How eFlow Projector works

The computation runs on a site by site basis. Initially importing all flow data, environmental water data, prevailing climate conditions, flow rule data, then running through each rule defined for that site. ​ For each rule, an overall result for magnitude (all discharge rule types), duration (all discharge rule types), count (freshes) and independence (freshes) is calculated, potentially across multiple prevailing climate conditions with varying rule attributes. The product of these results is then calculated to give an overall score for that rule. The environmental water data is then used to calculate the contribution of delivered environmental water to the success of a rule, given that environmental water data is available for that site (this also takes into account lag). In the event that zero percent success is achieved for any relevant attribute within a rule (magnitude, duration, count, independence), the overall success will be 0, but the success of each attribute will still be reported, to show why a rule is failing.​

How low flow (and oversupply) flow rule assessment work​s

Steps for assessment

  1. Create a subset of the data for the given time period, split this into periods that relate to different prevailing climate conditions.​
  2. For each climate condition period, get the flow rule requirements for that climate condition, if there are no flow rule parameters for this prevailing climatic condition note this as a warning and skip the affected days.​
  3. For each day in this climate condition period, get a magnitude score for the day (partial score if applies).​
  4. Sort the magnitude scores.​
  5. Remove days with a score of 0.​
  6. If the list is greater then the desired rule duration (# days), take the top N (# days) scores, as defined by the duration of the rule.​
  7. If there is ewater data, get a daily % score of how much environmental water data contributed up to the rule magnitude (i.e. if 50ML/D was recorded and ewater provided 25ML/D that’s a 50% score, always prioritises environmental water over natural flow)​.
  8. Average the magnitude scores and get a duration score (partial score if applies), if the duration score is 0 then return a result with the magnitude score, duration score of 0 and overall score of 0​.
  9. Multiply the magnitude and duration score to get an overall score.

How freshes flow rule assessments work​

A key concept for freshes is the ‘event opportunity’. The event opportunity is a collection of continuous days that could contain an event. eflow Projector identifies event opportunities, then determines the collection of days within that opportunity that provides the best event magnitude score. This subset of days is the ‘event’ used for scoring.​

Steps for assessment

    1. Create a subset of the data for the given time period, split this into periods that relate to different prevailing climate conditions.​
    2. For each climate condition period, get the flow rule requirements for that climate condition, if there are no flow rule parameters for this prevailing climatic condition note this as a warning and skip the affected days.​
    3. Loop through each day in the dataset, check if the day is over the minimum magnitude needed to constitute a fresh event (determined by partial success).​
    4. If the minimum magnitude is met, create and store a fresh ‘event opportunity’ with the day in question and all subsequent days that meet the minimum magnitude criteria.​
    5. If the magnitude score drops below the minimum magnitude level, start counting the days it is below this level, once this count is equal to the minimum independence score (determined by partial success) stop the current fresh event opportunity.​
    6. Examine the fresh event opportunity and extract the ‘best’ event from the opportunity. The best event is the one where the average of magnitude score for each day in the event (up to the target event length) is maximised.​
    7. Return to step 3 from the day after the extracted event and continue searching for further event opportunities.​
    8. Once a list of acceptable fresh events has been achieved, sort through them to order them from best score to worst.​
    9. Take the top N events up to the count desired, if there aren’t enough events to reach the rule count attribute, take all of the events. ​
    10. Determine seasonal magnitude score. For each event, the average of the daily magnitude scores gives and event based magnitude score. The event based magnitude scores are averaged to give a season magnitude score.​
    11. Determine season duration score. Firstly determine the partial duration score for every event. Average the partial scores for all selected events to determine the season duration score.​
    12. Determine season count score. The season count score is simply based on the partial success table for the count of the number of events recorded for the season.​
    13. Determine season independence score. Calculate the partial success independence score based on the period between each event. The partial independence scores are averaged to give a seasonal independence score.​
    14. If there is ewater data, get a daily % score of how much ewater data contributed up to the rule magnitude (i.e. if 50ML/D was recorded and ewater provided 25ML/D that’s a 50% score, always prioritises ewater over natural flow) for the identified event days.​
    15. Multiply the magnitude, duration, count and independence scores to get an overall result for the season.​

How partial success works

Traditionally, flow rule success would be considered as either a pass or fail. A pass occurs when all target flow components are met (magnitude, duration, count, independence). If any component does not meet the target requirement then the flow rule fails. A variation on this ‘strict’ compliance is to maintain this strict pass/fail criteria for magnitude and duration for freshes, but the score is scaled by the proportion of the target count achieved. ​ The partial success approach of the eflow Projector is to allow the creation of a continuous function to define success for each flow component. The approach scales the success based on the proportion of the target that was achieved.