Modification of a Commercial Gas Filter Correlation CO Detector for Enhanced Sensitivity

View More View Less
  • 1 Department of Meteorology University of Maryland College Park, Maryland
  • | 2 National Center for Atmospheric Research, Boulder, Colorado
© Get Permissions
Full access

Abstract

Carbon monoxide is important in tropospheric chemsitry and useful in studies of cloud dynamics, yet measurements of this gas, especially from aircraft, are too few to characterize fully the atmospheric sources, sinks and distribution of CO. This article describes how a commercial infrared gas filter correlation analyzer (GFC) can be modified to provide sufficient sensitivity and response time for clean air CO measurements. Modifications include improved IR detection, a chemical zero, and sample gas preparation to eliminate interferences effectively, including a minor interference from ozone. The modified instrument demonstrates a detection limit of ∼24 ppb (signal-to-noise ratio 2:1 at the ±1σ noise level, with a 60 s time constant), a broad linear dynamic range, temperature and pressure independence, calibration stability, and a minimum response time of ∼30 s. We present examples of surface and airborne measurements including results from 12 flights over New Mexico showing rapid vertical transport of CO near convective clouds.

Abstract

Carbon monoxide is important in tropospheric chemsitry and useful in studies of cloud dynamics, yet measurements of this gas, especially from aircraft, are too few to characterize fully the atmospheric sources, sinks and distribution of CO. This article describes how a commercial infrared gas filter correlation analyzer (GFC) can be modified to provide sufficient sensitivity and response time for clean air CO measurements. Modifications include improved IR detection, a chemical zero, and sample gas preparation to eliminate interferences effectively, including a minor interference from ozone. The modified instrument demonstrates a detection limit of ∼24 ppb (signal-to-noise ratio 2:1 at the ±1σ noise level, with a 60 s time constant), a broad linear dynamic range, temperature and pressure independence, calibration stability, and a minimum response time of ∼30 s. We present examples of surface and airborne measurements including results from 12 flights over New Mexico showing rapid vertical transport of CO near convective clouds.

Save