Confidence in estimates of 20th century climate change is limited by known issues with near-surface air temperature observations from land stations. Station siting, site moves, instrument changes, changing observing practices, urban "heat island" effects, land cover, land use variations, and statistical processing have all been hypothesized as affecting the trends presented by the Intergovernmental Panel on Climate Change and others. The observed rise in global land temperatures since the 1950s is important in the assessment of anthropogenic effects on climate, so any artifacts associated with these issues in the observed decadal and centennial variations could have important consequences for scientific understanding and climate policy. We test those observations using a completely different approach. We have ignored all land temperature observations and instead inferred the temperature from the observed variations of carbon dioxide concentration, solar and volcanic radiative forcing agents, monthly sea surface temperature and sea-ice concentration, and subdaily barometric pressure observations, using a physically-based data assimilation system called the 20th Century Reanalysis (20CR). This independent dataset reproduces both annual variations and centennial trends in the observation-based land station temperature datasets. The patterns and magnitudes of the multi-decadal trends, the annual and even monthly variations of temperature are all similar among the 20CR and station-based datasets, both over the 1901 to 2010 period and the more rapidly- warming 1952 to 2010 period. These similarities indicate that procedures for determining these quantities using the station temperatures are robust and reliable. The comparison is insensitive to the choice of station dataset. That observed global land warming can be reproduced using ocean data and surface pressure observations confirms that the warming is not an artifact of deficiencies in station temperature measurements. The results demonstrate the robustness of previous conclusions using the station temperature datasets to assess global warming.