Bitflips on HDD's are more common than you think, especially when the data is not read (to allow the disk firmware to detect the degradation and rewrite the data).
Only way to keep safe from this is to use a checksumming filesystem + having more than 1 copy of the data.
Possible setups with a BTRFS filesystem: (probably possible with ZFS or any of the other filesystems with checksumming support)
Option 1: One HDD with two partitions. BTRFS setup as raid1 between the partitions and files are copied here. for history create a snapshot before copying the next round of data.
Option 1 protects against bit-flips and general corruption in the filesystem, but you are required to "scrub" the filesystem every time you do a backup..
Option 2: Two HDD's. BTRFS setup as raid1. The rest is the same as option 1.
Option 2 protects against one backup-device having a hardware failure.
Option 3: Two HDD's. BTRFS setup as raid1. The rest the same as option 1. After each backup you take one of the disks and store it offsite and add a fresh disk to the raid1. Swapping between 3 disks (one local, one in transit and one stored remotly) could be an idea too.
Option 3 protects against hw failure of the backup and at the same time protects against data-loss due to a breakin or a fire.