What do we measure?

Water quality variables measured

A range of physical and chemical variables are measured in the NRWQN, covering four main categories:

Physico - chemical variables: dissolved oxygen, temperature, pH, conductivity
Optical variables: Visual clarity, turbidity, coloured dissolved organic matter
Nutrients: Total and dissolved forms of nitrogen and phosphorus
Microbial Indicator: E. coli (since 2005).

In addition, we also measure several biological indicators of water quality, specifically macroinvertebrates and periphyton. Dissolved oxygen, temperature and visual clarity are collected directly while in the field. Water samples are obtained from each site and freighted by overnight courier to the NIWA laboratory in Hamilton, where the remaining water quality variables are measured.

Find out more about biological indicators

The following table explains the methods to capture parameters in more detail   -  as at  March 2014.

 

COLUMN ABBRV.  EXPANDED NOTATION  UNITS  METHOD DESCRIPTION

REFERENCE

Refer Ref.

Method Notes

Detection Limit

(indicative)

Precision

(estimate)

within  +/-

FILT

0.45u

MF

MISC.

COMMENTS

MISC.

COMMENTS 2

Time Time (NZST) hh:mm New Zealand Standard Time (NZST) 11




C Cloud cover x/8 1/8 to 8/8 observed cloud cover 11




Temp Temperature Deg C Glass Thermometer, or DO meter thermistor (both rated) 3, 11 0.1 0.1
majority readings via DO meter thermistor
DO% Dissolved Oxygen % saturation Meter (YSI model 57) to Dec 2008, then HACH HQ40d from Jan 2009  3, 11 0.5 1.0


DO mg/m3 Dissolved Oxygen mg/m3  O Calculated from % saturation, water temperature, and pressure at site (to Dec 2008).  Then as direct reading from replacement HACH meters   11 0.1 0.1


Flow Instantaneous Discharge m3/sec Gauging or rating 11 - n/a
Provisional data only, may be subject to later applied rating changes etc. Most accurate flow data may be obtained from the National Archived flow database
Clar Visual Water Clarity  m Horizontal Black Disc. 10 0.01 0.01 - 0.1


Turb Turbidity NTU Nephelometer (HACH 2100A) to March 2012, then switched to Hach model 2100AN 3 0.1 no info
Hach 2100A became faulty, and replaced with newer model 
pH Hydrogen ion concentration pH Meter (Radiometer, using Radiometer GK2401C electrode) 2 0.05 0.05


Cond Electrical conductivity uS/cm @ 25 oC Meter (Radiometer CDM83) 3 0.5 1, or 2%


BOD5 Biochemical Oxygen Demand - 5 day incubation.   ppm O 5 day oxygen depletion test @ 20 Deg C 2 0.2 0.2
Analyses ceased after JULY 2002 3  indicator sites continued though - AK2, RO2 and WA9
NH4 Ammonia   mg/m3 as N automated phenol hypochlorite 7 or 14 1 2, or 3%, also Refer Caveats Filt Samples may be prone to laboratory background NH4 contamination (especially in 1989-2002 period)
NO3+NO2 Nitrate + Nitrite nitrogen mg/m3 as N automated hydrazine or cadmium column reduction, diazotisation.  4 or 12 1 0.5, or 3% Filt Hydrazine reduction to Dec 2001. Cadmium reduction from Jan 2002. Predominantly Nitrate only present
TN Total Nitrogen (Organic N+NO3+NH4)  mg/m3 as N Alkaline persulphate digestion to Nitrate, then Nitrate determination 8 (digestion) 10 10, or 3%


DRP Dissolved Reactive Phosphate mg/m3 as P automated molybdenum blue - ascorbic acid reduction  6 or 13 0.5 0.5, or 3% Filt

TP Total Phosphate mg/m3 as P Acid persulphate digestion to DRP, then DRP determination 2 (digestion) 2 1, or 3%


A340 Absorbance @ 340nm Abs x 1000 Spectrophotometric absorbance reading x 1000, 40 mm cell path 11 1 5 Filt whole number convention adopted
A440 Absorbance @ 440 nm Abs x 1000 Spectrophotometric absorbance reading x 1000, 40 mm cell path 11 1 2 Filt whole number convention adopted
A740 Absorbance @ 740 nm Abs x 1000 Spectrophotometric absorbance reading x 1000, 40 mm cell path 11 1 1 Filt whole number convention adopted
g340 Absorbance co-efficient @ 340 nm  /m Calculated from 340 & 740 nm absorbances, 9 0.05 n/a


g440 Absorbance co-efficient @ 440 nm  /m Calculated from 440 & 740 nm absorbances,  9 0.05 n/a


