Recurring Dip Detection Guide

              
This tool detects recurring downward blips in pCO2 data that occur approximately every 8 hours. These patterns indicate equipment issues that require attention.

              
Detection Method

              
    
                
  • 
                  Identifies local minima
                   in the pCO2 time series using smoothed data
                
    • 
                
  • 
                  Finds dip pairs
                   separated by approximately 8 hours (7-9 hour range)
                
    • 
                
  • 
                  Flags all qualifying occurrences
                   once dips exceed the spike threshold slider
                
    • 
                
  • Pattern indicates equipment malfunction requiring investigation
    • 
              

              
Common Causes

              
    
                
  • Kinked or damaged tubing affecting measurements
    • 
                
  • Clogged intake creating periodic blockages
    • 
                
  • Shared equipment cycling between stations
    • 
                
  • Pump or valve timing issues
    • 
              

              

              
Understanding the Window Scores Tab

              
The Window Scores tab shows detailed analysis results for every 24-hour sliding window analyzed across the selected date range.

              
Column Descriptions:

              
    
                
  • 
                  station:
                   Station identifier (e.g., IRL-VB-WQ, IRL-BR-WQ)
                
    • 
                
  • 
                  start_time:
                   Beginning of the 24-hour analysis window
                
    • 
                
  • 
                  period_hours:
                   Approximate hours between detected dips
                
    • 
                
  • 
                  num_dips:
                   Total number of local minima found in the window
                
    • 
                
  • 
                  dip_pairs:
                   Number of dip pairs separated by ~8 hours
                
    • 
                
  • 
                  amplitude:
                   Maximum variation in pCO2 within the window (µatm)
                
    • 
              

              
How to Use This Data:

              
    
                
  • 
                  Pattern Verification:
                   Review the plots to visually confirm detected dip patterns
                
    • 
                
  • 
                  Timing Analysis:
                   Check the period_hours to confirm ~8-hour spacing
                
    • 
                
  • 
                  Cross-Station Comparison:
                   Look for synchronized patterns across multiple stations
                
    • 
              

              
Note: Windows are overlapping (6-hour step size), so consecutive rows may analyze similar time periods. This overlap improves detection sensitivity.