What controls humidity on hot days in the tropics?

Hot days on tropical land warm faster than the average day, and evidence is growing that this enhanced warming is due to hot days being dry. Mechanisms rooted in tropical atmospheric dynamics and thermodynamics explain how low near-surface air humidity tends to give rise to high surface temperatures. But what makes the air dry in the first place and how the underlying processes may be affected by a changing climate remain open questions, impeding reliable predictions of tropical heat extremes in a warming world.

Here we present a composite moisture budget analysis for hot days on tropical land based on 45 years of daily ERA5 reanalysis data. Using a statistical approach and sampling the hottest day per location and year, we investigate the contributions of horizontal and vertical moisture transports, evapotranspiration, and precipitation to the change in specific humidity during the run-up to hot days.

Preliminary findings based on the vertically-integrated moisture budget indicate that moisture anomalies develop over a characteristic time scale of three to four days prior to the hot day. However, counter to our initial expectations, these anomalies are not generally negative. Only in about half of all cases, the atmosphere undergoes net drying leading up to a hot day. We find that the sign of the moisture anomaly correlates with the background climate as measured by the aridity index: In moist regions, hot days tend to be dry while in arid regions, hot days tend to be moist. In both cases, the anomaly is explained by a regime shift of the horizontal moisture transport, from convergence to divergence in the case of dry anomalies, and from divergence to convergence in the case of moist anomalies.

We discuss the implications of these results for understanding humidity on tropical hot days in a warmer climate.