Assumebally, If the scale of the impact of extra-solar powers overcome to local and close-environmental albedo, images could be nearly identical, 

If you map data to the 24-category USGS data, then albedo will be given at the model start time, based on the LANDUSE.TBL. Since the initial value is dependent on the landuse type, however, the Noah LSM (since V3.1) does not use the albedo (or any other land properties) from the LANDUSE.TBL. It uses values from the VEGPARM.TBL, and albedo is computed based on vegetation fraction, and boudned by the max and min values given in the VEGPARM.TBL (according to UCAR).

 Do you use the Noah scheme? The difference between the two variables will be important depending the type of soil, the time of the year and the ice/snow (or not) cover. Which area does your domain cover? The "albedo" variable would be an average number depending on the time of the year of your simulations. The background would be the actual one, for instance if you had snowfall during your run, you should have more albedo than without it.