Ca Calcium   g/m3 Atomic Absorption Spectroscopy (emission) 5 0.05 no info Filt Data for 1989 only
Mg Magnesium    g/m3 Atomic Absorption Spectroscopy (absorbance) 5 0.01 no info Filt Data for 1989 only
Na Sodium g/m3 Atomic Absorption Spectroscopy (emission) 5 0.05 no info Filt Data for 1989 only
K Potassium g/m3 Atomic Absorption Spectroscopy (absorbance) 5 0.01 no info Filt Data for 1989 only
Alk Total Alkalinity (as Calcium Carbonate) g/m3 as CaCO3

Titration, 0.02N HCl tiration to pH 4.9 end-point (using pH meter)

N.B. essentially bicarbonate expressed as ppm CaCO3

1 (403) 0.5 no info
Data for 1989 only
Cl Chloride g/m3 Mercuric thiocyanate (manual adaptation) 1 (407D) 0.1 no info Filt Data for 1989 only
SO4 Sulphate g/m3 SO4 Turbidimetric (with prior sample evaporation/concentration technique for samples with less than 5 ppm) 1 (426C) 1 & 0.2 no info Filt Data for 1989 only
T coli Total coliforms  /100ml Colilert Quanti-tray MPN enumeration, 24hr incubation @ 35 deg C 15 1 refer Caveats
Commenced FEB 2005
E coli Feacal coliforms  /100ml Colilert Quanti-tray MPN enumeration, 24hr incubation @ 35 deg C, (fluorescent finish) 15 1 refer Caveats
Commenced FEB 2005

 

A. PROCEDURAL NOTES:

1) Sampling for the NRWQN programme is carried out by NIWA Environmental Data field teams (14 teams nationwide), covering 77 sites

2) Field observations are made of time (NZST), cloud cover, water temperature, dissolved oxygen (% saturation), clarity (horizontal black disc), stage height, flow, and atmospheric pressure. A conversion of %DO to ppm DO is made back at the laboratory.

3) Samples are colleced in 1 litre polyethylene bottles, sealed air-tight, packed in ice in chilly bins, and dispatched directly to the NIWA Hamilton Chemistry Laboratory (usually received within a 24 hour period).

4) Samples are processed as soon as possible upon receipt, being first warmed to room temperature (sample bottles placed in a water bath). Immediate analyses are made for pH, conductivity and turbidity. Remaining sample is then sub-sampled (both unfiltered and membrane filtered) into smaller 100 ml containers for total nutrients (unfiltered for TN &TP -samples prior shaken for representative aliquot) and dissolved nutrients (membrane filtered for DRP, NH4, NO3 and absorbance measurments for g340 and
g440.) For 1989 (first year only of NRWQN) membrane filtered subsamples were also taken for major anions ( Cl, SO4) and cations (Ca, Mg, Na, K).  Alkalinity analyses were carried out on unfiltered samples.

5) Membrane filtration is through a 0.45 micron membrane filter to remove turbidity and microorgansims

6) Sample containers for total and dissolved nutrients are stored frozen until time of analysis.

7) Remaining sample in the initial 1 litre sample bottles was utilised for the BOD test (5 day incubation at 20 degrees C), in duplicate. N.B. BOD analyses ceased after July 2002, except for sites AK2, RO2, and WA9 which have been left in as indicator sites.

8) TP test: colorimetric finish from 1989 to 1995 incl. done by manual test tube method. From 1996 onward, by automated method.

9)Total & Faecal coliform (E coli) analyses commenced February 2005. Sample collected in separate container, processed immediately
upon receipt at lab (utilising Colilert Quanti-tray MPN enumeration).

10) Periphton observations made directly if possible by wading and using underwater viewer (as for water clarity), ten observation circles  of 0.5m radius, full cross section if possible otherwise a number of transects from either bank. NOTE: A high degree of sujectivity inherent in this measurement, more so for some sites than others. Enquiries should be directed to Dr John Quinn of NIWA Hamilton regarding any data interpretation/clarification preceding preparation/publication of any reports.

