I recently constructed a stacked bowl type of radiation shield to experiment with and found some interesting results in the tests.
The construction and materials are pretty much the usual or standard stuff. However, I wanted to make sure that the bowls overlapped enough so no outside area was visible confirmed by looking down through with the top bowl off. The top and the bottom bowls are the only two not cut out.
The interesting results now. 1st chart shows 3 days temp readings of a standard WS-2310 with it's "hat" on located in the shade with only the ground below lit by the sun. The sun keeps the ground lit until around noon where it is then blocked by my house for the rest of the day. The peaks on the charts show the sunshine until the house blocks it. The 2nd chart is two days with the "hat" removed and the shield on the sensor.
Measurements in the same location and all clear sunny days. The sensor "sees" the sun's radiation on the ground almost like a photocell.
Oh, total cost of the bowl shield came to $2.80. I like it better than the WS-2310's supplied weather "hat".
I hope this helps anyone thinking of building one in the future.
Dan
edit: Since the time of this project I've learned that this type of thermo/hygro sensor uses a thermistor type temp sensor which is sensitive to sunlight. Most wx stations today use a temp/hygro combination chip that is not sensitive to sunlight so this isn't an issue.