1Data sources & licences
The map is built entirely from published New Zealand government and Crown Research Institute datasets. Each is reproduced here with its original licence; you must comply with that licence (typically requiring attribution) if you reuse the data further.
Base map
- NZ Topo50 Maps (JPEG)
- Land Information New Zealand (LINZ), via the LINZ Data Service. The published 1:50 000 topographic map series — 445 georeferenced sheets, mosaicked here into a single image and resampled from 4.23 m to 8.5 m pixels. Crown copyright, licensed Creative Commons Attribution 4.0 International.
Soils
- FSL Cation Exchange Capacity, FSL Soil Carbon, FSL Salinity, FSL New Zealand Soil Classification v1.1
- Manaaki Whenua – Landcare Research, via the LRIS Portal (lris.scinfo.org.nz). Licensed under the Landcare Data Use Licence v1. Note from the publisher: this data has not been updated since the late 1990s; for current soil data, see S-Map Online.
- Fundamental Soil Layers — New Zealand Soil Classification
- Ministry for the Environment, via the MfE Data Service. Creative Commons Attribution 3.0 New Zealand. This is the same Fundamental Soil Layer polygon set and the same class / group / order coding as the LRIS release above (107,298 polygons in both), republished by MfE for the Environmental Indicators programme; it is the source of the NZ Soil Order and NZ Soil Classification maps rather than a further layer of its own.
- S-map: Soil Depth, Soil Drainage, Soil Texture, Relative Bypass Flow Susceptibility (August 2025)
- Manaaki Whenua – Landcare Research, via the LRIS Portal. The "dissolved" single-attribute releases of S-map, the current national soil survey. Licensed Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International — note this is more restrictive than the other datasets here: non-commercial use only, and no distribution of modified versions. S-map covers about 43% of the land area and is blank elsewhere; unlike the Fundamental Soil Layer, it does map most urban areas.
- NZEnvDS Soil pH (0–10 cm) v1.0
- Manaaki Whenua – Landcare Research (Roudier et al. 2020), via the LRIS Portal. Landcare Data Use Licence v1.
- Geochemical Atlas of Aotearoa New Zealand
- GNS Science (Martin, Turnbull, Roudier, Cavanagh, Rattenbury, Rogers, Vandergoes, Reyes, Gard, Richardson, et al. 2023). GNS Science Report 2023/23. doi.org/10.21420/P9BK-701 ↗. Licensed Creative Commons Attribution 4.0 International (CC BY 4.0).
Land cover
- NZ Cost-Effective Land Cover Map 2023/24
- Manaaki Whenua – Landcare Research (Dymond, Shepherd, Law, Martin, Schindler & Belliss). Derived from Sentinel-2 satellite imagery. Landcare Data Use Licence v1, via the LRIS Portal.
Land use
- LUCAS NZ Land Use Map 2020 (v005)
- Ministry for the Environment (MfE), produced with Manaaki Whenua – Landcare Research. Only the 2020 snapshot is used here (the source dataset also maps 1989, 2007, 2012 and 2016). Published under MfE's open data terms (Creative Commons Attribution 4.0 International unless stated otherwise on the source page).
Ecology
- Ecological Districts and Ecological Regions
- Department of Conservation (DOC) / Protected Natural Areas Programme (McEwen et al.). Crown copyright; DOC data is generally available under Creative Commons Attribution 4.0 International.
Climate
- Downscaled CMIP6 climate projections (CCAM)
- National Institute of Water and Atmospheric Research (NIWA), for the Ministry for the Environment. Only the historical baseline (1995–2014), multi-model median is used — future emissions-scenario projections in the source dataset are deliberately excluded from this map. NIWA / MfE open data terms (enquiries@niwa.co.nz).
- Climate Station Statistics
- NIWA / Cliflo. Station locations and available parameters only — this catalogue carries no gridded values of its own.
- WorldClim 2.1
- Fick & Hijmans (2017), International Journal of Climatology 37(12): 4302–4315. 1970–2000 climatological normals at 30 arc-seconds (~700 m at this latitude). The source of Monthly Rainfall, and of the temperature, solar radiation, wind and vapour-pressure inputs behind Potential Evapotranspiration. Free for academic and other non-commercial use. Note that this is a different source, a different period and a finer grid than the NIWA layers above: a rainfall month and a temperature season are not two measurements of the same thing.
- Potential evapotranspiration (derived)
- Not a published dataset. WorldClim carries no PET, so reference evapotranspiration (ET0) is computed here for the FAO short grass reference crop by the FAO-56 Penman-Monteith method (Allen et al. 1998), from the WorldClim variables that feed it plus SRTM elevation — the same method and inputs as the CGIAR-CSI Global-PET database. It is an atmospheric demand, not the water actually lost by whatever grows at that point, and soil heat flux is taken as zero.
Every map's ? panel, every row of the Data tab and every JSON or CSV download names the dataset the value came from, so you can trace any number back to this list.
2How the data was processed
All datasets were converted to Cloud-Optimized GeoTIFFs (COG) on a single national analysis grid (NZGD2000 / NZTM2000, EPSG:2193, 100 m pixels), so that every layer lines up with every other layer at a given point. The land cover map is the one exception, kept at its native ~10 m resolution because resampling it to 100 m would have discarded most of its value.
- Polygon layers (soil classification, cation exchange capacity, soil carbon, salinity, the four S-map attributes, ecological districts, land use) were rasterized onto the shared grid.
