Temporal decorrelation is probably too high to use image at a yearly base.
with ABCD i mean the single images you have. You create pairs of every subsequent ones. First+second, second+third, third+fourth,…
It is unlikely that you will get good results with one image per year (unless you are in arid regions without any vegetation cover).
Maybe these materials help you to understand the theories behind InSAR for subsidence monitoring: http://www.esa.int/About_Us/ESA_Publications/InSAR_Principles_Guidelines_for_SAR_Interferometry_Processing_and_Interpretation_br_ESA_TM-19