Climate risk gains two counters beyond rainfall and temperature
Atmospheric water demand changes the Amazon drought account, while building age and compactness reshape urban exposure. The studies show how indicator choice redistributes measured climate vulnerability.
Science··Morning
Atmospheric demand for water in the Amazon
A study ranking Amazon droughts from 1980 to 2024 found that the 2023–2024 event affected 88% of the basin and reached four times the average magnitude of the five largest droughts. Rainfall shortage was the main component in major events before 2005; after 2005, warming-enhanced atmospheric evaporative demand became the dominant intensifier. The measure adds how much water hotter air draws from soil and vegetation to the account of water that did not fall as rain.[1]
Building-stock proxies in cities
The second study combines building footprints and Earth observations in 100 × 100 km tiles across more than 9,000 urban centres. Some 43% of buildings predate 1980. Using construction period and building compactness as proxies for local heating and cooling needs, the model places about 48 million people today in areas where vulnerable buildings concentrate. Under the selected scenarios the number grows 10–30% by 2100, shifting towards the Sahel, Gulf countries and Pakistan.[2]
What the proxy makes visible
Neither directly reads the target. The Amazon analysis derives atmospheric demand from rainfall and temperature; the urban model uses construction period and form. Those proxies expose a blind spot: the same rainfall deficit can differ in hotter air, and the same outdoor temperature across building stocks. The limits follow from that choice. Renovation, HVAC efficiency, occupant behaviour and local adaptation can alter the urban account; drought-index choice and ocean-atmosphere links can alter the Amazon account. A risk map is only as reliable as its indicator definition.[1], [2]
Related columns
For more information on this topic, you can read the related columns.
