General, global, long-term, and comprehensive phenological observations are required to accurately evaluate the variability of ecosystem structures, functions, and diversity due to environmental changes. Satellite remote-sensing (S-RS) is an useful tool to evaluate interannual and seasonal variations in canopy phenology from plot to global scales with a high temporal resolution. However, from the in situ ecological research viewpoint, further field observations are required to obtain ground-truth for S-RS on various ecosystem sites. Toward this aim, we established a robust, universal, continuous, and long-term ground observation network for various ecosystems: “Phenological Eyes Network” (PEN). For instance, we found that the green-red ratio vegetation index (GRVI) value of 0.0 could detect the timings of leaf-expansion and -fall with robust and general conditions for various ecosystem sites. However, wrong evaluations of the timings of leaf-expansion and -fall by using S-RS data were caused by noise or missing data. Although further ground-truth for a range of sites are required, these results could provide scientific evidences for S-RS observations from the in situ ecological research viewpoint.