Rapidly growing medium-sized cities in the tropics may exhibit complex spatial patterns of combustion-related air pollutants while remaining poorly represented by conventional ground-based monitoring networks. This study characterizes nitrogen dioxide (NO₂), sulfur dioxide (SO₂), and carbon monoxide (CO) over Malang City, East Java, Indonesia, using Sentinel-5P TROPOMI observations for selected periods in 2021 and 2025. Tropospheric column densities were processed in Google Earth Engine and converted to estimated ambient concentrations using a mixing-height formulation. The estimates were classified according to the Indonesian Air Pollution Standard Index and compared with available ministry in situ observations and Copernicus Atmosphere Monitoring Service (CAMS) reanalysis using regression, root mean square error, and mean absolute error. City-mean concentrations were lower in the analysed 2025 periods than in 2021 for NO₂ (5.36 to 4.60 μg/m3) and CO (1.571 to 1.465 μg/m3). Because the two years were not sampled in the same months, each difference was recomputed across every possible month subset. Only the CO difference proved robust (−6.7%, envelope −7.3% to −6.1%); the NO₂ difference was robust in direction but not in magnitude (−24.9% to −3.5%); and the apparent SO₂ decrease reversed to an increase when restricted to the months common to both years (−60% to +34%) and is therefore not reported as a difference. Spatial patterns differed among pollutants: NO₂ maxima occurred along northern and western built-up corridors, SO₂ was higher near the eastern and southern industrial periphery, whereas CO was relatively homogeneous, with a coefficient of variation below 2%. Agreement with in situ observations was weak for NO₂ and absent for SO₂, with no satellite–reference correlation reaching statistical significance (p ≥ 0.09) and every confidence interval on r spanning zero, indicating that satellite-derived concentrations are more appropriately interpreted as spatial screening estimates. The principal contribution is a multi-pollutant characterization of Malang as a tropical highland city with limited ground-based monitoring, revealing contrasting spatial signatures of NO₂, SO₂, and CO. These findings demonstrate the potential of satellite observations to complement sparse ground-based monitoring in data-limited tropical urban environments.