Observations of very low amounts of precipitable water vapor (PWV) by means of the Ground‐Based Millimeter wave Spectrometer (GBMS) are discussed. Low amounts of column water vapor (between 0.5 and 4 mm) are typical of high mountain sites and polar regions, especially during winter, and are difficult to measure accurately because of the lack of sensitivity of conventional instruments to such low PWV contents. The technique used involves the measurement of atmospheric opacity in the range between 230 and 280 GHz with a spectral resolution of 4 GHz, followed by the conversion to precipitable water vapor using a linear relationship. We present the intercomparison of this data set with simultaneous PWV observations obtained with Vaisala RS92k radiosondes, a Raman lidar, and an IR Fourier transform spectrometer. These sets of measurements were carried out during the primary field campaign of the Earth Cooling by Water vapor Radiation (ECOWAR) project which took place at Breuil‐Cervinia (45.9°N, 7.6°E, elevation 1990 m) and Plateau Rosa (45.9°N, 7.7°E, elevation 3490 m), Italy, from 3 to 16 March 2007. GBMS PWV measurements show a good agreement with the other three data sets exhibiting a mean difference between observations of ≃9%. The considerable number of data points available for the GBMS versus lidar PWV correlation allows an additional analysis which indicates negligible systematic differences between the two data sets.

A novel approach for the analysis of aerosol absorption coefficient measurements is presented. A 7-wavelenghts aethalometer has been employed to identify brown carbon (BrC) and black carbon (BC) and to detect organic carbon (OC) in fine atmospheric aerosols (PM_2.5). The Magee Aethalometer estimates the BC content in atmospheric particulate by measuring the light attenuation in the aerosols accumulated on a quartz filter, at the standard wavelength λ = 0.88 μm. The known Magee algorithm is based on the hypothesis of a mass absorption coefficient inversely proportional to the wavelength. The new algorithm has been developed and applied to the whole spectral range; it verifies the spectral absorption behavior and, thus, it distinguishes between black and brown carbon. Moreover, it allows also to correct the absorption estimation at the UV wavelength commonly used to qualitatively detect the presence of mixed hydrocarbons. The algorithm has been applied to data collected in Agri Valley, located in Southern Italy, where torched crude oil undergoes a pre-treatment process. The Magee Aethalometer has been set to measure Aerosol absorption coefficients τ_aer (λ, t) every 5 min. Wavelength dependence of τ_aer (λ, t) has been analyzed by a best-fit technique and, excluding UV-wavelengths, both the absorption Angstrom coefficient α and the BC (or BrC) concentration have been determined. Finally, daily histograms of α provide information on optical properties of carbonaceous aerosol, while the extrapolation at UV-wavelengths gives information on the presence of semivolatile organic carbon (OC) particles.

The vertical extinction of the aerosol has been measured with a multiwavelength sunphotometer at a rural location in southern Italy. All measurements were performed during a period with relatively stable weather condition, thus it could be expected to have measured always the same air mass at least within each day. The direct solar radiation was recorded at wavelengths between 400 and 1100 nm with a resolution of 0.5 nm. Calibration of the instrument was performed by means of a Langley plot at an elevation of approximately 3000 m. From these measurements, vertical optical depths have been derived for a series of days in June 1993. The spectral extinction coefficient of the atmospheric aerosol at the same location has been measured by means of a telephotometer. The radiance of a target at a distance of 12 km was measured at nine wavelengths in the visible. Thus the extinction coefficient of the aerosol contained within a conical volume of about 450,000 m³ and a length of 12 km has been measured. The daily variation usually followed the same pattern: a decrease in extinction coefficient in the morning and in the early afternoon and an increase toward the evening. This variation correlated well with the change in humidity, so that it can be assumed that the aerosol particles have decreased their size due to evaporation of water in the morning and uptake of water in the evening. The regular daily pattern of the horizontal aerosol extinction coefficient normally was not followed by the vertical aerosol optical density, but on the other hand, changes in vertical extinction have occurred on some days, which had no effect on the horizontal extinction. Vertical extinction measurements were inverted to obtain aerosol size distributions. Inversions were obtained by the Twomey‐Phillips method and they confirm the presence of a hygroscopic aerosol as suggested directly by the variation of the extinction coefficients against the relative humidity. The aerosol undergoes very clear processes of growth which are well correlated with the behavior of the humidity. Overall, the type of aerosol we have measured could be classified as rural aerosol. This kind of aerosol is typical of the boundary layer of clean continental area and is composed prevalently by water‐soluble substances and in minor abundance by dustlike aerosol.

The dependence of Ångstrom's turbidity parameters on the wavelength has been analyzed in the wavelength interval 0.45–0.70 μm. Measurements of aerosol optical thicknesses in this range were obtained by a wide-range grating spectrometer with a resolution of ∼ 2.5 nm. Based on such measurements, two different methods for obtaining turbidity parameters have been validated and intercompared: (a) the ratio method based on processing couples of aerosol optical depths, and (b) a linear best-fitting method based on log-log plot of aerosol optical thicknesses vs. wavelength. We found that Angstr̊m's turbidity relation is quite effective in characterizing atmospheric aerosol in low-medium turbidity cases. However, the usual method of computing the turbidity parameters on the basis of a few measurements (< 10) does not produce adequate results. Accurate estimates can be only obtained using a larger number of data points