In digital image processing of remote sensing data:

• Supervised classification relies on user-defined training sites (ground truth areas of known land cover types) to train statistical algorithms (e.g., Maximum Likelihood, Random Forest) to classify the rest of the image.
• Unsupervised classification groups pixels into spectral clusters based purely on numerical similarity without prior training datasets.

Microwave remote sensing operates in the $1\text{ mm}$ to $1\text{ m}$ wavelength domain and penetrates clouds, haze, and rain day and night. Its proven applications encompass all five domains:

• Soil moisture estimation via radar backscatter and microwave dielectric constant variations.
• Forest and crop canopy volumetric structure via Synthetic Aperture Radar (SAR) polarimetry.
• Snow cover and snow water equivalent mapping.
• Atmospheric moisture sounding profiles via satellite microwave sounders (e.g., AMSU).
• Ocean circulation, altimetry, currents, and surface wind vectors via radar altimeters and scatterometers.

Atmospheric optical phenomena:
Assertion (A) is true: The blue color of the clear daytime sky is caused by Rayleigh scattering of incident solar radiation by atmospheric gas molecules ($ ext{N}_2, ext{O}_2$). Scattering intensity is inversely proportional to the fourth power of wavelength ($I \propto 1/\lambda^4$), causing shorter blue wavelengths to scatter much more efficiently than longer red wavelengths.
Reason (R) is false: The blueness of the sky is an intrinsic scattering phenomenon in the upper atmosphere, completely independent of ocean reflection.