Comparing the output of general circulation models to observations is essential for assessing and improving the quality of models. While numerical weather prediction models are routinely assessed against a large array of observations, comparing climate models and observations usually requires long time series to build robust statistics. Here, we show that by nudging the large-scale atmospheric circulation in coupled climate models, model output can be compared to local observations for individual days. We illustrate this for three climate models during a period in April 2020 when a warm air intrusion reached the MOSAiC (Multidisciplinary drifting Observatory for the Study of Arctic Climate) expedition in the central Arctic. Radiosondes, cloud remote sensing and surface flux observations from the MOSAiC expedition serve as reference observations. The climate models AWI-CM1/ECHAM and AWI-CM3/IFS miss the diurnal cycle of surface temperature in spring, likely because both models assume the snowpack on ice to have a uniform temperature. CAM6, a model that uses three layers to represent snow temperature, represents the diurnal cycle more realistically. During a cold and dry period with pervasive thin mixed-phase clouds, AWI-CM1/ECHAM only produces partial cloud cover and overestimates downwelling shortwave radiation at the surface. AWI-CM3/IFS produces a closed cloud cover but misses cloud liquid water. Our results show that nudging the large-scale circulation to the observed state allows a meaningful comparison of climate model output even to short-term observational campaigns. We suggest that nudging can simplify and accelerate the pathway from observations to climate model improvements and substantially extends the range of observations suitable for model evaluation.