measures how often predictable outages had a prior signal within the configured radius and lead window.
Forecast accountability
Outage prediction should measure itself.
Forecast accountability tracks whether predictions were timely, useful, and trustworthy enough for operational decisions.
prediction accuracy
outage forecast validation
model trust
Lead time
Lead time measures whether forecasts arrived early enough to support staging, briefing, or monitoring.
False positives and false negatives
False positives and misses help operators understand model behavior and where data quality or thresholds need refinement.
Model trust
GeoGridIQ tracks artifact availability, feature-contract compatibility, temporal validation, leakage controls, and separately governed fallback status. Fallback rows are not currently public-serving.
Frequently asked questions
Frequently asked questions for operators and planners.
Why is forecast accountability important?
It prevents prediction dashboards from becoming unmeasured claims and helps teams improve models over time.
Does a rules fallback mean failure?
No. It can be a separately governed internal scoring path when model trust or data quality is insufficient. It is not currently a public-serving track; public output remains unavailable unless a separate serving contract is implemented and verified.
Related GeoGridIQ resources
Continue exploring utility intelligence.
Documentation
Read GeoGridIQ documentation for platform scope, contract-bound data sources, prediction and GIS methods, limitations, and conditional crew-staging support.
Open Data Reports
Public utility intelligence reports covering Quebec outage risk, vegetation threats, storm impact, and critical infrastructure exposure.
Explainable Outage-Risk Intelligence
See how GeoGridIQ presents time-bound outage-risk outputs with source, model, validity, uncertainty, and explanation context for human review.
Vegetation Risk Analysis
Learn how permitted satellite vegetation indicators, weather, and generalized infrastructure context may support bounded vegetation-exposure screening for grid reliability.