Extended Data Fig. 1: Changes in hot, dry days for JRA-55, ERA5, ISD1979, and ERA5 at locations of Integrated Surface Database stations.
From: Hot extremes have become drier in the United States Southwest
As in Fig. 2, but with data from (a, b) JRA-55, (c, d) ERA5, (e, f) Integrated Surface Database (ISD) stations from 1979–2019, and (g, h) ERA5 subset to the locations of the ISD stations in (e). Trends are calculated over the 1958–2019 period for JRA-55, and over the 1979–2019 period for all other maps. The ERA5-based trends in panel (d) uniquely do not show the amplification of drying at hot temperatures. To test whether the difference is due to the different starting point of the ERA5 trends, we recalculate trends in the ISD stations for the 1979–2019 period (panels e and f), which still show the amplification. To test whether the difference is due to the specific locations of the ISD stations, we subset the ERA5 trends to the locations of the ISD stations (panels g and h); this subset does not show the amplification behaviour. In the line plots (panels b, d, f, h), thin grey lines show the trends at each gridbox, and the thick black line is the area-weighted average across gridboxes. Note the different y-scales across the line plots.
