Remote sensing depends on detecting electromagnetic radiation that has interacted with Earth's surface or atmosphere. Different portions of the spectrum — visible (0.4-0.7 um), near-infrared, shortwave infrared, thermal infrared, and microwave — carry different information about surface materials because each material has a characteristic spectral signature: a pattern of reflectance, absorption, and emission that varies with wavelength. Atmospheric windows — wavelength bands where the atmosphere is relatively transparent — dictate which spectral regions can be observed from space. Water vapor absorbs strongly in parts of the infrared; ozone absorbs ultraviolet; oxygen and CO2 have specific absorption bands. Sensor design therefore targets atmospheric windows to maximize signal and avoid absorption.
Examine spectral reflectance curves for common materials (vegetation, water, bare soil, snow) plotted against wavelength, and overlay atmospheric transmission curves. This reveals both why certain bands are chosen for satellite sensors and why materials that look identical in visible light can be distinguished in the infrared.
From your understanding of the electromagnetic spectrum, you know that electromagnetic radiation spans a continuous range of wavelengths from gamma rays to radio waves, and that different wavelengths interact with matter in fundamentally different ways. Remote sensing applies this physics to observe Earth from a distance — typically from aircraft or satellites — by measuring the electromagnetic radiation that is reflected, emitted, or scattered by the surface and atmosphere.
The spectral regions most important to remote sensing are visible (0.4-0.7 um), near-infrared (0.7-1.3 um), shortwave infrared (1.3-3.0 um), thermal infrared (3-15 um), and microwave (1 mm-1 m). Each region carries different information. Visible and near-infrared reflectance reveals surface color, vegetation health (chlorophyll absorbs red and blue, reflects green and strongly reflects near-infrared), and water turbidity. Shortwave infrared is sensitive to mineral composition and soil moisture. Thermal infrared measures emitted heat, enabling temperature mapping of land, ocean, and clouds. Microwave radiation penetrates clouds and can sense soil moisture, ice thickness, and surface roughness.
The atmosphere is not uniformly transparent. Atmospheric windows are wavelength bands where absorption by water vapor, CO2, ozone, and other gases is low enough for surface radiation to reach a sensor in space. The visible band is a window (which is why human eyes evolved to use it). There are thermal infrared windows around 3-5 um and 8-14 um, separated by a strong water vapor absorption band near 6-7 um. Satellite sensors are designed to observe within these windows; bands that fall in absorption regions would see mostly atmospheric signal, not the surface. Even within windows, residual atmospheric effects — scattering by aerosols, absorption by trace gases — must be corrected to recover accurate surface measurements, a process called atmospheric correction.
The practical power of spectral remote sensing lies in spectral signatures — the characteristic pattern of how a material reflects or emits radiation across wavelengths. Healthy vegetation, for example, absorbs strongly in red (chlorophyll absorption at 0.68 um) and reflects strongly in near-infrared (leaf cell structure scattering), creating a dramatic "red edge" that is the basis of vegetation indices like NDVI. Water absorbs strongly in the infrared, making it appear dark in those bands. Different minerals have diagnostic absorption features in the shortwave infrared that allow geological mapping from space. The art of remote sensing is exploiting these spectral differences — which are invisible to the naked eye — to classify, map, and monitor the Earth's surface at scales from individual trees to entire continents.