Background¶

STATEMENT

The ability to anticipate problems is critical to the effectiveness of weather resilience teams. In the case of Scotland's resilience team, an item of interest is river levels because unusual levels, or unusual river-level-rates-of-change, hint at possible or impending bank breaches. At present

• During storms the team neither has access to real-time forecasts nor supplementary measures/metrics.
• Outwith storms the team does not have access to river levels intelligence, such intelligence enables hydrologists to identify problems, study patterns, etc., in-time.

Hence, [in search of] an illustrative development of a plausible river levels intelligence hub, that hosts measures, metrics, actionable insights, predictions, etc., and exists within a continuous integration & delivery setting.

The outcome expectations during a weather warning period are frequent

1. forecasting of river levels, that fall within a warning area.
2. measures & metrics updates; every three, or fewer, hours.

The outcome expectations outwith a weather warning period are

1. daily, within schedule, updates of measures & metrics updates.

Hence, the underlying aims are

1. hourly-point forecasts/predictions of river levels, at least thirteen hours ahead; forecasting/predicting ahead every eleven or fewer hours.
2. hourly-point forecasts/predictions within ± 0.025% error.
3. the wherewithal to update measures & metrics, and actions thereof, every three, or fewer, hours

A continuously & automatically updated online intelligence hub ...

\(\ldots\) that hosts river level intelligence metrics and forecasts. Outwith a warning period, the metrics shall be updated daily, and should include, at least,

1. Raw river levels per gauge, and annual river level pattern comparison per gauge
2. Daily extrema & medians
3. Drift
4. Weighted rates of change of river levels
5. A comparison of percentage river level changes w.r.t gauges within the same catchment.