^This is correct. I average over an inch of rainfall a week. Also, I'm in "sunny" Florida.
Took a thermometer to test radiant heat. 6" away from the roof, with proper ventilation, there is less than two degrees of difference between ambiant air temp and measured air temp in my builds. At 1' distance, there is no measurable difference - in my biggest build its actually cooler than ambient, and generating its own convection pattern.
That's at 88 degrees. I **ASSUME** It would be much the same at 110, but I'm not in TX anymore, and don't dare consider AZ or NM, where it can be hotter still.
This year when we hit high 90s at similar humidity in advance of a storm front while the skies are still clear, I plan to test it again.
Hopefuly, someone with medium color shingles (tans, greys, or greens) will perform a similar experiment. I know the result in a TX attic - the shingles did NOT fare well - but in spite of all effort, I couldn't get the airflow I wanted.