Abstract. Atmospheric aerosol is a key component of the chemistry and climate of the
Earth's atmosphere. Accurate measurement of the concentration of atmospheric
particles as a function of their size is fundamental to investigations of
particle microphysics, optical characteristics, and chemical processes. We
describe the modification, calibration, and performance of two commercially
available, Ultra-High Sensitivity Aerosol Spectrometers (UHSASs) as used on
the NASA DC-8 aircraft during the Atmospheric Tomography Mission (ATom). To
avoid sample flow issues related to pressure variations during aircraft
altitude changes, we installed a laminar flow meter on each instrument to
measure sample flow directly at the inlet as well as flow controllers to
maintain constant volumetric sheath flows. In addition, we added a compact
thermodenuder operating at 300 ∘C to the inlet line of one of the
instruments. With these modifications, the instruments are capable of making
accurate (ranging from 7 % for Dp < 0.07 µm to 1 %
for Dp > 0.13 µm), precise
(< ±1.2 %), and continuous (1 Hz) measurements of size-resolved particle number
concentration over the diameter range of 0.063–1.0 µm at ambient
pressures of > 1000 to 225 hPa, while simultaneously providing
information on particle volatility. We assessed the effect of uncertainty in the refractive index (n) of ambient
particles that are sized by the UHSAS assuming the refractive index of
ammonium sulfate (n= 1.52). For calibration particles with n between 1.44 and
1.58, the UHSAS diameter varies by +4/−10 % relative to ammonium
sulfate. This diameter uncertainty associated with the range of refractive
indices (i.e., particle composition) translates to aerosol surface area and
volume uncertainties of +8.4/−17.8 and +12.4/−27.5 %,
respectively. In addition to sizing uncertainty, low counting statistics can
lead to uncertainties of < 20 % for aerosol surface area and
< 30 % for volume with 10 s time resolution. The UHSAS reduction
in counting efficiency was corrected for concentrations > 1000 cm−3. Examples of thermodenuded and non-thermodenuded aerosol number and volume
size distributions as well as propagated uncertainties are shown for several
cases encountered during the ATom project. Uncertainties in particle number
concentration were limited by counting statistics, especially in the
tropical upper troposphere where accumulation-mode concentrations were
sometimes < 20 cm−3 (counting rates ∼ 5 Hz) at
standard temperature and pressure.