Suppose you find an image of the world that is just that: an image with no spatial information attached to it. You can tell that someone used GIS to produce it, but when they published it, they converted it to a JPEG or another web-friendly format. It looks like the projection is WGS1984. Here’s an easy way to give that image some spatial information so that you can use it as a background, or as an animation. Keep in mind that you won’t be able to use the pixel values for analysis.
Load the image into ArcMap, open the layer properties, and navigate to the Source tab. You can use the number of columns and rows to derive the extent of the image because an image with no spatial reference always has its bottom-left corner at (0, 0). The top-left corner will be (0, # Rows), top- right will be (# Columns, # Rows) and the bottom-right corner will be (# Columns, 0). Using the Link Table from the Georeferencing toolbar, you’ll want to pair up each of the four corner points with its counterpart on the global WGS 1984 layout. It might help to draw a diagram:
The extent of the WGS 1984 image will always be what’s shown below, with one caveat: you may have downloaded an image that doesn’t include data beyond certain latitudes (±60 and ±75 are common). It’s a good idea to use an Esri basemap to verify that the continents line up the way they should. If it’s off, you’ll have to shrink the vertical extent by adjusting the Y Map values accordingly.