الفهرس 
REVIEW OF LITERATUREOnly 14 pages are availabe for public view
Abstract
The main objectives of the current study are to assess the capabilities of spectral remote sensing data in quantitatively forecasting certain soil properties, develop prediction models for some soil variables using PLSR modeling approach and utilizing digital photos to estimate soil color quantitively. Soil samples were collected from the study location in Al Wadi El Gadid Desert, New Valley Governorate, Egypt. The ”ASD FieldSpec 4” equipment was used to gather the reflection of 93 selected soil samples across the whole optical spectral range of 350 to 2500 nm.
The prediction models for soil clay content and sand content exhibited a relatively poor correlation (R2) for both the training and validation soil groups. Moreover, the soil salinity and calcium carbonate contents showed a good correlation with soil reflectance in some specific spectral regions. In the examined soil samples, the SWIR1 and SWIR2 bands were demonstrated to efficiently differentiate between different soil ECe values. For both the training and validation groups, the soil ECe prediction model showed an excellent correlation (R2) of 0.93 and 0.98. Regarding CaCO3 it has been demonstrated that hyperspectral data in the Red, NIR, and SWIR1 bands can efficiently discriminate between different soil calcium carbonate levels in the studied soil samples. Finally, two maps of soil salinity and calcium carbonate were constructed using the PLSR model that produced using reflectance data applied to a Landsat 8 image to show the geographical distribution of soil salinity (ECe) and calcium carbonate (CaCO3) in the study area. The maps were produced by the IDW method to interpolate the results of soil sample laboratory investigations.
To assess the color property of soil, the Munsell color system is widely utilized. The Munsell color chart (HVC) (descriptive technique) and a digital camera (RGB) (Canon) with 20.1 megapixels were used to identify the color of the soil sample. The resulting HVC values were analyzed and compared to HVC obtained from Munsell Soil color evaluation using correlation techniques.
In conclusion, this study found that data from soil spectral reflectance in the 350-2500 nm region may be used to estimate soil salinity and calcium carbonate contents. The use of multivariate calibration (PLSR) allowed the important information from reflectance spectra to be extracted and correlated with soil salinity and calcium carbonate contents.