1st International Conference of Soil and Agriculture:
TOWARDS SOIL SUSTAINABILITY
11–13 June 2024 • Lublin–Nałęczów, Poland

https://doi.org/10.24326/ICSA1.PP.03
Published online: 4 October 2024

Problems of deriving pedotransfer functions for
the assessment of soil bulk density from 3D-imagery porosity data
Maja Bryk*

Institute of Soil Science, Environmental Engineering and Management, University of Life Sciences in Lublin, Leszczyńskiego 7, 20-069 Lublin, Poland
* Corresponding author:

Abstract
Recently, Bryk and Kołodziej (2023, https://doi.org/10.3390/s23041852) proposed new pedotransfer functions (PTFs) for predicting soil bulk density using data on soil macrostructure obtained from 2D image analysis. Building on this concept, we sought to validate the usefulness of 3D soil structure analysis for predicting soil bulk density, assuming that no other data is available. The research utilised the Soil Structure Library of the Helmholtz Centre for Environmental Research and the BonaRes Centre for Soil Research (https://structurelib.ufz.de/lit/). The information on 128 soils was extracted from the database: soil type according to the WRB classification, texture, bulk density, total organic carbon content (TOC), and porosity measured by CT-tomography (3D-porosity) at the voxel size of 0.019 or 0.02 mm. Preliminary study indicated that: (i) Samples of the same soil, with the same TOC and texture, exhibited different 3D-porosities at equivalent bulk density. The bulk density ranged from 1.20 to 1.64 Mg/m³, while the 3D-porosity was in the range of 0.011–0.218 cm³/cm³. (ii) For the same soil, the coefficient of variation of 3D-porosity at the same bulk density value ranged from 6.4% to 66.4%, and no relationship was observed between this value and the number of replicates, which ranged from 2 to 9. (iii) Consequently, the relationship between 3D-porosity (Y) and bulk density (X) was very weak (Y = −0.194X + 0.389, R² = 0.169). However, it correctly showed a general decreasing trend of bulk density with increasing 3D-porosity. In conclusion, the data collected in the analysed database did not allow the development of a useful PTF. Upon analysis of the data, it can be assumed that the samples for bulk density measurement by the weight method and the samples for CT-tomography originated from different locations or soil layers of the study site. However, the analysis indicated which factors should be taken into account to establish the actual relationship between bulk density and porosity from 3D-image analysis. Namely, it is recommended that measurements of bulk density and 3D-porosity be conducted on the same soil sample whenever possible. Also, the voxel size should be constant throughout each set of CT-tomography measurements. The creation of a PTF relating bulk density and 3D-porosity could facilitate the use of existing resources of 3D soil structure images for the estimation of missing bulk density values, particularly for archival data.

Keywords: bulk density, 3D soil structure, image analysis, porosity

How to cite
Bryk M., 2024. Problems of deriving pedotransfer functions for the assessment of soil bulk density from 3D-imagery porosity data. 1st International Conference of Soil and Agriculture: Towards Soil Sustainability. https://doi.org/10.24326/ICSA1.PP.03