A comparison of soil texture measurements using mid-infrared spectroscopy (MIRS) and laser diffraction analysis (LDA) in diverse soils

Summary

This comparative study evaluated two rapid spectroscopic and diffraction-based methods for estimating soil texture against conventional sieve-pipette analysis in a geographically and texturally diverse soil sample set. Mid-infrared spectroscopy proved significantly more reliable than laser diffraction for clay prediction, with both methods performing well for typical agricultural soils containing less than 5% organic carbon and less than 60% clay. The work identifies organic carbon as a confounding factor in texture estimation and suggests that the widely adopted < 8 µm clay threshold in laser diffraction should be revised to < 4 µm for improved accuracy.

UK applicability

For United Kingdom agricultural soils, which typically fall within the low-to-moderate organic carbon and clay ranges identified as suitable for both methods, these findings suggest MIRS offers a cost-effective alternative to conventional texture analysis. The results are directly applicable to UK soil survey and monitoring programmes, though soils with unusually high organic matter content (> 5%) may require method adjustment or organic matter removal preprocessing.

Key measures

Clay, sand, and silt content predictions; coefficient of determination (R²) values for calibration sets; accuracy stratified by organic carbon content (< 5% vs > 5%); particle size threshold evaluation (< 8 µm vs < 4 µm)

Outcomes reported

The study compared mid-infrared spectroscopy (MIRS) and laser diffraction analysis (LDA) against conventional sieve-pipette methods for soil texture determination across diverse European and Kenyan soils. MIRS predictions of clay content were substantially more accurate than LDA, though both methods performed well for sand content estimation.

Topic tags