Arctic Sea ice, a thin layer between the Arctic Ocean and the atmosphere, strongly influences freshwater flux, ocean mixing, mass transport, and radiation properties. Given recent climate change and rapid sea ice melting, assessing sea ice motion is critical to the navigational safety of Arctic waterways. However, the quality of the motion products from Earth-observing satellite sensors largely depends on satellite resolution, choice of tracking algorithm, and sparse buoy observations. We present an intercomparison of the daily mean sea ice velocity products from Estimating the Circulation and the Climate of the Ocean (ECCO) and National Snow and Ice Data Center (NSIDC) Polar Pathfinder Daily 25 km EASE-grid Sea Ice Motion vectors from 1992 to 2017. ECCO provides an observationally constrained and dynamically consistent global ocean estimate driven by realistic boundary conditions and atmospheric forcing. This study aims to evaluate the seasonal and regional variability of the wind-ice coupling in ECCO and its underlying physics. Across all four seasons, NSIDC underestimates the sea ice drift speed compared to ECCO, where sea ice concentration is at least 15%. This difference follows the Gaussian distribution. During the freezing (October-November) and winter (December-March) months, there is a stronger correlation (r ~ 0.6) in Fram Strait, Chukchi, Laptev, and East Siberian Seas compared to the melting (April-May) and summer (June-September) months (r ~ 0.4). In contrast, the Central Arctic shows a consistent correlation (r ~ 0.7) between both products throughout the year. The Pearson correlation between the daily mean ice motion and the ERA-Interim wind field is lower, and sometimes even negative, during the summer when the ice pack is not in contact with the coasts. This correlation increases during the winter. The influence of the coastline extends to several hundred kilometers from the coast.