The DSMART prediction accuracy can be largely explained by characteristics of soil survey and map unit
components, and the landscape variability (map + landscape model accuracy for training (testing) data is 69.
This allows both soil and landscape limitations and other parameters to be portrayed using the same map unit
Survey creation methods, soil map, and soil map unit
descriptions have been reported by Hartsock et al.
The weighted mean property value is spatially invariant within map unit
However, available reports and databases contain little if any information on soil map unit
purity/composition or on soil property variability within map units
Through this process, as soil surveys are updated, the overall number of soil series exhibited on soil survey maps (and used to formulate map unit
attributes) could be changed through the combination or reclassification of some series.
In this, the soil-landscape subsystem map unit
became the primary analysis framework.
For these products, each map unit
represents an area dominated by one to three soil series (components) each of which has up to 6 layers (horizons).
A map unit
of an area-class map is a set of delineations that arc supposed to have the same soils/properties except for their geographic position (Rossiter 2000).
Hall and Olson (1991) state that according to this paradigm, a soil map unit
should be limited to a single landform position, though in practice this may not be the case.
As attributes are highly variable and may also be sensitive parameters in models, users need to know the expected range of values (preferably as a statistical distribution) for attributes of interest within the map unit
(Foussereau et al.
Soil map icons identify map unit
polygons, and they also represent specific characteristics associated with the natural group for which that particular map unit
polygon is a member.