Multisensor Earth observations reveal divergent human and land-surface responses across wartime Ukraine

War-related changes in night-time illumination and vegetation are frequently assessed separately, leaving uncertain whether human and land-surface responses follow a common spatial pattern. This study examined 6,402 equal-area, 10 km grid cells across Ukraine using Visible Infrared Imaging Radiometer Suite (VIIRS) night-time radiance, Sentinel-2 and Landsat 8/9 normalized difference vegetation index (NDVI) and normalized burn ratio (NBR), and independently mapped conflict events from the Armed Conflict Location & Event Data Project (ACLED). Pre-war (2018–2021) and wartime (from 24 February 2022) observations defined changes in illumination and optical indices. Robust standardization and principal component analysis yielded human and land-surface response metrics; residuals from their national linear relationship represented relative divergence. The responses were nearly uncorrelated (Pearson’s r = −0.022). The highest conflict-exposure quintile showed the strongest decline in night-time illumination despite a positive mean land-surface response. Local conflict density explained limited variation in the human response (adjusted R2 = 0.034), whereas adding neighbouring exposure increased the land-response model’s adjusted R2 from 0.009 to 0.527. Residual spatial autocorrelation nevertheless remained high (Moran’s I = 0.727). These descriptive results indicate that night-time darkening and optical land-surface change capture distinct dimensions of wartime change rather than interchangeable indicators of conflict impact.