134-4 A Simple Method for Disaggregating Soil Map Polygons to Produce Estimates of Soil Properties At a Grid Resolution of 90 m for the Globalsoilmap.Net Project.

See more from this Division: S05 Pedology
See more from this Session: New Challenges for Digital Soil Mapping: I
Monday, October 22, 2012: 8:50 AM
Duke Energy Convention Center, Room 252, Level 2
Share |

Glenn Lelyk, Agriculture and Agri-Food Canada, Winnipeg, MB, Canada, Robert A. MacMillan, LandMapper Environmental Solutions Inc, Edmonton, AB, Canada and Scott Smith, Pacific Agri-Food Research Centre, Agricutlure and Agri-Food Canada, Summerland, BC, Canada
The GlobalSoilMap.net project has a goal of producing maps of predictions of soil properties at a spatial resolution of 90 m for the entire world. Agriculture and Agri-Food Canada, as the lead agency representing Canada in the North American node of this project, is working with the USDA-Natural Resources Conservation Service to jointly develop and test methods to create GlobalSoilMap.net-compliant gridded soil map products.  A pilot area straddling the border of Manitoba, in Canada and North Dakota, in the USA was selected for developing, applying and comparing alternative methods of disaggregating soil polygons and predicting soil properties. The initial approach favored for use in both Canada and the USA involves investigating the disaggregation of existing legacy soil polygonal maps into their component soils and associated reported soil properties by relating soils to landform positions along toposequences or catenas. The three inputs for the disaggregation process are a polygonal soil map, a 90 m shuttle radar topography mission digital elevation model and a database containing representative soil property values at depths specified for GlobalSoilMap.net for all occurring soils. The polygon disaggregation is based on computing fuzzy likelihood values for landform classes for every 90 m grid cell then associating a single most likely soil with each defined landform class in each polygon, or group of similar polygons (e.g. a map unit). The proposed method is demonstrated to be capable of producing estimates of within-polygon variation in soil properties that are consistent with empirical knowledge about how soils and soil properties vary with landform position within typical soil-landform map delineations.  It is felt that the method has the potential to be generic enough to be applied as is, or in slightly modified form, and in an automated fashion, almost anywhere in Canada or the USA.
See more from this Division: S05 Pedology
See more from this Session: New Challenges for Digital Soil Mapping: I