Monitoring plans established under the National Policy Statement for Freshwater Management 2014 (NPS-FM) must identify representative monitoring sites and recognise the importance of long-term trends.
This page contains excerpts from the NPS-FM, cabinet papers, ministers’ speeches, implementation guidance and comments from a recent Environment Court decision which may be useful to councils looking for guidance on the monitoring requirements in the NPS-FM. They highlight the importance of representativeness and trends (ie, the requirement is not about monitoring every drop of water all the time).
National Policy Statement
for Freshwater Management 2014
“Monitoring plans are intended to be practical and affordable. It is not possible for regional councils to monitor every drop of fresh water. Monitoring against freshwater objectives need only be undertaken at representative sites within a region as identified by regional councils. Monitoring plans are also intended to recognise the importance of long term trends in data.”
See page 4, Preamble
National Policy Statement
for Freshwater Management 2014
“By every regional council developing a monitoring plan that:
b) identifies a site or sites at which monitoring will be undertaken that are representative
for each freshwater management unit; and
c) recognises the importance of long-term trends in monitoring results.”
See page 16, Section CB. Monitoring plans.
|Ngāti Kahungunu Iwi Inc v The Hawkes Bay Regional Council  NZEnvC 50 at || |
“It might, perhaps, be appropriate for a Council to regard overall quality as permitting some increase in a type of contaminant (nitrate-nitrogen, for instance) in a particular water body, so long as that was matched or exceeded in its adverse effects by, say, a reduction in some other contaminant, so that the … quality of the water … taken overall, was at least no worse.”
Note that the Court saw some difficulties with the practical implementation of this principle. We are currently looking at options to better enable this approach.
Decision of appeal [PDF, 3.87 MB] [Environment Protection Authority website]
|National Policy Statement for Freshwater Management – Cabinet Paper (May 2011)||The policy intent of objective A2 of the NPS-FM is to recognise “that a bottom line of at least maintaining water quality everywhere is not possible. It allows for some variability in terms of water quality as long as the overall water quality is maintained in a region. Essentially it allows for offsets within a region, including between catchments.”|| |
|New start for fresh water – Cabinet Paper||“New Zealand’s water resources are finite, and so it is not possible to fully meet all demands and expectations in all areas at all times. Outcomes will only be achieved by considering and making trade-offs between values, within a decision-making framework that sets limits and bottom lines.”||Cabinet paper: New start for fresh water|
Freshwater Reform: Consultation on Amendments to the National Policy Statement for Freshwater Management (August 2013)
|“Councils will group water bodies into management units for the purpose of planning and will identify the appropriate points within each unit at which monitoring will be undertaken. The scale of the unit chosen for planning will be the same as that used for the purpose of freshwater accounting. Freshwater objectives (at or above national bottom lines) will apply at the scale of the management unit, rather than individually to each component part of that unit (i.e. not to every tributary individually).”||Cabinet paper:|
|Speech by Amy Adams Govt invests further $2.1 million in community freshwater action (15 March, 2014)||“A large network of sites is currently used for assessing the state of our rivers. These sites were established for a variety of reasons, but the data collected is not necessarily representative of the whole country.”|| |
Govt invests further $2.1 million in community freshwater action [Beehive website]
A Guide to the
National Policy Statement for Freshwater Management 2014 (August 2015)
“Importance of long-term trends
Water quality varies a lot over short time periods. To be able to reliably estimate water quality statistics, it needs to be consistently measured over a long period. Long-term monitoring data should be used to calculate the sample statistics (eg, median, 95th percentile) required both to assess the current state and to determine progress towards freshwater objectives. Short periods of record cannot reliably calculate the sample statistics, and may produce erroneous conclusions about progress towards, and the achievement of, freshwater objectives.
Long-term data should also be used to determine trends to show changes in water quality. This can help in estimating timeframes in which freshwater objectives may be met, and show the direction of progress. In some cases, trends may initially show changes in the opposite direction to the freshwater objective. This could be due to historical contamination or natural variability (eg, seasonal/decadal climate patterns); however it could also indicate previously unidentified or new pressures, or a need to revisit limits and/or methods to ensure they are appropriate to meet the freshwater objectives.”
|Ministry for the Environment. 2014. Freshwater Accounting: Guidance for Regional Authorities.|| |
“As stated by George Box, “All models are wrong, but some are useful” (quoted in Motu, 2013). It is important to remember that models are abstract representations of reality and that they necessarily contain important assumptions and limitations. Furthermore, the amount of data and information available to construct a robust model can also be limited. It is therefore important to be open about the design and inherent assumptions used in a model, because if its structure and input data are inaccurate or uncertain, this will be directly reflected in the model’s outputs. Being transparent and expressing the assumptions, limitations, error and uncertainty associated with predictions is an essential part of using models. It is important to understand these limitations, particularly when they are to be used to communicate with communities involved in freshwater management decisions for an FMU.”
“Measurements and estimates are inherently associated with errors in that they are inaccurate (they vary from the ‘true’ value) and contain uncertainty (there is some level of confidence in how well the estimate reflects the ‘true’ value). This uncertainty arises due to many factors such as limited or missing data, poor quality data, conceptual and structural errors of a model, and boundary condition errors. The expression of uncertainty for an estimate represents the compounding errors of data and the model itself.”
“It is not easy to develop a robust accounting system for a complex real world system.”
Representativeness and statistical power of the New
Zealand river monitoring network. (July 2012)
|“A representative estimate of river condition is one in which the influence or leverage of a sampling site or environmental class is proportional to its abundance or dominance. Data from a highly representative monitoring network will give accurate estimates of water quality and ecological state and trend. It is important to note that accurate estimates of [sic] may be imprecise, which can prelude identifying state and trends with a high level of certainty.”||Representativeness and statistical power of the New |
Zealand river monitoring network
Freshwater programme: Amendments to the National Policy Statement for Freshwater Management – Cabinet Paper (June 2014)
|“We propose an amendment to the National Policy Statement to require councils to develop plans to monitor progress toward freshwater objectives. The new requirement makes it clear that monitoring for freshwater objectives should be undertaken at representative sites within each catchment, as identified by regional councils. The policy is intended to signal to councils that monitoring regimes should be practical and affordable.”|