Add US-wide data processing pipeline scripts

- build_us_zarr.py: Download and process state-by-state forest data
  - Checkpoint/resume capability for long-running jobs
  - Progress tracking with rich UI
  - Support for priority states (high forest coverage)
  - Optimal chunking (512x512) for cloud streaming

- create_manifest.py: Generate state data manifest
  - Indexes all available state Zarr stores
  - Supports local and cloud URL paths
  - Enables API discovery of state data

- upload_states_to_cloud.py: Upload states to cloud storage
  - Supports both Cloudflare R2 and Backblaze B2
  - Batch upload with progress tracking
  - Updates manifest with public URLs

Author: mihiarc