The Mars Color Imager (MARCI) camera on the Mars Reconnaissance Orbiter (MRO) is a very wide angle push frame camera with 5 visible color channels. The camera acquires pole-to-pole image swaths on the sunlit side of the planet. The polar orbit of the spacecraft provides daily synoptic coverage that allows monitoring of seasonal cap retreat and interannual changes that occur between Mars years (MY) and over the northern hemisphere summer.  MARCI has a pixel size of 1.9 km and each frame provides a limb-to-limb view with significant overlap of image frames at the pole so that a mosaic can be made to cover all longitudes of the polar region each day. Each MARCI mosaic is composed of images acquired over 12 to 13 orbits, combined in a north polar stereo projection, and red, green, blue (RGB) color images are constructed from three filters centered on wavelengths of  604 nm, 546 nm, and 437 nm, respectively. Each image is time stamped by Earth day and year and solar longitude (Ls). Images for each Earth day are joined in time ordered sequence into an MP4 video.

Prior to Ls ~ 0 and after Ls ~ 172, the pole is not sunlit and MARCI does not image latitudes all the way to the pole so that some frames exhibit a dark center until image acquistion covers the poles. Additionally, off-nadir rolls can create large gores at the poles and these images are removed from the time sequence.   In two previous publications these image mosaic frames were used to monitor seasonal polar cap recession over multiple Mars years (MY) (see related publications).This data set gathers the original MP4 movies of the MARCI mosaics in date and Ls order. Each movie file is titled by the Mars year over which the data were acquired. All original MARCI images are freely available through NASA's Planetary Data System.

Individual image mosaics are created for each Earth day and time stamped with Ls value as described in Calvin et al. (2015) Icarus, doi: 10.1016/j.icarus.2014.08.026. These images frames are then combined into seasonal cap recession movies using Matlab at 4 frames/second as .mp4 image files.  Users should refer to the associated articles for description of image artifacts.