Showing a historical trend, which is what you demanded proof of, requires data from past years. Pretty simple concept.
If you want to talk about predictions for the future, that's a different argument. But let's be clear about something: a prediction for the future is not factual data. The original point, disputed by you, was that coal use IS declining. "Is", not "will be". Understand?
Regardless, imagine what coal use predictions were in 2005, before we knew about abundant and cheap natural gas? Think that prediction changed? The current landscape could just as easily change, forcing current predictions to need modification. Again, predictions are not facts.