Abstract. Long-range transport of black carbon (BC) is a growing concern as a result of the efficiency of BC in warming the climate and its adverse impact on human health. We study transpacific transport of BC during HIPPO-3 using a combination of inverse modeling and sensitivity analysis. We use the GEOS-Chem chemical transport model and its adjoint to constrain Asian BC emissions and estimate the source of BC over the North Pacific. We find that different sources of BC dominate the transport to the North Pacific during the southbound (29 March 2010) and northbound (13 April 2010) measurements in HIPPO-3. While biomass burning in Southeast Asia (SE) contributes about 60 % of BC in March, more than 90 % of BC comes from fossil fuel and biofuel combustion in East Asia (EA) during the April mission. GEOS-Chem simulations generally resolve the spatial and temporal variation of BC concentrations over the North Pacific, but are unable to reproduce the low and high tails of the observed BC distribution. We find that the optimized BC emissions derived from inverse modeling fail to improve model simulations significantly. This failure indicates that uncertainties in BC removal as well as transport, rather than in emissions, account for the major biases in GEOS-Chem simulations of BC over the North Pacific.The aging process, transforming BC from hydrophobic into hydrophilic form, is one of the key factors controlling wet scavenging and remote concentrations of BC. Sensitivity tests on BC aging (ignoring uncertainties of other factors controlling BC long range transport) suggest that in order to fit HIPPO-3 observations, the aging timescale of anthropogenic BC from EA may be several hours (faster than assumed in most global models), while the aging process of biomass burning BC from SE may occur much slower, with a timescale of a few days. To evaluate the effects of BC aging and wet deposition on transpacific transport of BC, we develop an idealized model of BC transport. We find that the mid-latitude air masses sampled during HIPPO-3 may have experienced a series of precipitation events, particularly near the EA and SE source region. Transpacific transport of BC is sensitive to BC aging when the aging rate is fast; this sensitivity peaks when the aging timescale is in the range of 1-1.5 d. Our findings indicate that BC aging close to the source must be simulated accurately at a process level in order to simulate better the global abundance and climate forcing of BC.

No abstract

Secondary organic aerosols (SOA) exert a significant influence on ambient air quality and regional climate. Recent field, laboratorial and modeling studies have confirmed that in-cloud processes contribute to a large fraction of SOA production with large space-time heterogeneity. This study evaluates the key factors that govern the production of cloud-process SOA (SOA_cld) on a global scale based on the GFDL coupled chemistry-climate model AM3 in which full cloud chemistry is employed. The association between SOA_cld production rate and six factors (i.e., liquid water content (LWC), total carbon chemical loss rate (TC_loss), temperature, VOC/NO_x, OH, and O₃) is examined. We find that LWC alone determines the spatial pattern of SOA_cld production, particularly over the tropical, subtropical and temperate forest regions, and is strongly correlated with SOA_cld production. TC_loss ranks the second and mainly represents the seasonal variability of vegetation growth. Other individual factors are essentially uncorrelated spatiotemporally to SOA_cld production. We find that the rate of SOA_cld production is simultaneously determined by both LWC and TC_loss, but responds linearly to LWC and nonlinearly (or concavely) to TC_loss. A parameterization based on LWC and TC_loss can capture well the spatial and temporal variability of the process-based SOA_cld formation (<i>R</i>² = 0.5) and can be easily applied to global three dimensional models to represent the SOA production from cloud processes

No abstract

Secondary organic aerosols (SOA) exert a significant influence on ambient air quality and regional climate. Recent field, laboratorial and modeling studies have confirmed that in-cloud processes contribute to a large fraction of SOA production. This study evaluates the key factors that govern the production of cloud-process SOA (SOA_cld) in a global scale based on the GFDL coupled chemistry-climate model AM3 in which full cloud chemistry is employed. The association between SOA_cld production rate and six factors (i.e. liquid water content (LWC), total carbon chemical loss rate (TC_loss), temperature, VOC/NO_x, OH, and O₃) is examined. We find that LWC alone determines the spatial pattern of SOA_cld production, particularly over the tropical, subtropical and temperate forest regions, and is strongly correlated with SOA_cld production. TC_loss ranks the second and mainly represents the seasonal variability of vegetation growth. Other individual factors are essentially uncorrelated to SOA_cld production. We find that the rate of SOA_cld production is simultaneously determined by both LWC and TC_loss, but responds linearly to LWC and nonlinearly (or concavely) to TC_loss. A parameterization based on LWC and TC_loss can capture well the spatial and temporal variability of the process-based SOA_cld formation (<i>R</i>²=0.5) and can be easily applied to global three dimensional models to represent the SOA production from cloud processes