Do in depth analysis of adjusted fields

[cchwala]

Ideas for detailed analysis of adjusted fields (because long-term hexbin plots and metrics look pretty equal for all methods):

• take into account distance from CML to gauge
• rain-intensity based analysis
• event-based analysis (convective vs stratiform)
• CML density vs error
• Kriging uncertainty vs error

to be extended and discussed...

More ideas from meeting 2025-05-14:

• adjust IDW parameters
• adjust Variogram parameters
  (maybe do that for specific events)

[maxmargraf]

Additional figure for spatial overview of the results:

[cchwala]

Priorities:

1. Do hexbin and metrics for selected period with interesting events (those with high rainfall rates)
2. take into account distance from CML to gauge
3. rain-intensity based analysis

[cchwala]

Discussion based on new restults from @EliaCovi

• Interesting pattern here, but also unclear why worse metrics for raw radar, maybe due to small subset of gauges for very close to CML

• issue with KED due to super high rain rate values in extrapolation regions