At the beginning of the year, my h-index on Google Scholar was 45.
This is not going to be a rant about the vacuity of the h-index, or only indirectly. I don’t spend time working this out; Google Scholar does it for you, so it is right up there at the top of the front page.
Last month it was up to 47. Out of curiosity, I checked my papers with citations in the middle 40s, and indeed, several of them had had a few recent citations.
Today I looked again, and the number was up to 49. But a brief check showed that this was entirely bogus. Somewhere along the line, Google Scholar had dumped a whole lot of papers with authors sharing my surname and initial into my list. I went through deleting them (a somewhat tedious business) until I got bored; this brought the h-index back down to 47.
But on the way through, I was disturbed by the dodginess of the data, even in supposedly genuine citations. Various papers have been split into multiple copies (probably because of small errors by the authors citing them); sometimes having a paper in a journal gets me credited as an author of the entire journal issue; and as to the citations themselves, they are as unreliable as you might expect.
Probably this process can be controlled. If you are about to apply for a job or promotion, how nice it would be if you could artificially inflate your h-index. I know that, at least in the recent past, hiring committees in some subjects (not mathematics, as far as I am aware) are influenced by h-index. But perhaps they have moved on to altmetrics now, and the problem has gone away.