Humid heat metrics in Google Earth Engine

“It’s not the heat, it’s the humidity.”

Temperature as commonly reported in your favorite weather app or on your local news station is usually the “dry bulb” variety. While dry-bulb temperature is easier to measure (just use a standard a thermometer in the shade), it omits the effects of humidity, insolation, and wind speed, which are also important for thermal comfort. By contrast, wet-bulb temperature, which accounts for some of these factors and thus may better approximate our actual experience of humid heat, is harder to measure. Other metrics—like apparent temperature, humidex, wet bulb globe temperature, and dewpoint temperature—are proposed as alternate solutions. But none of these metrics is particularly straightforward to calculate. Partially as a result of this difficulty, humid heat metrics are not widely used in studies of the effect of weather on social or economic outcomes. I am hoping to lower the barrier to entry.

Below, I have a link to some Google Earth Engine code that takes as an input hourly ERA5-Land reanalysis data and can output properly aggregated nonlinear functions (see Hsiang 2016) of a set of daily humid heat metrics for geographic units of interest as a CSV straight into your Google Drive. This is very similar to the code I used recently in a paper on wet-bulb temperature and mortality in Mexico (see here). And it’s pretty easy to use. If you discover any bugs, please let me know!

Code here.