Evaluating the particle densities of subarctic soils using pedotransfer functions and vis–NIR spectroscopy

Peter L. Weber*, Cecilie Hermansen, Trine Norgaard, Charles Pesch, Per Møldrup, Mogens Humlekrog Greve, Emmanuel Arthur, Lis Wollesen de Jonge

*Kontaktforfatter

Publikation: Bidrag til tidsskriftTidsskriftartikelForskningpeer review

1 Citationer (Scopus)
40 Downloads (Pure)

Abstract

The particle density (ρ s) is a fundamental physical property needed for calculating the soil porosity and phase distributions. While ρ s is often estimated using soil organic matter (SOM) content and particle size distribution, the specific densities of each soil component remain unclear in a subarctic agricultural setting. This study aimed to evaluate the ρ s of soils from Southwest Greenland using a three-compartment model (3CM) based on the mixing ratio of SOM derived from loss-on-ignition, mineral particles <20 μm (FC), and mineral particles ≥20 μm (CC). We further evaluated the accuracy of the 3CM against pedotransfer functions (PTFs) and visible near-infrared (vis–NIR) spectroscopic models. A total of 324 soil samples from 16 Greenlandic agricultural fields were investigated, covering a wide range in SOM content (0.021–0.602 kg kg –1) and clay content (0.020–0.185 kg kg –1). Despite their high SOM content, the Greenlandic soils exhibited relatively high ρ s (1.936–3.044 Mg m –3), which together with a large SOM/organic carbon ratio of 2.16 indicated a high SOM density of 1.493 Mg m –3. The 3CM fit on all soils indicated FC and CC densities of 3.047 and 2.713 Mg m –3, respectively, while a subset of soils (n = 203) from the same geological setting resulted in FC and CC densities of 2.738 and 2.731 Mg m –3. Prediction accuracy of the 3CM (RMSE = 0.067 Mg m –3) was similar to PTFs (RMSE = 0.068–0.070 Mg m –3) and better than vis–NIR spectroscopic models (RMSE = 0.091 Mg m –3).

OriginalsprogEngelsk
TidsskriftSoil Science Society of America Journal
Vol/bind86
Udgave nummer4
Sider (fra-til)964-978
Antal sider15
ISSN0361-5995
DOI
StatusUdgivet - 1 jul. 2022

Fingeraftryk

Dyk ned i forskningsemnerne om 'Evaluating the particle densities of subarctic soils using pedotransfer functions and vis–NIR spectroscopy'. Sammen danner de et unikt fingeraftryk.

Citationsformater