On the use of carbon monoxide and aerosol optical depth retrievals in constraining the black carbon distribution

Understanding the distribution of black carbon (BC) is important for air quality, weather and climate research. However, our understanding of its distribution is hindered at present by limited BC measurements. Here, we exploit the availability of carbon monoxide (CO) and aerosol optical depth (AOD) observations from space to provide constraints on the model simulations of black carbon. Studies have shown that there is a strong correlation between these observations and the concentrations of BC observed in the atmosphere. Our approach is to simulate this correlation through an ensemble of model simulations and use this to correct simulated BC given CO and/or AOD retrievals. We use an ensemble-based chemical data assimilation system that includes the Community Atmosphere Model with Chemistry (CAM-Chem) and the Data Assimilation Research Testbed (DART). This system mimics a numerical weather prediction (NWP) system with chemistry. Currently, it assimilates conventional meteorological observations, CO retrievals from the Measurement of Pollution In The Troposphere (MOPITT) instrument and AOD retrievals from the Moderate Resolution Imaging Spectroradiometer (MODIS) instrument. We demonstrate the performance of this approach through model experiments with and without the BC corrections. Our results show significant improvements in the BC profiles and surface concentrations, based on comparisons with field campaign and surface measurements. We find that assimilating MOPITT and MODIS provides considerable impact on simulated BC concentrations, especially over the source regions. This approach offers an opportunity to augment our current ability to predict the distribution of BC.