Flow-induced errors in airborne in-situ measurements of aerosols and clouds

Abstract

Abstract. Aerosols and clouds affect atmospheric radiative processes and climate in many complex ways and still pose the largest uncertainty in current estimates of the Earth’s changing energy budget. Airborne in-situ sensors such as the Cloud, Aerosol, and Precipitation Spectrometer (CAPS) or other optical spectrometers and optical array probes provide detailed information about the horizontal and vertical distribution of aerosol and cloud properties. However, flow distortions occurring at the location where these instruments are mounted on the outside of an aircraft may directly produce artifacts in detected particle number concentration and also cause droplet deformation and/or break-up during the measurement process. Several studies have investigated flow-induced errors assuming that air is incompressible. However, for fast-flying aircraft, the impact of air compressibility is no longer negligible. In this study, we combine airborne data with numerical simulations to investigate the flow around wing-mounted instruments and the induced errors for different realistic flight conditions. A correction scheme for deriving particle number concentrations from in-situ aerosol and cloud probes is proposed, and a new formula is provided for deriving the droplet volume from images taken by optical array probes, reducing errors by up to one order of magnitude. Shape distortions of liquid droplets can either be caused by errors in the speed with which the images are recorded or by aerodynamic forces acting at the droplet surface caused by changes in the airflow around the instrument. These forces can lead to the dynamic breakup of droplets causing artifacts in particle number concentration and size. Furthermore, an estimation of the critical breakup diameter, as a function of flight conditions is provided. Experimental data show that flow speed at the instrument location is smaller than the ambient flow speed. Our simulations confirm the observed difference and reveal a size-dependent impact on particle speed and concentration. This leads, on average, to a 25 % overestimation of the number concentration of particles larger than ~10 μm} diameter and causes distorted images of droplets and ice crystals if the flow values recorded at the instrument are used. With the proposed correction scheme both errors are significantly reduced by a factor 10. Although the presented correction scheme is derived for the DLR Falcon research aircraft (SALTRACE campaign) and validated for the DLR Falcon (A-LIFE campaign) and the NASA DC-8 (ATom campaign), the general conclusions hold for any fast-flying research airplane.

CCN concentration and INP-relevant aerosol profiles in the Saharan Air Layer over Barbados from polarization lidar and airborne in situ measurements

“…An excellent agreement is obtained for the coarse mode. This indicates that the Falcon measurements capture well the large particles in the SAL (Spanu et al, 2019). The coarse mode mass concentration from POLIPHON is around 16 times higher than the fine mode mass concentration leading to a mass fine mode fraction of 0.06.…”

“…(iii) Horizontal and temporal inhomogeneities in the dust concentration along the flight tracks and over the lidar site may have also contributed to the found differences. (iv) Although the Falcon data are corrected for the particles losses at the inlets (Spanu et al, 2019), there are several uncertainty sources in the in situ CCN retrieval, that may have contributed to the found bias.…”

Profiles of cloud condensation nuclei, dust mass concentration, and ice-nucleating-particle-relevant aerosol properties in the Saharan Air Layer over Barbados from polarization lidar and airborne in situ measurements

“…Methods have been developed to retrieve CCN-and INP-relevant particle microphysical properties from particle extinction coefficients measured with lidar (Mamouri and Ansmann, 2016;Sawamura et al, 2017;Lv et al, 2018). In the case of INP profiling, particle extinction coefficients are converted to particle number concentrations n 250 (particles with dry radius > 250 nm) and particle surface-area concentrations s. The n 250 profile is input in the INP parameterization schemes of DeMott et al (2010DeMott et al ( , 2015 and Tobo et al (2013), and s profiles are input in respective INP parameterizations by Niemand et al (2012), Steinke et al (2015), Ullrich et al (2017), McCluskey et al (2018), and Harrison et al (2019). The entire lidar-based INP retrieval procedure is described by Mamouri and Ansmann (2016).…”

Aerosol size distributions during the Atmospheric Tomography (ATom) mission: methods, uncertainties, and data products

“…The CAPS probe measures aerosol and cloud particle size distributions in situ without substantial heating and drying of the sample stream as is the case for all the other aerosol instruments operating within the fuselage. However, pressure at the location of measurement below the wing is up to 100 hPa higher than ambient (Spanu et al, 2019), which implies some heating and potential water loss from particles measured by the CAS. The CAPS data are used here to evaluate the performance of the LAS, to examine the effects of efflorescence and inlet losses on the dry size distributions, and to identify cloudy periods for data filtering.…”

“…The underwing CAPS does not have an active controlled sample flow, but is passively pumped by airflow through the probe. Therefore, its sample flow increases with altitude from ~1 up to ~3.5 vlpm depending on true airspeed (Spanu et al, 2019).…”