This dissertation makes three scientific contributions to the understanding of how human populations react to environmental and epidemiological shocks. The first chapter investigates the impact of tropical storms on migration in the United States. It finds that migration rates remain stable after most tropical storms, with large changes in migration and population losses or gains confined to few county-years. Individuals leaving affected counties after tropical storms tend to relocate to counties at a high risk of experiencing future natural disasters while individuals moving to counties affected by tropical storms partly come from counties at low risk of experiencing tropical storms. Overall, there is thus no evidence that migration triggered by tropical storms reduces the risk of those moving. Damage was found to increase the number of excess out-migrants. Large metropolitan counties were also found to have higher excess in- and out-migration. The second chapter explores the impact of the COVID-19 pandemic on mortality from March 2020 and February 2022. As the first paper to produce county-level estimates of excess mortality, it finds a total of 1,179,024 excess deaths. The paper additionally finds that excess mortality decreased in large metropolitan counties from the first year to the second one but increased in nonmetropolitan counties. Despite the initial concentration of mortality in large metropolitan Northeastern counties, nonmetropolitan Southern counties had the highest cumulative relative excess mortality by February 2021. The third chapter explores the relationship between temperature and mortality and its relationship with seasonal mortality variation in California counties. Assessing the impact of different methodological choices, it finds that models not adjusting for seasonality likely overestimate the negative effect of cold temperatures. Additionally, it shows that when using common but unrealistic counterfactual scenarios researchers might attribute an excessive number of deaths to temperature. Overall, the paper concludes that when using appropriate models and counterfactuals the burden of temperature-related mortality is small though likely to increase because of global warming.