Time for a closer look at nitrates?

Nitrate (oxidised nitrogen) is the most widespread contaminant present in New Zealand groundwater—a concern for human health, the environment, and a potential barrier for primary sector exports. We investigate whether we need to monitor nitrate concentration continuously and look at progress in practical nitrate monitoring.

Nitrate is highly soluble and poorly bound to soils. It is easily mobilised by rainfall and seeps into shallow groundwater, agricultural drains, streams, and eventually into rivers, lakes and estuaries. 

A recent review of water quality trends confirms nitrate concentrations are increasing nationally. The  Government’s new framework for managing freshwater requires Regional Councils—in consultation with water users—to specify water-use values and set appropriate nitrate limits.

NIWA’s National River Water Quality Network (NRWQN), established in 1989, is New Zealand’s most comprehensive freshwater quality monitoring network. A physical water sample is taken from each of these sites (located on 35 rivers distributed evenly throughout NZ) once a month and analysed at NIWA’s Hamilton laboratory.

But does monthly sampling deliver the nitrate data set now required by regional councils or is it time for more frequent sampling?

Time for a closer look?

Where nitrate concentration can change significantly between samples, the simple answer is yes. Insufficient sampling can lead to over or under-estimation of nitrate loads, inaccurate pollution assessments and reduce our ability to accurately detect cycles.

The accompanying graph shows what can happen when monthly nitrate sampling (orange trace) is set against monitoring data recorded every 15 minutes (blue trace):

A real data example showing the ‘information’ lost by infrequent sampling. The monthly value (orange trace) under-represents the peak by 5% and provides little indication of how nitrate concentration varies cyclically or with flow dynamics. [Graph: ESR]

Less can sometimes be more

In an ideal world we would collect continuous data but not all aquatic environments are subject to rapid changes in nitrate concentration. For example, nitrate concentration in groundwater usually changes slowly. As a consequence, some Regional Councils sample groundwater yearly.

Being able to detect a ‘transient spike’ in nitrate concentration is likely to be more important for assessing the effects of nitrate toxicity on fish and aquatic life rather than when considering its impact on plant growth, which is less immediate.

Deploying an in-situ continuous nitrate sensor, and frequent measurement of nitrate concentration over a short period of time, would provide information about nitrate variability over this period. This data could then be used to determine whether longer-term or permanent continuous monitoring is required.  This requirement could be determined by deploying sensors for fixed periods of time at a number of sites, as a rolling campaign. The high-frequency data would then indicate whether the transient or short-term fine structure in nitrate concentration needs to be characterised.

Can nitrates be monitored continuously?

Yes… of the five forms of nutrient that NIWA currently measures in the NRWQN, nitrate is the only one we can measure continuously. Currently the best sensors for continuous monitoring are based on UV spectrometry.

NIWA, Otago Regional Council (ORC) and the Institute of Environmental Science and Research (ESR) are currently trialling TriOS Opus multispectral nitrate sensors.

TriOS Opus multispectral nitrate sensors. [Photo: ESR]

These operate on the principle that nitrate absorbs radiation in the ultra-violet (UV) region of the electromagnetic spectrum at a characteristic wavelength of about 220 nm.

Pulses of UV light are transmitted through a sample path of water (typically 20 mm), received and spectrally analysed. The amount of UV absorbed by the water sample is proportional to the concentration of nitrate in the water. The processor compares the intensities of the two received beams, derives the nitrate concentration and sends this to a recorder such as a web-enabled datalogger.

Multispectral UV nitrate sensor diagram (generic) [Image: NIWA]

The sensor may be located in a stream, or alternatively, above water-level and have water pumped to it. The optical windows the UV passes through, are usually made from transparent quartz. This can resist abrasion by the hardest materials likely to be present in the water stream.

These windows can be automatically cleaned by compressed air, a built-in wiper or ultrasonic vibrator. Monthly site visits enable a physical verification sample and also provide an opportunity to inspect and clean the instrument manually.

Strengths and weaknesses of UV multispectral sensors



+ Precise, accurate and stable

+ Inexpensive to operate

+ Minimal maintenance

+ Measure over a wide range

+ No chemicals needed

+ Robust and durable

+ Fast response

+ Spectra may also detect e.g. turbidity

+ Easy to deploy… ‘plug and measure’

+ Factory pre-calibration (may need ‘local water’ calibration)

+ Independent of water temperature

X Up front cost

X Power use

X Other constituents (notably CDOM and turbidity) can interfere with optics

Otago Regional Council nitrate monitoring trials

ORC has been carrying out an extensive water quality programme in North Otago’s Kakanui-Kauru region since March 2014. The programme is providing new knowledge of groundwater-surface water interaction, focussing on nitrate mobility and its effects on the Kakanui River.

This programme includes monitoring 15 groundwater bores and 14 surface water quality and/or flow sites fortnightly or monthly. ORC has recently added new, continuous, in-situ nitrate sensing technology (TriOS Opus), at two important indicator sites (Gemmel’s Crossing Bridge and McCones), to validate and evaluate the sensors by comparing their data with that from the monthly ‘grab sample’ analyses.

Why is ORC monitoring these sites?

To avoid a build-up of algae, ORC set a ‘nitrogen cap’ of 0.075 mg/l on the Kakanui River and a nitrogen (in nutrients) leaching rate limit of 20 kgN/ha/yr over the adjacent aquifer. Results so far verify that nitrogen concentrations below 0.075 mg/l do limit algae growth in the river, as expected.  ORC will next carry out groundwater modelling to determine if the set nitrogen leaching rate limit will sustainably support the river’s nitrogen cap. 

Otago Regional Council staff install a TriOS nitrate sensor in the Kakanui River. The yellow circle on the inset at left shows where the nitrate concentration in the water is measured. [Photos: ORC]


How does ORC rate TriOSs performance so far?

Rachel Ozanne is leading the ORC’s work and says the council is very happy with the performance of the TriOS Opus sensors. 

“The installations went smoothly and they have required very little maintenance since they went in at the end of September 2016. We’re using Zebra-tech wipers on both probes (with a 5-minute sweep) which are doing a good job of keeping the lenses clean, according to the ABS360 signal we get.  We do notice that during a rain event and/or flush down the river, there is increased noise, but once the flows recede the noise subsides. We check the probes monthly and give the lenses a light cleaning, as recommended by the manufacturer.”

“We are taking regular grab samples to provide validation for the probes and will continue doing so through the year.”

“The probes are robust, reliable and appear to be giving accurate data. They require very little maintenance.  We haven’t noticed any serious limitations yet.” 

Future directions

From what they have observed so far, ORC anticipates that continuous monitoring is the way of the future. ORC has installed three TriOS NICO (nitrate-focussed) sensors in the Shag River (further south), as part of another surface water/groundwater project. 


Jeremy Bulleid – NIWA Instrument Systems 

Rachel Ozanne – Otago Regional Council

TriOS2 Opus multispectral nitrate sensor. [Photo NIWA]