Physiographic Environments of New Zealand

A new landscape classification for water quality that better accounts for the influence of the natural landscape.

A Physiographic Environment is an area of similar landscape characteristics.

What are Physiographic Environments and what do they tell us about water quality?

There are two main factors controlling water quality outcomes – the landscape and us.

Water quality varies widely between regions around New Zealand, even where there are similar land uses and pressures. This is because the natural landscape can have a much bigger influence on water quality outcomes than land use on its own. This means in areas with lots of landscape variation, the effect from similar land use activities can be very different. In areas with little landscape variation, land use pressure is the main control over water quality.

Differences in the hydrological, chemical, and physical processes that occur in the landscape alter the composition of the water as it moves through the landscape. The soil and geology impart a unique chemical signature on the water that it interacts with. By analysing the water chemistry and the combination of landscape processes in the contributing area, it is possible to understand ‘how’ and ‘why’ water quality varies.

Areas with similar landscape factors are likely to respond to land use pressure in a similar and predictable manner. This is the basis of the Physiographic Environments of New Zealand landscape classification for water quality. By understanding the functions that the landscape is performing to minimise the pressures of land use, we can match our land use and management decisions to maintain and improve our freshwater resources.

Physiographic Environments of New Zealand Classification

The Physiographic Environments of New Zealand classification is hierarchical, with a basic 10 Environment Family Classification, and a higher-resolution Sibling Classification, which provides more information over water source and hydrological response (flow pathway), and the role the landscape has in removing or reducing potential water quality contaminants. These environments each have a defining set of landscape characteristics that affect water quality in a predictable manner.

Associated with each classification level are variants. Variants provide additional detail where there is modification of the natural hydrology by artificial drainage, or seasonal and episodic variation in the pathway water takes to leave the land. During heavy rainfall events overland flow may become the dominant pathway or under dry or drought conditions, natural soil zone bypass may occur due to cracking soils. Variants change the predicted contaminant profile for the Physiographic Environment.

Physiographic Environment Families

The summary descriptions provided here are for the Family and Sibling Classes of the Physiographic Environments Classification. For each Environment a video summarises the key hydrological pathway water takes to leave the land, how our actions have modified natural hydrology through drainage, and the inherent risk of contaminant loss from the land. We then match the likely water quality effects in the Physiographic Environment as a result of land use pressure and some best-suited actions to minimise contaminant losses.

Use the information from this section in ACTIONS to learn more about land management strategies and intervention options for reducing contaminant generation and loss.

Variants

Variations to the predicted hydrological pathway occur due to human modification of the natural hydrology or seasonal dryness which alters the pathway water takes after rainfall. These variations result in different contaminant risks to water quality and will require different actions to minimise the effects of land use activities.

Physiographic Method

Researchers at Land and Water Science have developed a new methodology to integrate water quality data with existing map layers (such as soil, geology, topography, and land cover) to conceptualise and model the processes that control the spatial variability of water quality across New Zealand. Physiographic science works ‘backwards’, using the composition of water to trace the water’s journey back through the landscape to understand the landscape processes that control water composition and quality. Physiographics is based on the inherent (or natural) properties of the landscape.

The scientific approach was first developed in 2016 at Environment Southland and used by the council within its regional plan (Rissmann et al., 2016, Hughes et al., 2016). Since then, the method has evolved as part of the Our Land and Water National Science Challenge ‘Physiographic Environments of New Zealand’ project and was applied nationally with the support of several regional councils.

There are three papers on the physiographic approach available through the following links:

How are Physiographic Environments mapped?

The physiographic method brings together map data for climate, topography, geology, soils, and hydrological controls with analytical chemistry at a national scale to produce a water quality specific landscape classification. The method utilised twelve pre-existing geospatial datasets (e.g., topography, soil, geology, land cover, climate, etc) and >10,000 ground and surface water samples from 2,921 monitoring sites across 8 regions of New Zealand.

A beautiful 3D diagram with nine sections each listing the data sources used for classification.
Data sources used to develop Physiographic Environments.

A total of 16 process attribute gradient maps were defined nationally (2x climatic, 8x hydrological, 2x redox, 4x weathering, and 1x geothermal). Of these 16 maps the majority were classed according to hydrochemical measurement data from surface and groundwater sampled between 2014 and 2018. The process attribute gradient maps directly related to water quality (shaded) are available in Maps.

