Tidal creeks – connections between freshwater and saltwater

An experiment in Henderson Creek, Auckland, has demonstrated how tidal creeks variously import, export and deposit sediment, depending on the wind and freshwater runoff, and modulated by the tide.


The issue

Tidal creeks connect freshwater stream networks to saltwater estuaries and beyond, but they are much more than a simple conduit. The mixing of freshwater and saltwater along the length of the tidal creek controls how the plume at the mouth of the creek is formed, which in turn controls what happens to freshwater runoff and its associated contaminant load (e.g., sediments and heavy metals) in the wider estuary receiving waters.

At the same time, contaminants themselves are transformed as they move through the tidal creek. For instance, sediment particulate matter flocculates (comes out of suspension to form larger clusters) in response to changing salinity and mixing. Additionally, heavy metals initially in solution at the head of the tidal creek begin to attach to suspended sediments as they flocculate and as the water chemistry (primarily salinity) changes. These three processes – freshwater/saltwater mixing, sediment flocculation and metal transformations – are intricately linked.

We conducted a field experiment in Henderson Creek, Auckland with two goals:

  • to extend our understanding of how tidal creeks function as the link between land and sea
  • to validate and improve how these processes are represented in numerical source-to-sink models that predict how catchment-sourced contaminant accumulates in estuaries.

The approach

The Henderson Creek field experiment was designed to measure inputs of sediments and metals at the head of the tidal creek, and outputs of the same at the base of the tidal creek where it discharges into the wider Central Waitemata Harbour.

The experiment ran for three months, from 8 April 2010 to 24 June 2010, and captured data under a wide range of tides and weather (wind and rain). Within that three-month period, three one-day experiments that involved a more intensive sampling effort were conducted. These were designed to capture more detail on metal and sediment processes, which would allow for a more robust interpretation of the less intensive, longer-term data.

The experimental work was intended to complement the Central Waitemata Harbour Contaminant Study, which used the USC-3 source-to-sink model to predict sedimentation, rates of heavy metal (zinc, copper) accumulation in harbour bed sediments, and the time it will take in the future to exceed sediment quality guideline thresholds under four catchment development scenarios

The result

The data demonstrate how sediments move through and deposit within the tidal creek. This includes sediments that derive from catchment erosion and that are transported to the headwaters of the tidal creek by freshwater runoff, and sediments that derive from erosion by waves of seabed sediments in the wider Central Waitemata Harbour and that are transported to the mouth of the tidal creek by tidal currents. Data from rainstorms, from periods when it was calm and dry, and from periods when it was windy and dry were analysed.

The data show that Henderson Creek variously imports, exports and deposits sediment, depending on the wind and freshwater runoff, and modulated by the tide.

During heavy rainfall, freshwater-borne contaminants are mostly flushed through the tidal creek into the wider Central Waitemata Harbour, beyond the mouth of the creek (see Figure) – this verifies previous model predictions. In contrast, between rainstorms and particularly when it is windy, contaminants in the harbour are resuspended from the seabed into the water column by small wind-waves, and then progressively transported by tidal currents back into the tidal creek, where they accumulate.

The long-term fate of the tidal creek and of sediment that is eroded from the land will depend on the balance that is established between the export of sediment during periods of high freshwater runoff and the return of sediment to the tidal creek by tidal currents between those periods. The return of sediment to the tidal creek is greatly enhanced by wind waves that resuspend sediments on the intertidal flats of the wider Central Waitemata Harbour. Waves can, in fact, be seen as cleansing the wider estuary of fine sediments, since they initiate the net transport of sediment off the intertidal flats and back towards the land.

The analysis of the Henderson Creek dataset has shown that key aspects of the modelling developed specifically for the Central Waitemata Harbour Contaminant Study are broadly correct. These are:

  1. the flushing of sediments (and attached metals) from tidal creeks and the dependence of that flushing on freshwater discharge, and
  2. the way in which waves cleanse the wider harbour of fine sediment.

The proportion of any given metal dissolved in the water or attached to particulate material (primarily sediments) suspended in the water column is an important control on how the metal is dispersed in the estuarine receiving environment. It also affects the metal toxicity. Further analysis of the dataset will allow us to quantify this, for future inclusion in predictive models.


Green, M.O. and Hancock, N.J., 2012. Sediment transport through a tidal creek. Estuarine, Coastal and Shelf Science, 109: 116–132. DOI: 10.1016/j.ecss.2012.05.030 

Towards the head of the tidal reaches of Henderson Creek, Auckland. Here, the tidal creek is confined between mudbanks that are topped by mangroves. The mudbanks are inundated for a short time at high tide. Credit: Malcolm Green
An extract of about 5 days of data from the Henderson Creek experiment, showing what happens in a rainstorm and in the aftermath of a rainstorm. Data from the two measurement stations are shown: station Confluence is at the head of Henderson Creek, and station Mouth is at the mouth of Henderson Creek. Panel (a) shows the tide at both stations. There is about 12 hours between successive high tides (and low tides). Panel (b) shows cumulative rainfall in the catchment. Notice that some 80 mm of rain fell in 12 hours on May 20. Panel (c) shows freshwater runoff to the head of Henderson Creek; the runoff increased during the rainstorm on May 20, and quickly subsided when the rain stopped. Panel (d) shows salinity. Notice that, at low tide when it was raining, salinity was zero at both stations, which indicates that Henderson Creek had essentially turned into a freshwater river. The longitudinal salinity gradient re-established soon after it stopped raining. Panel (g) shows sediment transport through each station, where positive denotes seaward transport and negative denotes shoreward transport. During the rainstorm, transport was strongly seaward at both stations, which indicates that sediment delivered to the head of the creek by freshwater runoff was being flushed through Henderson Creek and into the wider Central Waitemata Harbour. After the rainstorm, sediment transport reduced and turned around into the tidal creek, reflecting the return to, and sequestration in, the tidal creek of sediment that was flushed into the harbour during the preceding rainstorm. Panels (e), (f) and (g) show current speed, water flux and suspended-sediment concentration, respectively.