Climate change has emerged as a global political issue. Evidence of the devastating impact of human activities on the planet is accumulating, including record heat, increasing number of weather disasters, melting ice, decline of nature. Many scientific studies are based on meteorological data that has become abundant and easily accessible. In this work, we apply a recent statistical tool from risk handling in finance and insurance to environmental data for assessing some aspects of climate change. More specifically, we use the concept of extremiles, which defines an asymmetric least squares analog to quantiles (Daouia et al., 2019, 2022), as an alternative risk measure to the standard Value at Risk. In order to quantify how extreme precipitation and heatwaves have become over time, we perform extremile estimation and inference at different time periods between 1980 and 2023 on two different geographical scales: the satellite data itself besides the country scale. The results show an overall increasing evolution in heavy precipitation and extreme heatwaves. However, the tail exposure differs following the geographical location. The risks of high precipitation are concentrated around the tropics, while the risks of heatwaves are very strong in regions near the poles, in Central Africa, northern South America, Europe, and North America.