Failure and Redemption of Multifilter Rotating Shadowband Radiometer (MFRSR)/Normal Incidence Multifilter Radiometer (NIMFR) Cloud Screening: Contrasting Algorithm Performance at Atmospheric Radiation Measurement (ARM) North Slope of Alaska (NSA) and Southern Great Plains (SGP) Sites

Abstract

Well-known cloud-screening algorithms, which are designed to remove cloud-contaminated aerosol optical depths (AOD) from Multifilter Rotating Shadowband Radiometer (MFRSR) and Normal Incidence Multifilter Radiometer (NIMFR) measurements, have exhibited excellent performance at many middle-to-low latitude sites around world. However, they may occasionally fail under challenging observational conditions, such as when the sun is low (near the horizon) and when optically thin clouds with small spatial inhomogeneity occur. Such conditions have been observed quite frequently at the high-latitude Atmospheric Radiation Measurement (ARM) North Slope of Alaska (NSA) sites. A slightly modified cloud-screening version of the standard algorithm is proposed here with a focus on the ARM-supported MFRSR and NIMFR data. The modified version uses approximately the same techniques as the standard algorithm, but it additionally examines the magnitude of the slant-path line of sight transmittance and eliminates points when the observed magnitude is below a specified threshold. Substantial improvement of the multi-year (1999-2012) aerosol product (AOD and its Angstrom exponent) is shown for the NSA sites when the modified version is applied. Moreover, this version reproduces the AOD product at the ARM Southern Great Plains (SGP) site, which OPEN ACCESSAtmosphere 2013, 4 300 was originally generated by the standard cloud-screening algorithms. The proposed minor modification is easy to implement and its application to existing and future cloud-screening algorithms can be particularly beneficial for challenging observational conditions.

The thin border between cloud and aerosol: Sensitivity of several ground based observation techniques

“…With all this in mind, we find that the variability of the optical depth (which is intended to reflect the variability in the atmosphere, although we cannot rule out some contribution from the instrument's noise when the signal is very low) is not sufficient to discriminate between cloud and aerosol. The optical depth itself and its wavelength dependence (AE) do not help very much in this distinction (despite that Kassianov et al 2013 suggested a method that worked well for their Arctic site). Figure 8 shows the histograms of the populations of points screened out as clouds (this is for Girona, so 88,000 points), points that passed the strict filter so they are considered aerosol (39,000), and points that belong to the transitional zone (81,000), as a function of OD and AE.…”

“…For example, the difficulties in trying to distinguish clouds and aerosol in sunshine duration records have been pointed out elsewhere (Sanchez-Romero et al, 2014). In addition, many works focus on removing cloud -contamination‖ from aerosol observations performed with sunphotometers or shadowband radiometers (Alexandrov et al, 2004;Kassianov et al, 2013;Michalsky et al, 2010). The problem further expands when considering other views (satellite) or other wavelengths (ceilometers in the infrared, microwave radiometers, weather radars).…”

Aerosol Optical Depth Value-Added Product for the SAS-He Instrument

“…The parameter is adjusted to be conservative; that is, it tends to identify some AODs as being contaminated, when in fact they are not, thereby embracing the philosophy that it is better to error on the side of removing a few good AODs rather than letting a significant number of cloud contaminated AODs slip through. Threshold values used in the VAP were derived following the procedures described in Kassianov et al (2013). Figure 9 shows the cloud screen applied to the AODs.…”

“…The blue dots in this are the AODs that have been screened, and deemed acceptable, whereas the red dots show "AODs" that are likely to be cloud contaminated. For additional cloud screening considerations, see Kassianov et al (2013).…”