Coogee Beach: When water-quality data alone isn’t enough
Coogee Beach has received a Poor water-quality rating for the fifth consecutive year. It’s an example of why water quality data analysis needs to go beyond individual monitoring results to understand recurring issues.
One poor result might reflect conditions at a particular point in time. Five years of Poor ratings suggests there is something more persistent to understand.
And that raises a more useful question than simply what the latest rating was.
What can the monitoring record tell us about what is actually happening?
At Coogee, years of monitoring provide more than a history of individual results. They allow those results to be considered over time and alongside the conditions in which they occurred.
That is where a monitoring record becomes particularly useful.
The pattern behind Coogee’s Poor rating
Coogee’s five consecutive Poor ratings stand out, particularly when 97% of monitored ocean beaches across NSW were rated Good or Very Good in the latest State of the Beaches report.
But a Poor annual rating doesn’t mean water quality at Coogee is poor every day.
Conditions are generally suitable for swimming during dry weather. The problems tend to occur following rainfall, when stormwater can affect microbial water quality at the beach.
That distinction matters.
If we looked only at individual results, we would see water quality changing from one sampling event to another.
Looking across the monitoring history allows us to ask different questions.
Does the same thing happen each time it rains? How much rainfall tends to precede poorer results? How long does the impact last? Is the problem becoming more or less frequent? Are the same locations affected each time?
The monitoring history doesn’t necessarily answer every one of those questions on its own.
But it gives us somewhere much more useful to start.
Water quality data analysis: Recognising patterns and understanding causes
Seeing that poorer microbial results tend to occur following rainfall is important.
Understanding why requires context.
Depending on the water-quality issue being investigated, that might mean considering monitoring results alongside rainfall, flow, catchment conditions, treatment processes, operational changes, seasonality or results from other locations.
This is an important distinction in water-quality data analysis.
A recurring relationship in the data can help identify where further investigation is needed, but it doesn’t automatically establish cause.
The Australian Water Quality Guidelines make a similar distinction between identifying whether a trend exists and determining why it exists.
For water-quality teams, the practical challenge is often having the relevant information available when those questions need to be answered.
Historical monitoring data may exist across laboratory systems, spreadsheets, SCADA, reports and other sources. The information may already have been collected, but using it to understand what happened three months or three years ago isn’talways straightforward.
And that can limit how much value we get from the monitoring history we’ve built.
The real value comes from being able to turn that history into something useful.
When water-quality results can be considered alongside the other information that gives them context, the data starts to tell us more. It can help narrow an investigation, identify potential contributing factors and build a clearer understanding of what may be driving a recurring issue.
That understanding can then inform what happens next.
Where should further investigation be focused? What action might be needed? And, if an intervention is made, what should we continue monitoring to determine whether it worked?
In that sense, the monitoring record isn’t simply a record of what has happened. It becomes part of how we decide what to do next.
What happens after you act?
This is where Coogee becomes particularly interesting.
A $7 million project involving the NSW Government, Randwick City Council and Sydney Water is underway to address longstanding stormwater pollution affecting the beach.
That creates another role for the monitoring record.
Once the work is complete, how will we know whether it has worked?
The same historical data that helped establish the recurring problem provides a baseline against which future results can be considered.
Do microbial results following rainfall improve?
Do periods of poorer water quality occur less frequently?
Do they persist for less time?
Does the annual rating eventually change?
Monitoring then becomes part of a continuous process:
Monitor → Identify → Investigate → Act → Measure
The value isn’t only in identifying that something has gone wrong. It’s in using the information to understand the problem, inform what happens next and then determine whether the action taken has made a measurable difference.
What could this look like in Information Engine?
The following is an illustrative example only. D2K Information was not involved in the Coogee Beach monitoring program, and this does not represent an analysis of Coogee Beach data.
Consider a similar scenario within a water-quality organisation.
A result comes in that requires attention.
The first question is obvious: what happened?
But the investigation probably doesn’t stop there.
Has it happened before?
A team might want to look back through previous results to see whether similar events have occurred and how frequently.
Is there a pattern?
Looking across the monitoring history might show whether the issue is isolated, recurring or changing over time.
Is it happening elsewhere?
Comparing monitoring locations could help determine whether the issue is confined to one part of the system or occurring more broadly.
What else was happening?
This is where the water-quality result needs context.
Depending on the issue, that could mean looking at rainfall, wind, tides, flow, operational information or other environmental and process data alongside the water-quality results.
The purpose isn’t simply to see another trend on a dashboard. It’s to understand what the different pieces of information can tell us when they’re considered together.
This is the type of investigation Information Engine is designed to support.
Information Engine takes water-quality data and analyses it alongside other relevant data, such as rainfall, wind, tides, flow and operational or process information.
It helps teams investigate recurring issues, uncover relationships within the data, explore potential contributing factors and expose possible root causes.
The value isn’t simply in seeing the data together. It’s in extracting insight and meaning from it and converting that into actionable information.
That gives water-quality teams a stronger basis for deciding where to focus their attention, what needs further investigation and what action may be required.
It changes the conversation from simply:
What was the latest result?
to:
What is happening, what might be driving it, and what should we do next?
Making more use of the monitoring data we already have.
Water-quality organisations have spent years building monitoring records.
The question isn’t simply how much data has been collected. It’s how effectively that data can be used when it matters.
Coogee is a useful example because five consecutive Poor ratings make the recurring issue particularly visible. In other situations, the pattern may be much less obvious.
Looking beyond the latest result and considering monitoring data alongside the conditions, events and operational information surrounding it can help turn individual results into a much clearer understanding of what is happening.
That understanding can help focus investigation, inform action and, over time, show whether the action taken has made a measurable difference.
That’s ultimately where the value of monitoring data sits.
Not simply in collecting it, storing it or displaying it, but in extracting the insight and meaning needed to make better-informed decisions.
How much value are you getting from the water-quality data you already collect?
Information Engine is designed to help turn water-quality data and its surrounding context into actionable information.
