The Effect of the Geomorphologic Type as Surrogate to the Time Factor on Digital Soil Mapping

摘要

Many environmental variables are frequently used to predict values of soil in locations where they are not measured. Digital soil mapping (DSM) has a long-standing convention to describe soils as a function of climate, organisms, topography, parent material, time and space. It is obvious that terrain, climate, parent material and organisms are used frequently in the prediction of soil properties while time and space factors are rarely used. Time is the indirect factor for the formation and development of soil. Moreover, it is very useful to explicit and implicit estimates of soil age for DSM. However, it is often difficult to obtain time factor. In the absence of explicit soil age data, geomorphologic data are commonly related to soil relative age. Consequently, this study adopts the geomorphologic types (genesis type of geomorphology) as surrogate to the time factor and analyzes its effect on DSM. To examine this idea, we selected the Ili region of northwestern China as the study area. This paper uses geomorphologic data from a new digital geomorphology map as the implicit soil age in predictive soil mapping. For this study, Soil-landscape inference model (SoLIM) was used to predict soil properties based on the individual representation of each sample. This model applies the terrain (topography), climate, parent material (geology) and time (geomorphologic type) to predict soil values in the study area where they are not measured. And the independent sample validation method was used to estimate the precision of results. The validation result shows that the use of geomorphologic data as surrogate to the time factor in the individual representation leads to a considerable and significant increase in the accuracy of results. In other words, implicit estimates of soil age by genesis type of geomorphology are very useful for DSM. This increase was due to the high purity of the geomorphologic data. This means that the geomorphologic variable, if used, can improve the quality of DSM. Predicted value through the proposed approach comes closer to the real value.