A band ratio is a remote-sensing technique that divides the pixel values of one spectral band by those of another. The resulting image enhances spectral features tied to specific materials while cancelling much of the brightness variation caused by topography and illumination.
Why it matters
Because slope and shadow scale all bands roughly together, dividing one band by another largely removes that effect, leaving differences that reflect surface composition rather than terrain lighting. This makes band ratios a fast, robust tool for geological mapping — picking out iron oxides, clays, carbonates, and hydrothermal alteration that are hard to see in a true-color image.
A concrete example
For Landsat 8/9 OLI, the iron-oxide ratio is Band 4 (red, ~0.65 µm) / Band 2 (blue, ~0.48 µm); the ferrous-iron ratio is Band 6 (SWIR1) / Band 5 (NIR); and a clay/hydroxyl ratio uses Band 6 (SWIR1, ~1.6 µm) / Band 7 (SWIR2, ~2.2 µm), since clays and micas absorb near 2.2 µm. These three are often combined as an RGB composite to flag alteration zones for follow-up. ASTER's finer SWIR bands give even more diagnostic ratios for specific minerals.
Common pitfall
Band ratios should be computed on atmospherically corrected (surface reflectance) data; running them on raw digital numbers mixes in atmospheric and sensor effects. Ratios also exaggerate noise where the denominator is small, and they indicate spectral similarity, not proof — vegetation, soil, and man-made surfaces can mimic mineral signatures, so ground truth remains essential.