Add 3D variable plotting support for multi-dimensional datasets in report generation

[eleanorfrajka]

The report generation system currently only supports plotting of 1D and 2D timeseries. Several datasets contain multi-dimensional variables that would benefit from 3D visualization in their reports:

MOVE 16°N datasets:

• VCUR variables with shape (time, latitude) - should plot first latitude index with squeeze
• Multi-dimensional transport components

CALAFAT2025 datasets:

• MHT variables with 3+ dimensions - should average over 3rd dimension for plotting
• Spatial heat transport fields

Arctic Gateway datasets:

• Multi-dimensional transport arrays across different passages

Requirements:

• Extend amocatlas/plotters.py with functions to handle multi-dimensional variables
• Update amocatlas/report.py to detect and appropriately plot 3D variables
• Add dimension reduction logic (squeeze, average, select indices) based on variable type
• Ensure generated plots are informative and properly labeled
• Maintain backward compatibility with existing 1D plotting

Acceptance Criteria:

• Reports for MOVE, CALAFAT2025, and Arctic Gateway include meaningful plots of their multi-dimensional variables
• Plot generation handles dimension mismatches gracefully
• Generated plots have clear titles indicating any dimension reduction applied