Soil mapping: different information needs require different approaches
Congratulations to our colleagues across the Tasman for the recent launch of the Australia Soil and Landscape Grid.
The Grid is a way of collecting together the best available information on soil attributes (such as pH, soil carbon, clay etc.) down to a depth of 2m at a 90m2 resolution (about the size of a football pitch).
Because it is compiled from a range of information, included derived or interpreted data, it also provides estimates of uncertainties.
The Grid is based upon a framework agreed by the Global Soil Map initiative and relies heavily on advances made in informatics, particularly around data standards. In time, the Grid will support the delivery of a range of outcomes (e.g. where to site infrastructure projects or which habitats are vulnerable), by providing key soil attributes to incorporate into scientific models.
So how do we compare?
It’s interesting to compare Australia’s Grid with our own national ‘soil information system’, S-map. Our system provides soil information down to 1m depth rather than 2m, but we can provide information at a finer scale than Australia’s 90m2 resolution.
Our system includes the National Soil Database (containing point data) as well as an inference engine which takes this data, together with large amounts of legacy data (some dating back to the 1930s) and new polygon data (maps, surveys, reports), to produce an online national map of soil spatial variability.
At present 26% of New Zealand and 56% of high class (productive) land has S-map coverage – largely paid for by investment from regional councils, and to a lesser extent the primary industries. A hybrid approach to increasing coverage has been used, focusing on conventional mapping and the use of polygon data for the intensive lowlands, and applying digital soil mapping techniques – like the Australians (using the same techniques as in Global Soil Map) – for the hilly terrain.
So why the different approaches?
It’s a question of needs. In New Zealand, we need soil information at a higher resolution than that offered by the Australia Soil and Landscape Grid. Essentially, New Zealand needs sub-catchment to regional-scale information to support both primary production and the water reforms. The Australian Grid product has been developed primarily to support Australia’s national research infrastructure, and is focused on providing soil attributes for modelling and research purposes.
Here in New Zealand, we have a long investment in collecting soil data, dating back to the 1930s, and so our system aims to capture and use as much of this legacy data as possible. In addition, the soil series names used in the polygon approach have meaning for many landowners and managers.
We are also focussed on providing a good information supply chain. S-map has been established online, which allows users to access information, free of charge, for their locality, in an easy to understand way (e.g. through factsheets). 21,000 factsheets have been downloaded by the public in the last six months alone. Data is also available for download from a GIS portal by scientists and other GIS users, or streamed directly through web services to support models and tools such as Overseer.
While we and our Australian counterparts are taking different approaches to soil mapping, we are both working towards the same Global Soil Map (GSM) goal of using the latest technologies to create a digital soil map for the world.
Our key soil experts on both sides of the Tasman are working together in leadership roles on the Oceania Node for the GSM. The New Zealand Landcare Research team helped create the critical data standard, ANZSoilML, that supported the development of the Grid and we had team members present in Australia at the launch.