Minimize of moisture effects from laboratory simulations of in-situ Vis-NIR spectral for the prediction of soil salinity

In-situ spectral is a superior means of determining soil properties than indoor spectral, but soil moisture (SM) has been a key limiting factor in the accuracy of in-situ spectral. The effective removal of the SM effects is an ongoing challenge for in-situ spectral of soil properties. The present studies have typically focused on SM removal only for a smaller SM interval, but there is a wide range of SM variability in field environments. We assume different optimal humidity correction algorithms for different SM intervals. Our research collected 207 soil samples (0–20 cm) from Kongterik Pasture Nature Reserve in Xinjiang, China, categorized by salinity (EC < 15, 15–45, >45 dS m−1). Wetting experiments were conducted, and spectral data were collected at different SM levels during air-drying. Comprehensive comparison of the ability of external parameter orthogonalization (EPO), direct standardization (DS), and orthogonal signal correction (OSC) to remove SM from different salinization soils. Results showed varying SM correction strategies based on salinity and moisture. For soils with EC < 15 dS m−1 and SM < 10 %, 15 < EC < 45 dS m−1 and SM < 15 %, and EC > 45 dS m−1 and SM < 21 %, high accuracy predictions were achieved without SM correction. As soil salinity increases at the same SM level, moisture's impact on spectral estimation decreases. Optimal SM correction algorithms differ for various SM intervals and salinity levels: EPO removes the effects of moisture best for EC < 15 dS m−1 and SM < 21 %, DS for SM > 21 %. For soils with 15 < EC < 45 dS m−1 and SM below 10 %, EPO proves optimal for minimize SM, DS for 10–37 %, and OSC is most suitable when SM exceeds 37 %. OSC correction was best for EC > 45 dS m−1 and SM > 21 %, EPO for SM < 10 %, and DS prediction is best when SM is between 10 and 21 %. Therefore, to achieve precise in-situ spectral soil salinity detection, select the appropriate moisture correction algorithm based on SM and salinity levels.