The UTF8 encoding of a non-breaking space (U+00A0) is 0xC2 0xA0. If you decode as ISO-8859-1 or CP-1252, that is 0xC2 (Â) and then NBSP (0xA0).

So the content is supposed to have an NBSP between the sentences, was encoded as UTF8, but declared as ISO-8859-1 or similar at some stage in its history. The page seems now to declare UTF8, so it's presumably had the wrong decoding reencoded as UTF8.

Because someone's messed up their encodings.Â