- Non-soil map units are nodata, not zero. The Fundamental Soil Layer gives towns, lakes, rivers, estuaries, ice, quarries and bare rock a map unit with no soil attributes attached, and the numeric fields on those polygons are left at zero. Rasterized literally, that made every city in the country read as a genuine measurement of 0% carbon and 0 cmol(+)/kg exchange capacity. Those units are now masked out, so they report no data. They are detected by a blank class field rather than by name.
- Coarse point and grid data were interpolated rather than left sparse: the 676 geochemistry sample points were interpolated across the country using inverse-distance weighting (12 nearest neighbours), and the ~5 km climate grid was resampled to the shared 100 m grid using cubic-spline interpolation.
- The Topo50 base map is not on the shared grid: the 445 published sheets were mosaicked in the same projection at 8.5 m per pixel — an exact 2:1 average of the 4.23 m scans — and stored as an RGB image with an alpha channel, so the ocean beyond the sheet coverage stays clear rather than becoming a white or black rectangle. Half resolution keeps sheet labels legible to the map's maximum zoom while holding the mosaic to 2 GB instead of roughly 30.
- Climate and geochemistry are clipped to the land. Both were built by interpolation — climate from a ~5 km model grid, geochemistry from 676 sample points — and neither interpolation stops at the coast, so both originally carried values hundreds of kilometres out to sea. Those values described nothing, and being dense floating-point they were most of the file size. Everything outside the land mask is now recorded as no data. The mask is the union of the datasets that are land-only by construction: LUCAS land use, the ecological districts, the Fundamental Soil Layer and the 10 m land cover map, which between them cover 16.6% of the analysis grid.
- Large layers are stored as per-area tiles. Every raster over about 100 MB — the climate surfaces, all 65 geochemistry elements, soil pH, land cover and the Topo50 mosaic — is held as a grid of 200 km cell files (50 km for the two native-resolution rasters) plus one national overview at an eighth of the resolution. The map opens only the cells in view, or the single overview when zoomed out past the point where cells would tell you more, and a click opens only the cell it landed in. Cells that are entirely nodata are never written. No value was altered, resampled or rounded in the process: the cells are cut straight out of the source grid.
- Original source files (shapefiles, GeoTIFFs, NetCDFs, the geochemistry geodatabase) are kept, unmodified, in this project's
archive/folder.
3Disclaimer & limitations
For research and planning use only. This map is not a survey-accurate or legally authoritative record of soil, land cover, land use, ecological or climate conditions at any location. Do not use it as the sole basis for a legal, financial, engineering, resource-consent or land-purchase decision — obtain a site-specific professional assessment for those purposes.
- Interpolated values are estimates. Anywhere data has been interpolated (soil geochemistry from 676 sample points; climate from a ~5 km model grid) the value shown for a given point is a modelled estimate between real measurements, not a direct observation at that exact location.
- Source data currency varies. The Fundamental Soil Layer datasets (soil classification, cation exchange capacity, soil carbon, salinity) reflect soil survey information last updated in the late 1990s; other layers are far more recent (S-map August 2025, land cover 2023/24, land use 2020, climate baseline 1995–2014).
- Soil data in built-up areas. The Fundamental Soil Layer holds no soil properties for towns and cities, and the modelled soil pH surface is masked over them by its publisher, so those maps report no data over Christchurch, Auckland, Wellington and every other urban area. The four S-map layers do cover most cities; where they do, use them. Nothing here fills the Fundamental Soil Layer's urban gaps by inference — an estimate borrowed from the nearest rural polygon would look like a measurement.
- S-map covers about 43% of the country, concentrated in productive lowlands. Outside it the four S-map maps are blank; that is missing coverage, not an absence of soil.
- Coordinate transformation. Source layers used four different coordinate reference systems; reprojecting between them can introduce a positional error of a few metres, immaterial at the 100 m scale of this map.
- Soil classification is shown at two levels. The source records the New Zealand Soil Classification subgroup as a code (BOT, PIT, ZDH). Subgroup names come from Appendix 3 of the LRIS data dictionary shipped with the source shapefile (nomenclature after Hewitt 1998); the NZ Soil Order layer is the same raster read back to the 15 NZSC orders, so the two layers always agree. Fourteen subgroup codes present in v1.1 are not listed in that appendix and are shown by their order instead.
- Displayed colours come from the source data itself, drawn at whatever resolution the current zoom needs; a click always reads the exact full-resolution value regardless of the zoom level or how the colour looks on screen. Continuous layers are colour-stretched to the 2nd–98th percentile of their values so skewed variables stay readable — the legend states both that range and the full range.
- Climate layers stop at the coastline. They used to extend over the sea, where the values came from an over-extended gap-fill in the processing rather than from the climate model. Those are gone: offshore points now report no data, which is what they always meant.
4Privacy
This map collects no personal data. It is a static site with no server-side component: every click, map pan/zoom and JSON download happens entirely in your own browser. No location you click, no query you run and no file you download is transmitted anywhere beyond the initial page and data files loading from the host serving this site. There is no analytics, tracking or cookies.
5Copyright & reuse
The map application code (HTML/CSS/JavaScript) is original to this project. The underlying data remains the property of its original publishers, listed in Section 1, and is used and redistributed here under their stated licences. If you republish or build on data from this map, attribute the original source dataset, not just this application.