Summary of the 16 national-scale process-attribute gradients (PAG). Relevant datasets and attributes are also shown. Shading shows the maps available to view in the maps section.
Process Process attribute gradient PAG Relevant datasets and scales Attributes
Atmospheric Precipitation (rain, snow, hail) source O18 8 m DEM, δ18O-H2O precipitation isoscape (4 km2 pixel) δ18O-H2O, altitude, distance from the coast
Precipitation volume PPT Annual average rainfall in mm (5 km2 pixel) Precipitation volume
Hydrological Recharge domain and connectivity RCD Soil surveys (1:50,000), Aquifer type and extent (1:50,000) Altitude, temperature isotherm, river network, Typic Fluvial Recent soils
Overland flow OLF Soil surveys (1:50,000), 8 m DEM Soil texture, drainage class, permeability, slope, area of developed land
Deep drainage (vertical drainage) DD Soil surveys (1:50,000) Drainage class, permeability, depth to slowly permeable horizon
Lateral drainage LAT Soil surveys (1:50,000) Drainage class, permeability, depth to slowly permeable horizon,
Artificial drainage ART Soil surveys, 8 m DEM, Land Cover (1 ha) Drainage class, permeability, depth to slowly permeable horizon, slope, agricultural land cover
Soil slaking and dispersion as a soil hydrological index HYD Soil surveys (1:50,000) Soil texture, drainage class, permeability, area of developed land
Natural soil zone bypass (cracking soil) NBP Soil surveys (1:50,000) Cation exchange capacity, pH
Equilibrium water table and aquifer potential EWT Water Table Model (0.04 km2 pixel) Modelled water table depth
Redox Soil reduction potential SRP Soil surveys (1:50,000); soil chemistry profile points. Drainage class, carbon content
Geological aquifer reduction potential GRP Geological surveys (1:50,000 - 1:250,000) Rock type (main and sub rocks, geological descriptions)
Weathering Soil acid neutralization capacity SANC Soil surveys (1:50,000); geochemical baseline survey (8 km2) Soil pH, cation exchange capacity
Geological acid neutralization capacity GANC Geological surveys (1:50,000 - 1:250,000); geochemical baseline survey (8 km2) Rock type
Surface/top regolith strength SGC Geological surveys (1:50,000 - 1:250,000) Rock type and strength
Basal regolith strength BGC Geological surveys (1:50,000 - 1:250,000) Rock type and strength
Geothermal High enthalpy geothermal (≥180 °C) GTH Log of resistivity (limited to extent of Taupo Volcanic Zone) Resistivity

PAG: process attribute-gradient; DEM: digital elevation model

The process-attribute gradient (PAG) maps listed above were subsequently combined into Physiographic Environments guided by the performance to represent dominant processes controlling water quality. This involved ranking the process maps according to their sensitivity to generate landscape units. Where each unit, “Physiographic Environment,” is assumed to respond to land use pressure in a predictable and broadly similar manner.

Five classification maps of New Zealand of showing the Atmospheric, Hydrological, Redox, Chemical and Physical Weathering maps.
Summary method for process attribute gradient map generation and validation for the national application.

What is the scale of the classification?

The spatial scale of the information varies between the datasets used. The resulting map is generally at 1:50,000 which is accurate to approximately ±100m on the ground. Incorporating higher resolution datasets is the goal of future research.

What is the performance of the classification?

The ability of the process maps to represent steady-state water quality, as indicated by nitrogen and phosphorus species, sediment, and E.coli (a microbial indicator), was assessed by combining an independent dataset of 811 long-term surface water quality monitoring sites and a map representing the gradient in land-use intensity. The modelling identified regional climatic and geological variation as key controls on water quality variation across New Zealand.

Two classification maps of New Zealand, one showing the three surface water capture zones (Hydrochemical, Water Quality, and Combined). The other map shows the land use intensity from Low to High.
In Figure A the red points show where surface water samples were taken, and yellow points show where groundwater samples were taken to develop and test the classification. Blue points show where surface water samples were taken to test the classification only. The capture zone represents the area that contributes to the water quality monitoring site and shows how well areas of New Zealand were represented. Areas with little to no water quality measurements are predominantly natural state. Figure B shows a representation of land use pressure nationally.

The performance was tested using a statistical approach known as a coefficient of determination (R² value and pronounced “R squared”). An R² value of 1 indicates an exact match between the actual and predicted value or 100% accuracy and a 0 R² value indicates no relationship or 0% accuracy. As a rough rule of thumb, R² values of greater than 0.7 are considered strongly correlated, R² values between 0.5 and 0.7 are moderately correlated, R² values between 0.3 and 0.5 are weakly corelated, and an R² values less than 0.3 can be described as very weak or no correlation. A cross validated approach was also used. This means some of the measurement data used to test performance is withheld from the analysis during testing and swapped in and out to see how the R² value varies. A cross-validated R² value is considered more robust.

The process gradient maps were tested in regional subsets due to the significant variation in landscape processes and land use pressure. Cross validated R² values for environmental contaminants are listed below as a minimum-maximum and median performance:

  • Total Nitrogen R² values ranged between 0.71 - 0.90 (median of 0.78)
  • Nitrate-Nitrite Nitrogen R² values between 0.71 - 0.83 (median of 0.79)
  • Total Phosphorus R² values between 0.63 - 0.85 (median of 0.73)
  • Dissolved Reactive Phosphorus R² values between 0.57 - 0. 76 (median of 0.73)
  • Turbidity between R² values 0.48 - 0. 92 (median of 0.69)
  • Clarity between R² values 0.50 - 0. 89 (median of 0.62)
  • E. coli between R² values 0.59 - 0. 75 (median of 0.74).

Learn more about these water quality indicators in SCIENCE – WATER QUALITY CONTAMINANTS.