No abstract

A new three-step idealized-profile method to estimate the mixing height from vertical profiles of ceilometer backscattering coefficient is developed to address the weaknesses found with such estimates that are based on the one-step idealized-profile method. This three-step idealized-profile method fits the backscattering coefficient profile of ceilometer measurements into an idealized scaled vertical profile of three error functions, thus having the potential to determine three aerosol layers (one for the surface layer, one for the mixing height, and one for the artificial layer caused by the weakened signal). This three-step idealized-profile method is tested with ceilometer and radiosounding data collected during the Helsinki Testbed campaign (2 January 2006-13 March 2007). Excluding cases with low aerosol concentration in the boundary layer, cases with clouds present, and cases with precipitation present, the resulting dataset consists of 97 simultaneous backscattering coefficient profiles and radiosoundings. The three-step method is compared with the one-step method and other commonly employed algorithms. A strong correlation (correlation coefficient r 5 0.91) between the mixing heights as determined by the three-step method using ceilometer data and those determined from radiosoundings is an improvement over the same correlation using the one-step method (r 5 0.28), as well as the other algorithms.

No abstract

Abstract.A field measurement campaign was conducted near a major road "Itäväylä" in an urban area in Helsinki in 17-20 February 2003. Aerosol measurements were conducted using a mobile laboratory "Sniffer" at various distances from the road, and at an urban background location. Measurements included particle size distribution in the size range of 7 nm-10 µm (aerodynamic diameter) by the Electrical Low Pressure Impactor (ELPI) and in the size range of 3-50 nm (mobility diameter) by Scanning Mobility Particle Sizer (SMPS), total number concentration of particles larger than 3 nm detected by an ultrafine condensation particle counter (UCPC), temperature, relative humidity, wind speed and direction, driving route of the mobile laboratory, and traffic density on the studied road. In this study, we have compared measured concentration data with the predictions of the road network dispersion model CAR-FMI used in combination with an aerosol process model MONO32. For model comparison purposes, one of the cases was additionally computed using the aerosol process model UHMA, combined with the CAR-FMI model. The vehicular exhaust emissions, and atmospheric dispersion and transformation of fine and ultrafine particles was evaluated within the distance scale of 200 m (corresponding to a time scale of a couple of minutes). We computed the temporal evolution of the number concentrations, size distributions and chemical compositions of various particle size classes. The atmospheric dilution rate of particles is obtained from the roadside dispersion model CAR-FMI. Considering the evoCorrespondence to: M. A. Pohjola (mia.pohjola@fmi.fi) lution of total number concentration, dilution was shown to be the most important process. The influence of coagulation and condensation on the number concentrations of particle size modes was found to be negligible on this distance scale. Condensation was found to affect the evolution of particle diameter in the two smallest particle modes. The assumed value of the concentration of condensable organic vapour of 10 12 molecules cm −3 was shown to be in a disagreement with the measured particle size evolution, while the modelling runs with the concentration of condensable organic vapour of 10 9 -10 10 molecules cm −3 resulted in particle sizes that were closest to the measured values.