No abstract

Abstract. A field campaign focused primarily on free tropospheric aerosol measurements over Mildura, Australia, at 34øS (Mildura Aerosol Tropospheric Experiment (MATE 98)) was conducted in the austral spring of 1998 to test for the current presence of a seasonal aerosol layering activity observed in the 1970s and to obtain additional characteristics that would lead to a better understanding of the phenomenon. Ground-based lidar as well as balloon-borne optical particle counters and backscattersondes with ozone sensors were employed. The results indicate that large horizontal scale layers are present and show that their structure is highly correlated with excess ozone. Particle concentrations in the layers were sufficient to measurably affect aerosol optical depth. The probable source is distant biomass burning regions, but a detailed understanding of associated smoke transport and evolution as observed over Mildura is incomplete. At the time of these measurements it was believed that the enhancements were related to biomass burning, but conclusive supporting evidence for this assumption was lacking. For example, since reliable relative humidity measurements were not made, it could be argued that the layers were simply caused by particle growth in layers of higher relative humidity that might be expected with springtime moisture. In addition, for biomass burning aerosols a good correlation with certain other trace species, such as ozone and/or carbon monoxide, might be expected, but this type of supporting information was unavailable. Also, it was unknown if the layers were localized or of large scale, as would be consistent with distant biomass burning sources.Only a two-point slope of the size distribution was measured in the 1970s, but a more complete size distribution was required to assess the physical and optical properties of the aerosol for relating it to the suspected source regions and estimating its radiative impact. For example, it could not be determined if the steep slope of the size distribution indicated that the aerosols were to be associated with the tail of a much higher concentration condensation nuclei mode or rather were part of a relatively narrow accumulation mode. Since the effective particle size for scattering could not be determined reliably from the available size information, it was not clear to us that the layers had sufficient optical cross sections to appreciably affect the radiative properties of the atmosphere. MATE 98In the austral spring of 1998 the Mildura Aerosol Tropospheric Experiment (MATE 98) was conducted for the purpose of reexamining the layers over Mildura in more detail and with the expectation that substantial evidence could be found linking them to biomass burning sources. The initial goal was to observe the presence of the layers and possibly identify changes in activity and/or magnitude. A more sophisticated OPC was employed to obtain extended size distributions for optical model calculations. Some optical properties of the 17,833

Stratospheric aerosol concentrations obtained on seven balloon soundings by in situ photoelectric counting of individual particles and jet impaction with later electron microscope analysis have been compared. The potential error sources in the two sizing techniques appear sufficient to explain the observed differences in concentration. This implies an underestimate in the impactor‐derived precollection radius (for sulfuric acid aerosol) of about 25% relative to the photoelectrically determined radius in the two ranges r > 0.15 μm and r > 0.25 μm. The variance in impactor‐derived column loadings is determined by comparison of the impactor and photoelectric detector loadings; over the 10‐ to 28‐km altitude range, for example, this gives an error estimate of ≈ ± 18% for a 95% confidence level. An estimate of the flight‐to‐flight accuracy of the impactor sizing models is also determined from the loading comparison.

The stratospheric aerosol particle size distribution has been measured at 34°S for the period 1974–1979. Results from two measuring techniques—in situ single‐particle counting and jet impaction—agree well. The size distribution is shown to be described equally well by either a log‐normal or zero‐order logarithmic function. Observed distributions below 22 km agree substantially with model predictions for r>0.1 μm but indicate fewer particles at smaller radii.