Heat Index: Why 90°F Can Feel Like 105°F

Why this guide

Heat index is the number that answers "how hot does it feel", and it is not a guess: it is a regression fit to a human heat-balance model, published by the US National Weather Service. It matters because it is the number used to issue heat advisories. This guide works the formula, shows a worked example, and explains why the result is undefined in exactly the conditions where people most want it.

Information Gain

The NWS heat index is a polynomial in air temperature T (°F) and relative humidity R (percent):

HI = c_1 + c_2 T + c_3 R + c_4 T R + c_5 T^2 + c_6 R^2 + c_7 T^2 R + c_8 T R^2 + c_9 T^2 R^2

with the coefficients c_1 = -42.379, c_2 = 2.04901523, c_3 = 10.14333127, c_4 = -0.22475541, c_5 = -6.83783\times10^{-3}, c_6 = -5.481717\times10^{-2}, c_7 = 1.22874\times10^{-3}, c_8 = 8.5282\times10^{-4}, c_9 = -1.99\times10^{-6}.

Two adjustments are then applied. If R < 13\% and T is between 80°F and 112°F, subtract:

ADJ_1 = \frac{13 - R}{4}\sqrt{\frac{17 - |T - 95|}{17}}

If R > 85\% and T is between 80°F and 87°F, add:

ADJ_2 = \frac{R - 85}{10} \times \frac{87 - T}{5}

The formula is only defined for T \ge 80\ ^\circ\text{F}. Below that the heat index is reported as the air temperature, because the model has no meaningful deviation from it.

Worked example

At T = 90\ ^\circ\text{F} and R = 65\%:

HI = -42.379 + 2.04901523(90) + 10.14333127(65) - 0.22475541(90)(65) - 6.83783\times10^{-3}(8100) - 5.481717\times10^{-2}(4225) + 1.22874\times10^{-3}(8100)(65) + 8.5282\times10^{-4}(90)(4225) - 1.99\times10^{-6}(8100)(4225)
= -42.379 + 184.411 + 659.317 - 1326.256 - 55.386 - 231.602 + 646.932 + 324.294 - 68.123
= 103.21\ ^\circ\text{F}

Neither adjustment applies (R = 65\% is not below 13% and not above 85%), so the heat index is about 103°F. The 65% humidity makes 90°F air feel like 103°F.

Common mistakes