In order to do that one would have to take the time to research, for every single train number, as to the cities along the route of THAT train, and then build another bot to pull the weather along THOSE cities only, in the correct order, and to be really useful, since LD trains can take several days to travel a route, you would have to go back in time and pull historical weather data, for the correct period of time, for each of those cities. That would be a horrendous amount of work, it seems to me, sitting down and reading the route of each train in the system, from the timetables on the web site (a bunch of PDF files) , and even then you can't know from the delay data whether meteorological conditions actually had anything to do with the delays for that particular train, or if it was a mechanical problem or if it was caused by being stuck behind a freight that hit a vehicle at a grade crossing, or if there was a medical emergency with a passenger and because of that the train had an extended stop while emergency medical assistance was called, or maybe that train or the freight in front of it had a crew that expired..... The possibilities are endless.