There are basically three types of satellite images used by weather forecasters. The most common is called "visible" imagery. Here, the satellite uses reflected sunlight to make a picture in much the same way that a camera works. Land and oceans are dark, clouds white, and the brighter the white, the thicker the clouds. Visible imagery isn't useful at night when there's no sunlight.

A second type of satellite imagery - infrared - is useful all the time. Here the satellite measures temperature - clouds, being high in the sky, are colder. Typically, the coldest, highest clouds are shaded the brightest white, like those along the eastern seaboard in this image, while a dull gray is used for warmer clouds closer to the ground, such as over Michigan and Ohio. The ground is usually warmest of all and is assigned the darkest shades.

Finally, there's water vapor imagery, which is specially tuned to detect high-altitude water above about 15,000 feet. This imagery not only can identify clouds but also invisible water vapor. By making a movie of these images, meteorologists can get a sense of upper-air winds, even when there are no clouds. This is a very important capability, especially to forecasters trying to predict the movement of storms over the ocean where upper-air winds are poorly observed using other means.