No need for convert ISO 8859-1 characters to entities on UTF-8 output.
See http://api.wunderground.com/api/YOURA... for an example.
It returns รค as & auml; and it is impossible for TTS readers to read.
...
Keijo Mukku, 8 years ago