Abstract. In the first part of the paper, from a large set of spectrum data, obtained with the UHF/VHF (INSU/METEO) radars during the FRONTS87 and PYREX experiments, statistical characteristics of the noise power spectral density and behavior of its mean are pointed out. For each range gate and various sets of number of coherent and incoherent integrations, histograms of the noise power spectral density are computed using data from different parts of the campaign. The probability density function, so computed, is asymmetrical for low number of incoherent integrations and tends toward a Gaussian shape as the incoherent integration number increases. This histogram behavior is accounted for by a statistical approach, which assumes that the input noise is a random variable normally distributed and has a white spectrum for UHF as well as for VHF. In the second part of this paper, the results obtained by two different methods are compared and are explained by the shape of the probability density function. In the third part of this paper, it is shown that the mean value of the noise power spectral density is a very sensitive parameter for spectrum quality control. Some examples of its behavior as a function of the range gate, the observing direction, and the total power received are given and interpreted; these examples show how the noise mean power density can reveal malfunctioning of the radar, environmental perturbation, or variation due to the sky noise.

Abstract. The capability of Stratosphere-Troposphere radars for estimating the energy dissipation rate oe and eddy diffusivity K z is still strongly questioned because the methods proposed in the literature generally make use of assumptions not yet fully justified. In order to clarify this point, two assessment campaigns, using the high-resolution UHF PROUST radar, have been performed. Results obtained with both "width" and "power" methods are systematically compared with a statistical approach proposed by Dewan [ 1979] and Woodman and Rastogi [1984], using the thickness and lifetime of the turbulent layers. The e and K z derived from these three methods in observed turbulent fields near the tropopause level are found to be reasonably similar for turbulent layers whose thickness LT is smaller than 300 m. For the thicker turbulent patches (LT>300 m), the statistical approach using turbulent thickness tends to give eddy diffusivity values greater than those obtained with width and power methods. This discrepancy could be due to the nonhomogeneity of the turbulence in these large turbulent structures.

No abstract

In this work we investigate the performances of a passive sampler (GABIE badge) in industrial atmospheres, in accordance with the general specifications of the EN 838 standard. The field experiment was carried out in a paint-manufacturing factory producing a large number of pollutants at the workplaces. A comparison was performed between the results obtained by passive sampling and the conventional tube/pump method (reference method) on nine solvents usually encountered in the different workshops: n-butanol, isobutanol, toluene, ethylbenzene, m-xylene, methylisobutylketone, methylethylketone (MEK), ethyl acetate, butyl acetate. Results were compared by use of the distribution of the relative difference between badge "passive sampling" and tube "active sampling" results (with the latter considered as the reference method). In general, results revealed good agreement between passive and active sampling (except in the case of MEK) and confirmed the accuracy of sampling rates determined for the GABIE sampler. Bias was generally low and variability could be considered to be satisfactory (generally < 20% with a maximum of 30% for ethylbenzene). For MEK, strong bias was noted together with probable underestimation of the tube results. Additional results lead us to suggest that this phenomenon could be due to poor desorption of the SKC tubes by carbon disulphide (CS2); (quantitative recovery for MEK is in fact possible using other desorption solvents).

No abstract