Ask what the pixel value represents

A digital science image contains values arranged in a grid. What one value means depends on the instrument, wavelength band, exposure and processing level. It may represent detector counts, a calibrated intensity or a derived quantity. A displayed square’s color is another layer: software maps those values into something a screen can show.

A pixel — open definition can include light from several unresolved features. Imagine a ground-viewing pixel containing both water and shoreline. Its recorded signal need not match either pure material. Increasing the display size makes the square easier to see but does not recover the separate details that were never resolved.

A stored integer is not automatically a physical unit. USGS documents scale factors and offsets for Landsat science products, with different conventions for different products and collections. The documentation tells the reader how to turn an encoded value into the intended quantity and which values should be excluded.

For a fictional product, suppose calibrated value equals stored value times 0.01, minus 2. A stored value of 500 becomes 3 calibrated units. Applying only the multiplication gives 5 and misses the offset — open definition. This is invented teaching arithmetic, not the conversion for a real Landsat product.

Separate three layers

Two displays can map the same measured values to different brightness ranges. One may reveal faint structure while the other emphasizes a bright region. Neither display change necessarily alters the underlying measurement, but the choices can strongly affect what a viewer notices.

Before comparing two pixels quantitatively, check that they represent compatible quantities and have valid calibration. Before interpreting a screenshot, find the source data and processing description. An attractive color gradient is useful communication, but the scientific meaning comes from the measurement and documented transformation behind it.