B. REFERENCE METHOD NOTES:

1 APHA, 1985
2 Smith et al, Water and Soil Miscellaneous Publication No. 38
3 Manufacturer's Instructions
4 Downes, 1978a; Technicon Auto-analyser (until December 2001)
5 DSIR Chemistry Division reports, CD 2151, 1971 and 1972
6 Downes, 1978b; Technicon auto-analyser (until December 2001)
7 Technicon, 1978; Technicon auto-analyser (until December 2001)
8 Koroleff, 1983
9 Davies-Colley & Vant, 1987 - Refer Miscellaneous Notes 2).
10 Davies-Colley, "Measuring water clarity with a black disk", Limno. Oceanogr., 33(4, part1), 1988, 616-623
11 NIWA Instruction/Adaptation
12 Quik Chem Method 31-107-04-1-A ; Lachat Flow Injection Analyser (from January 2002)
13 Quik Chem Method 31-111-5-01-1-1; Lachat Flow Injection Analyser (from January 2002)
14 Quik Chem Method 31-107-06-1-B ; Lachat Flow Injection Analyser (from January 2002)
15 IDEXX laboratories Inc, Colilert quanti-tray MPN enumeration method
16 Smith et al, 1989, Water Quality Centre Consultancy Report 8016/2, A National Water Quality Network for New Zealand

C. MISCELLANEOUS NOTES:

1) FILT column - indicates use of 0.45 micron membrane filtered sample for analysis

2) ppb (parts per billion) = ug/l (micrograms/litre) = mg/m3 (milligrams/cubic metre)

3) ppm (parts per million) = mg/l (milligrams/litre = g/m3 (grams/cubic metre)

4) Absorbance Coefficients (g340 & g440) - Derived from spectrophotometric absorbance measurements at 340nm and 440nm. These wavelengths are recommended for monitoring concentrations of yellow substance in water i.e. 0.45 micron filter-passing ("dissolved") organic matter. The expression for converting spectrophotometer absorbance measurements to an absorption coefficient is as follows:
g340 = ((In 10) * Corrected Abs 340) / cuvette path length (m); g440 = ((In10 * Corrected Abs 440) / cuvette path length (m);
where (In 10) = Log e 10 = 2.303; Corrected Abs 340 or 440 = spectrophotometric absorbance at the respective wavelengths
(340nm & 440 nm) corrected for absorbance by any residual particulates measured at 740nm. Correction values (which are subtracted from
the 340 and 440 nm spectro readings) comprise respectively the 740nm spectro reading x 2.176471 (ratio of 740/340), and the 740nm spectro
reading x 1.681818 (ratio of 740/440). EXAMPLE: Say sample absorbances at 340, 440 and 740 nm are 0.100, 0.025, 0.002 ABS units using
40 mm path length.
g340 therefore = (2.303*( 0.100-( 0.002*2.176471)))*(1000/40)) = 5.51. g440 therefore = (2.203*(0.025-(0.002*1.681818)))*(1000/40) = 1.25.
For a fuller discourse on absorbance coefficients refer to the following publication:
"Absorption of Light by Yellow Substance in Freshwater Lakes", Authors R.J. Davies-Colley and W.N.Vant,
published in Limnology and Oceanography, Vol.32, No. 2, March 1987.

5) Anions (sulphate, chloride & "alkalinity") & cations (calcium, sodium, magnesium, potassium) analysed monthly for year 1989 only

6) BOD analyses ceased after JULY 2002

7) Bacteriological analyses (Total & Faecal coliforms) commenced February 2005

 

CAVEATS - as at March 2014

1) Flow Data:
Flow data is Instantaneous flow data only, provided by NIWA field teams at the time of site visits from known ratings, estimates, or provided by Regional/District Councils, or Energy Provider, with some data not always accurate (e.g. due to later applied rating changes). For most accurate flow data this should be independently sourced from Flow Data Archives.

2) Nominal Detection Limits: (as presented in the Methods sheet)
For dissolved nutrients (DRP, NO3 and NH4) in particular - Finished laboratory values have been rounded to nearest whole numbers with values that registered between zero and 0.5 mg/m3 given a <1 designation. Initial uncensored data may be provided separately upon request.

3) Accuracy and Precision (General Guide only):
For dissolved nutrients (DRP, and NO3) these range from +/- 0.5 mg/m3 (P or N) at nominal detection limit to +/- 3% above 100 mg/m3.

For dissolved ammonia (NH4-N) the range is +/- 1mg/m3 N at nominal detection limit to +/-3% above 100 mg/m3 (based on repetitive low concentration standards).

However, Ammonia analysis at low concentration is inherently less stable than for DRP and NO3, as may be compromised by backgound laboratory atmospheric contamination during initial sample handling processes, especially for samples already with low NH4 concentrations and this may effectvely alter precision (in the main) up to +/- 2 mg/m3 N. This effect is difficult to counter, or to gauge the real extent of, within each analytical sample batch. Greater NH4 contamination prevalence and spikes were noted (retro-suspected) for period 1989-1994 incl. due to unsatisfactory (cramped and non-air conditiond) laboratory premises. Significant Improvements (tighter results banding) were obtained from 1995 onward following a shift into new premises, with further improvement from 2002 onward when dissolved nutrient analyses shifted from segmented autoanalyser system to unsegmented Flow injection Analyser (FIA), along with upgraded quality control measures.

