Quantitative assessment of soil parameters in Western Tajikistan using a soil spectral library approach

Seiler, Bruno; Kneubühler, Mathias; Wolfgramm, Bettina; Itten, Klaus (2007). Quantitative assessment of soil parameters in Western Tajikistan using a soil spectral library approach. Proceedings of the ISPRS Working Group VII/1 Workshop ISPMSRS'07, XXXVI(7/C50), pp. 451-455. International Society for Photogrammery and Remote Sensing

[img]
Preview
Text
25622.pdf - Published Version
Available under License BORIS Standard License.

Download (347kB) | Preview

Soil degradation is a major problem in the agriculturally dominated country of Tajikistan, which makes it necessary to determine and monitor the state of soils. For this purpose a soil spectral library was established as it enables the determination of soil properties with relatively low costs and effort. A total of 1465 soil samples were collected from three 10x10 km test sites in western Tajikistan. The diffuse reflectance of the samples was measured with a FieldSpec PRO FR from ASD in the spectral range from 380 to 2500 nm in laboratory. 166 samples were finally selected based on their spectral information and analysed on total C and N, organic C, pH, CaCO₃, extractable P, exchangeable Ca, Mg and K, and the fractions clay, silt and sand. Multiple linear regression was used to set up the models. Two third of the chemically analysed samples were used to calibrate the models, one third was used for hold-out validation. Very good prediction accuracy was obtained for total C (R² = 0.76, RMSEP = 4.36 g kg⁻¹), total N (R² = 0.83, RMSEP = 0.30 g kg⁻¹) and organic C (R² = 0.81, RMSEP = 3.30 g kg⁻¹), good accuracy for pH (R² = 0.61, RMSEP = 0.157) and CaCO3(R² = 0.72, RMSEP = 4.63 %). No models could be developed for extractable P, exchangeable Ca, Mg and K, and the fractions clay, silt and sand. It can be concluded that the spectral library approach has a high potential to substitute standard laboratory methods where rapid and inexpensive analysis is required.

Item Type:

Conference or Workshop Item (Paper)

Division/Institute:

08 Faculty of Science > Institute of Geography > Integrative Geography > Unit Sustainable Land Management
08 Faculty of Science > Institute of Geography
10 Strategic Research Centers > Centre for Development and Environment (CDE)

UniBE Contributor:

Wolfgramm, Bettina

ISSN:

1682-177

Publisher:

International Society for Photogrammery and Remote Sensing

Language:

English

Submitter:

Factscience Import

Date Deposited:

04 Oct 2013 14:59

Last Modified:

17 Oct 2015 16:12

Related URLs:

BORIS DOI:

10.7892/boris.25622

URI:

https://boris.unibe.ch/id/eprint/25622 (FactScience: 59958)

Actions (login required)

Edit item Edit item
Provide Feedback