Spectral index selection method for remote moisture sensing under challenging illumination conditions

Area of Science:

• Geosciences
• Remote Sensing
• Spectroscopy

Background:

• Passive solar illumination in the Short-Wave Infrared (SWIR) spectrum is susceptible to intensity variations from atmospheric conditions and spectral absorption.
• Accurate soil moisture assessment via remote sensing requires methods robust to these environmental and spectral challenges, especially under non-ideal illumination.

Purpose of the Study:

• To develop and demonstrate a computational hyperspectral image analysis method (HIAM) for deriving optimal reflectance indices.
• To enhance the accuracy and applicability of remote sensing for soil moisture content determination.

Main Methods:

• Utilized histogram analysis of hyperspectral images of wet and dry soil samples.
• Tested contrast ratios and wavelength pairings to identify optimal spectral indices for soil moisture recovery.
• Validated the method using local soil samples under laboratory and field conditions, and public soil databases.

Main Results:

• The developed spectral index demonstrated robustness to varying lighting conditions.
• Soil moisture content was recovered with a Root Mean Square (RMS) error better than 5% across diverse soil types and conditions.
• The HIAM method proved independent of specific material types.

Conclusions:

• The HIAM method offers a reliable approach for accurate soil moisture estimation using remote sensing.
• Its material independence suggests broad applicability to various biological and man-made samples beyond soil.
• This advancement extends the utility of SWIR remote sensing to challenging environmental conditions.