Consider a train horn mounted horizontally at an elevation of 14 feet
above ground level. A thin band of fog exists from 4 feet to 8 feet
above ground. Air is still. Fog has occurred because of radiational
cooling and solar insolation as sun comes up and begins to warm the
ground, causing microenvironmental evaporation at the surface. Warming
air rises with attendant adiabatic cooling. A thin band of fog forms
which will rise in the next minutes.

A train horn sounds above that band of fog.

Is there a thermocline at the fog interface? Will that thermocline
operate in a fashion similar to how one operates in water where a
submarine can hide from sonar below the thermocline?

Does total internal reflection occur so that all (most) of the sound
bounces off the thermocline layer and goes up into the air so that a
person residing and listening below the layer will hear no (almost no)
horn sound until the angle exceeds that where TIR occurs?