S. Riza, - Modeling soil landscapes and soil textures using hyperscale terrain attributes.
Abstract
Digital soil mapping using hyperscale modeling of topographic parameters is a promising approach to capture the multiscale relationship between soil properties and the landscape. Hyperscale analysis sampled and calculated topographic parameters at more than one resolution. This study aims to develop prediction models for estimating the soil particle size fraction by using a hyperscale terrain analysis. We collected 407 topsoil samples from a watershed of 236 km2 near the city of Malang, Indonesia. We related the observations to hyperscale terrain parameters in a multiple linear regression framework as the model’s parameters allow direct geomorphological process interpretations. We found that the accuracy of the model increased when the watershed area was divided in two based on its parent material. The local morphometric variable parameters included in the models provided an overview of the scale of topography parameters that influence the hydrology and soil forming factors. Our models indicated that, besides the parent material, the flow direction in the landscape had a large controlling effect on the spatial distribution of soil texture. The proposed model using simple linear relationships had a high accuracy in prediction and can be used to reveal soil formation processes.
English Abstract
Digital soil mapping using hyperscale modeling of topographic parameters is a promising approach to capture the multiscale relationship between soil properties and the landscape. Hyperscale analysis sampled and calculated topographic parameters at more than one resolution. This study aims to develop prediction models for estimating the soil particle size fraction by using a hyperscale terrain analysis. We collected 407 topsoil samples from a watershed of 236 km2 near the city of Malang, Indonesia. We related the observations to hyperscale terrain parameters in a multiple linear regression framework as the model’s parameters allow direct geomorphological process interpretations. We found that the accuracy of the model increased when the watershed area was divided in two based on its parent material. The local morphometric variable parameters included in the models provided an overview of the scale of topography parameters that influence the hydrology and soil forming factors. Our models indicated that, besides the parent material, the flow direction in the landscape had a large controlling effect on the spatial distribution of soil texture. The proposed model using simple linear relationships had a high accuracy in prediction and can be used to reveal soil formation processes.
| Item Type: | Article |
|---|---|
| Depositing User: | Christinia Minarso |
| Date Deposited: | 15 Dec 2021 09:05 |
| Last Modified: | 15 Dec 2021 09:05 |
| URI: | http://repository.ub.ac.id/id/eprint/187041 |
Actions (login required)
 |
View Item |