Automated Time Series Anomaly Detection Pipeline

Design a Python script using pandas and scikit-learn that can automatically detect and flag statistical anomalies in time series data across multiple numeric columns. The solution should include robust preprocessing, z-score and Interquartile Range (IQR) methods, and generate a comprehensive JSON report with anomaly details, timestamps, and statistical significance. Implement error handling for edge cases and ensure the script can process large datasets efficiently with minimal computational overhead.