Proposed Method for Data Substitutions 1

1
Proposed Method for Data Substitutions (1)

• Shining Rock (SHRO) and Breton (BRET) monitoring
  sites did not meet the Regional Haze Rule data
  capture requirements for 2002
• Data from the Great Smoky Mountains (GRSM)
  IMPROVE site were used for SHRO, and data from
  the Gulfport (GFP) SEARCH site were used for
  BRET. (SEARCH and IMPROVE methodologies, while
  not identical, are reasonably similar.)
• Data were scaled based on long-term comparisons
  among species between pairs of sites

2
Proposed Method for Data Substitutions (2)

• In most cases the R2 values associated with the
  long-term comparisons show weak to moderate
  correlations
• Therefore, the daily ratio of species mass
  between site pairs was calculated and averaged
  for the entire period
• For example, the average daily ratio for ammonium
  sulfate between SHRO and GRSM was determined to
  be 0.74 therefore, ammonium sulfate for a
  missing sample day at SHRO was filled in with the
  same days ammonium sulfate value from GRSM
  multiplied by 0.74

3
Aerosol Mass Comparisons SHRO vs. GRSM (1994
2004)
4
Aerosol Mass Comparisons BRET vs. GFP (2000
2004)
5
Summary of Site Comparisons
6
Potential Refinements to Method

• Review seasonal statistics for site pairs
• Review other site pairings to see if results are
  reasonably consistent
• SHRO vs. LIGO or COHU (both have short data
  record)
• BRET vs. SIKE (central Louisiana also short
  data record)
• Review back trajectory or model information to
  help decide how much emphasis to give site
  pairings