High Backup Storage Usage: Request for Detailed Breakdown

My Azure Database for PostgreSQL Flexible Server bill shows a significant increase in "Backup Storage LRS Data Stored" costs — from $255.72 in June to $580.90 in July.

• During this time I did not modify the provisioned storage capacity (512GiB, P20 = $58.88/month) nor backup retention period (35 days, GRS). In reality actual storage used is around 214GB.
• On July 23, I did upgrade the VM from D2ds_v4 to D4ds_v4. But the backup usage metrics doesn't show a significant increase corresponding to that (see attached chart):

My first question is how the GRS pricing is applied.

• Is this Usage plot before doubling for GRS? If so 35 days * 214GB actual storage used = 7.5TB of snapshots, so it means the backups contain significant bloat. Is that all from WAL continuous archiving?
• Or if this Usage Plot already includes double-counting of LRS data to achieve global redundancy, then... I'd actually expect the value to be much higher, unless the backups are significantly compressed. How is this the case?

Can you tell me the exact split between snapshot storage vs WAL storage? And provide insight into whether my WAL volume is typical, and whether there are specific time periods where WAL spiked unusually? If not typical then I need to understand what's driving WAL size.

It is reasonable for me to reduce retention duration down to 14 days and I'll do that after an answer from you. Beyond that, are there any other changes you recommend to optimize backup costs?