For other determinands, refer to precision data in Methods Note though that for TN and TP the precision data is based on clean standards. Samples values may be influenced outside these ranges (to an unknown extent) if turbidity is high. i.e. due to difficulty of achieving representative sub-sampling (of suspended particles) from sample container to test tube.


4) Special NH4 and TN Note for 1994 calendar year:
There is no reported NH4 or TN data for 1994 calendar year due to gross NH4 contamination arising from storage of sample pottles for that year in an off-site coolstore freezer later found to have had leaking ammonia-based refrigerant.

5) Total coliforms and E. Coliforms:
The official Colilert test method statistically derived values have been consistently applied i.e. Values may have a decimal placing. A null result is recorded as <1 (the data user however is at liberty to delete decimals, and convert <1 values to zeroes). Highest possible reading using 100ml undiluted sample is 2419.2 (readings greater than this are designated as >2419.2). Dilution of sample prior to analysis (e.g. x10) extends reportable range (dilutions carried out only if if thought warranted). Robust E.coliform results are higher priority than Total coliform results hence many Total coliform results record over-range because sample dilution is not required in the majority of cases to obtain optimum E. Coliform counts. Statistical 90% confidence limit data can be supplied separately upon request.

6) Dissolved Oxygen
From Jan1989 to Dec 2008 the YSI model 57 DO/temperature meter (DO probe employing polaragraphic principle) was used by the field teams. (DO probe employing polarographic principle). From Jan 2009 onward the YSI were replaced with the HACH Model HQ40d DO/temperature meter (DO probe employing luminescent principle). The new technology proved superior re calibration stability and ease of use. However, the probe longevity has proved to be a problem, with numerous probe replacements required, and resulting in loss of some field data (DO and temperature) whenever the probe suddenly ceases to function in the field. In January 2015 the HACH meters were replaced with new model YSI 'Pro ODO' meters and probes (also employing luminescent principle) with no reported problems (as at August 2015).

7) Missing data: represented by "missing"
Usually this indicates that a value could not be assigned due to some circumstance beyond remedial control. e.g. Sample site incaccessible (no sample collected), sample leaked/lost from sample container, sample inadvertently became contaminated, faulty instrumentation or analytical response, or a decision to delete out a grossly anomalous value that cannot be re-checked. On occasion though "missing" may represent data pending inclusion at a later date. e.g. late submission of flow data, repeated analytical results, etc.

8) Data updating:
Scheduled on a 3 monthly basis i.e. quarterly calender updates, but may take some further weeks to action.

9) Site & Data amendments:
May be effected at any time without further notice (Includes closing down of old sites and addition of new sites)

10) Changes in analytical regime:
7.1 Laboratory premises change - Jan 1989 - Dec 19944 the chemistry laboratory operated in a cramped non air-conditioned environment. From Jan 1995 the laboratory moved to new, larger air-conditioned premises that facilitated improved analytical operations.

7.2 Nutrient analyses (DRP, NO3, NH4, TN, TP) - analysed by segmented flow Technicon AutoAnalyser (AA) II, Jan 1989-Dec2001 Data output determined manually from recorder traces until Jan 1995, then replaced by electronic data capture. AA then replaced by unsegmented flow Latchat Flow Injection Analyser (FIA), Jan 2002 to date. For nitrate (NO3) this required a change from hydrazine reduction method to cadmium column reduction method, but retaining same colorimetric finish. Trial testing of all determinands between the Autoanalyser and the FIA showed excellent agreement, with FIA demonstrating better stability.

7.3 BOD - ceased July 2002 for majority of samples. Key indicator sites AK2, RO2 and WA9 retained to date.
Reason: BOD inherently very low at most sites (median all sites = 0.4 mg/m3 O ), with no discernible trends.

7.4 Total and E. Coilform bacteria - commenced February 2002 following stakeholder interest to institute testing.

7.5 Turbidity - change from original HACH model 2100A to model 2100AN March 2012.
Reason: Old meter became inoperable. Measurements considered reasonably comparable. (unable to run actual comparisons of NRWQN samples)

11) Quality Assurance:
In addition to established in-house laboratory QA practices, finished data from the laboratory is regualrly scrutinised for anomalies and dubious values with appropriate measures taken to resolve any issues. Refer narrative in relevant sections in Davies-colley et al, "Twenty Years of New Zealand's National Rivers Water Quality Monitoring Network".

Measuring dissolved oxygen in situ. [NIWA]
NIWA technician Jeremy Rutherford measures water clarity in the Motu river at Waitangirua. [NIWA]
NRWQN analyst, Graham Bryers in Hamilton’s water quality laboratory. [NIWA]