Description:
DSM is a collaborative workspace for researchers working on digital soil mapping in Australia. This workspace is the initiative of the Advisory Group on Digital Soil Assessment, a working group of the National Committee on Soil & Terrain. The website is part of an interoperable web-GIS maintained by the Centre for eResearch and Digital Innovation (CeRDI) at Federation University Australia (FedUni). The International Union of Soil Sciences Digital Soil Mapping Working Group defines DSM as "creation and the population of a geographically referenced soil database, generated at a given resolution by using field and laboratory observation methods coupled with environmental data through quantitative relationships." Digital Soil Mapping (DSM) utilises numerical methods and information technologies to produce predictive maps of soil types and their properties. DSM relies on traditional field mapping, observations and laboratory analyses for soil data but also utilises spatial models of landscape terrain and remotely-sensed properties. Typically numerical methods such as interpolation algorithms and data mining are used to create the maps. OzDSM in collaboration with CeRDI is developing a digital soil mapping tool, and a demonstrator version is displayed on the site. A range of data is included, such as nutrient budgets and soil pH, and with extra data available for the Corangamite Catchment Management Authority region including landslides, erosion and salinity.
Description:
Spatial soil information is used to support questions on agriculture and the environment from global to local scales. Historically, soil mapping has been used to inform and guide a multitude of land users with their decisions. Demand for specific spatial soil information is increasing in response from a wider range of users operating across agricultural and environmental domains. To satisfy these demands, users must be provided with practical and relevant spatial soil information. Novel approaches are required to deal with global deficiencies in available soil information. A major limitation to this is the plethora of incongruent legacy data with poor spatial and temporal coverage. This research study initially identifies the specific needs of users for spatial soil information with a focus on the requirements of biophysical modellers. Secondly, error sources that hamper Digital Soil Mapping (DSM) are identified, described and assessed using pH in practical and relevant examples. A final aim is to spatially predict soil properties (e.g. clay mineralogy) that underpin soil chemical behaviour. This is achieved by harmonising legacy data in combination with new spectroscopy techniques and a spatial inference approach. The spatial soil information needs of biophysical modellers in Victoria, Australia were found to be consistent with global needs for information including soil water characteristics, organic carbon and effective rooting depth. To accommodate stochastic and epistemic uncertainties in spatial soil information, uncertainty frameworks proved effective to deal with, and understand the limitations of legacy data in spatial inference models. Robust and reliable spectroscopic models for properties that are linked to functions and services delivered by soil were achieved and used in 3D spatial models. These findings will enable a tactical response through the delivery of pertinent spatial soil information that is contemporary, quality assured and sought by users. Learnings presented should enable producers of spatial soil information to be more comprehensive in their delivery of products that are easy to use, accessible and understood by a growing user community.
Description:
Spatial soil information is used to support questions on agriculture and the environment from global to local scales. Historically, soil mapping has been used to inform and guide a multitude of land users with their decisions. Demand for specific spatial soil information is increasing in response from a wider range of users operating across agricultural and environmental domains. To satisfy these demands, users must be provided with practical and relevant spatial soil information. Novel approaches are required to deal with global deficiencies in available soil information. A major limitation to this is the plethora of incongruent legacy data with poor spatial and temporal coverage. This research study initially identifies the specific needs of users for spatial soil information with a focus on the requirements of biophysical modellers. Secondly, error sources that hamper Digital Soil Mapping (DSM) are identified, described and assessed using pH in practical and relevant examples. A final aim is to spatially predict soil properties (e.g. clay mineralogy) that underpin soil chemical behaviour. This is achieved by harmonising legacy data in combination with new spectroscopy techniques and a spatial inference approach. The spatial soil information needs of biophysical modellers in Victoria, Australia were found to be consistent with global needs for information including soil water characteristics, organic carbon and effective rooting depth. To accommodate stochastic and epistemic uncertainties in spatial soil information, uncertainty frameworks proved effective to deal with, and understand the limitations of legacy data in spatial inference models. Robust and reliable spectroscopic models for properties that are linked to functions and services delivered by soil were achieved and used in 3D spatial models. These findings will enable a tactical response through the delivery of pertinent spatial soil information that is contemporary, quality assured and sought by users. Learnings presented should enable producers of spatial soil information to be more comprehensive in their delivery of products that are easy to use, accessible and understood by a growing user community.