One reason that the media are reluctant to cover it is that they do not know what is normal in science. I remember data crunches in speech recognition research

One reason that the media are reluctant to cover it is that they do not know what is normal in science. I remember data crunches in speech recognition research in the late 80's. We would run out of disk space and have to delete stuff. Usually it would be the value added data. If we wanted it back we would re-run the processing scripts to recreate it from the raw data. Sometimes the raw data would be archived on tape and deleted from the disks. If you needed to regenerate the value added data you had to get one of the system administrators to retrieve the raw data from the tape archive before you could start. Two take home messages. One: we never lost raw data. Two: we never lost the scripts for applying corrections and were always in the position of being able to replicate our work. Shortage of disk space meant that we sometimes had to do this, so it had to work. I suspect that journalists are slow to realise that deleting the raw data is a very odd thing to do. You delete your data when you have no students, have lost faith in your research and are committed to giving up and leaving the field. Do you delete your raw data if you are battling to save the planet? No.