Data in visible bands of red (0.64µm), green (0.51µm) and blue (0.47µm) recorded by the Japanese Himawari-8 satellite are combined and Rayleigh scattering modified to produce an RGB product called true-color image, so named because its color is similar to what may appear to the naked eye. The brightness, color and sharpness of a true-color image between sunrise and sunset are affected by different sun angles and water vapor content, and may vary even when the sky is clear. Generally speaking, varying shades of blue represent oceans or lakes; land areas may look brown or green depending on different level of vegetation cover. Phenomena such as dust storm, smoke, and volcanic dust are more easily discernible in true color images than in visible-light or infrared images.

True color relies on solar reflectance, which is not available at night. For the purpose of data continuity, nighttime images are produced by using the GeoColor algorithm, which is composed of infrared channel (10.8μm) and near-infrared channel (3.9μm), and which uses city lights as a static background.

Figure 1a. An example of true-color image, which looks milky white because of atmospheric scattering of sunlight. Figure 1b. The same image after Rayleigh correction, as appears on the CWB website.
Figure 2a. A true-color image at 4:00 p.m., April 28, 2017, when the sky was clear over Inner Mongolia. Figure 2b. A true-color image at 4:00 p.m., May 3, 2017, which shows a dust storm event (red box) over Inner Mongolia.
Figure 3a. A single infrared (11.2μm) image of Taiwan at 6:00 a.m., 13 September, 2017. The low clouds or fog inside the yellow box is not easily discernible since its brightness temperature is close to that of the surface. Figure 3b. The GeoColor image displayed on the CWB official website. It’s the same image as Figure 3a, but channel difference between 3.9μm and 11.2μm is used to highlight said low clouds or fog in the yellow box. Varying shades of gray represent higher clouds, coral red represents low clouds or fog, and city lights appear in yellow.
Figure 4. Image of the area between daytime and nighttime areas is produced by using the specific composite RGB method, in which the opacity increases from 0 to 1 as the solar zenith increases from 70 to 84 degrees.