The COVID-19 pandemic has upended daily life around the globe, posing a threat to public health. Intuitively, we expect that surging cases and deaths would lead to fear, distress and other negative emotions. However, using state-of-the-art methods to measure emotions and moral concerns in social media messages posted in the early stage of the pandemic, we see a counter-intuitive rise in some positive affect. In addition, we measure changes in emotions and moral concerns during notable events, such as the first US death and state-ordered lockdowns, and find that negativity shows a short-term increase and then a counter-intuitive decrease after COVID-19 takes hold in the US. We hypothesize that the long-term increase of positivity, as well as amelioration of short-term negative reactions, are associated with hedonic adaptation, emotion regulation, and a decrease of uncertainty. Finally, we identify a partisan divide in affect that emerged after the first US death, which resulted in differences in moral and emotional reactions of liberals and conservatives and among US regions. Overall, these results show how collective emotional states have changed since the pandemic began, and how social media can provide a useful tool to understand, and even regulate, diverse patterns underlying human affect.