Your question is probably somewhat more complex than it appears, because a Landsat image is essentially a quasi-instantaneous snapshot of the shoreline position at the time the image was acquired. There are multiple processes that affect the position of the shoreline (assuming even that this concept is itself well-defined), including:
- sea tides and earth tides, with a frequency of half a day,
- beach erosion, with typical time scales of months to years,
- sea level rise due to climate change, occurring over decades (currently about 4-6 mm per year), and
- in some areas, continental rebound, taking place over thousands of years.
To understand the nature of the mismatch between your observational tool (Landsat) and your process of interest (shoreline change), you should learn about aliasing and the Nyquist frequency; see, for instance:
http://en.wikipedia.org/wiki/Aliasing
http://en.wikipedia.org/wiki/Nyquist_frequency
The only rational approach is to first decide which of the processes mentioned above might be of interest to you, and then select the most appropriate observational tool to match your goal. This may or may not be a satellite instrument. In any case, don't start by selecting Landsat images and ask what you can do with them, as they may or may not be relevant for any particular problem.
Once you have selected the geophysical process that interests you, and its associated typical period (or frequency) of change,
- identify a suitable sensor that provides a high enough frequency of observation to deal with all the geophysical processes that operate at your selected frequency or faster,
- accumulate as much data as possible over the relevant time scales, and then
- use appropriate statistical methods to "smooth out" the undesirable high frequency perturbing events while highlighting the process you are interested in.