Since its inception in 1990, the central hypothesis guiding PAL research has been that the phenology, duration and extent of annual sea ice cover is the main determinant of key life history success and ecosystem processes. The Palmer LTER study site along the western Antarctic Peninsula (64-70oS, 64-78oW) extends 200 km from the nearshore (<100 m depth) to the continental slope region (>3000 m depth) south of the Antarctic Circumpolar Current; and 700 km from near Palmer Station on Anvers Island in the north to the scarcely-explored Charcot Island region in the south. Sea ice cover ranges from 2-3 months in the north to >6 months in the south (Figure 3), with much of the region now being ice-free in summer. The mean air temperatures are +4.1oC in summer and -2.4oC in winter (2003-14). Mean December through February surface air temperature has increased by 7oC since 1950, the most rapid seasonal increase on the planet. The WAP region is warming rapidly and experiencing a concomitant reduction in sea ice cover. As a consequence of the dependence of biotic systems on the ice, the ecosystem is changing from that typical of polar seas to a system with more microbes at the base, and declining populations of apex predators. Some changes expressed over large spatial and temporal scales (phytoplankton, krill) and still too small to be detected at the scale of the LTER survey grid. In contrast some key processes governing ecological and biogeochemical dynamics have characteristic time and space scales of just a few days and 1-10 km. One example is concentration of diatom blooms and Adelie penguin colonies near submarine, cross-shelf canyons. Thus for both large- and small scale phenomena, we require new, autonomous, high-resolution observing systems to gain new understanding of important mechanisms of ecosystem response to climate change.