By the looks of those plots, my first guess is this is the result of binned quality scores produced by machines like the Novaseq or Nextseq. There a few posts on the forum about moving forward with this type of data, maybe this one and it's links can be of help.
all the best 