Abstract. In this paper, a novel lidar/radar method for simultaneously determining cloud particle effective size profiles (for water and ice clouds) and the lidar attenuation profile is described. Simulations and application to real data show that this procedure can be quite robust even in cases where significant lidar attenuation is present. In addition, the concept of a suitable lidar/radar effective particle size is introduced, and the determination of water contents for both water and ice clouds using this effective size together with the radar reflectivity profile is discussed. This paper concludes by presenting examples of effective size profiles and water content profiles inferred from measurements made during the Dutch Clouds and Radiation (CLARA) campaign. In a companion paper, it is demonstrated that the results of the inversion procedure compare favorably with infrared radiometer measurements as well as with in-situ measurement results.

Abstract. A new combined iidar/'radar inversion procedure has been de -' .... _, cA__ cloud effective radius and water content retrievals. The algorithm treats the lidar extinction, derived effective particle size, and multiple-scattering effects together in a consistent fashion. This procedure has been applied to data taken during the Netherlands Cloud and Radiation (CLARA) campaign and the Cloud Lidar and Radar Experiment (CLARE'98) multisensor cloud measurement campaign. The results of the algorithm compare well with simultaneous IR radiometer cloud measurements as well as with measurements made by using aircraft-mounted two-dimensional probe particle-sizing instruments.

A newly developed procedure for estimating ice cloud effective particle size using combined lidar and radar signals has been applied to several months of data from the Atmospheric Radiation Measurement Program's Southern Great Plains site. Though considerable scatter is present in the retrieval results, on average, the data show a clear dependence of effective particle size on both temperature and ice water content. The data are used to construct a parameterization based on a bimodal ice particle size distribution whose characteristics depend both on temperature and ice water content.

ABSTRACT. Simultaneous measurements of oscillations of the electron temperature, the electron density and the poloidal magnetic field in the Rijnhuizen Tokamak RTP are combined to obtain insight into the structure of MHD mode perturbations. Diagnostics used are a 20 channel heterodyne ECE radiometer, a 19 channel FIR interferometer, and a set of magnetic pick-up coils. A m/n = 2/1 island preceding a disruption is analysed in detail. Whereas the density and temperature are significantly lower in the O-point than in the X-point, no local minimum or even flattening of the profiles could be demonstrated. The gradients across the island may be understood in terms of a very long connection length inside the island, in combination with the relatively short mean free path along the field lines due to the high collisionality in the outer region of the plasma. The transport inside the island and the poloidal distribution of the radial heat flux over X-and O-point are discussed.

The Thomson scattering spectrum represents the projection of the three-dimensional electron velocity distribution on the scattering vector. From this the local electron temperature and density can be derived. To determine the three-dimensional electron velocity distribution it is necessary to have several viewing directions, assuming axial symmetry in velocity space perpendicular to the magnetic field. Thomson scattering experiments in the TORTUR tokamak demonstrated the first experimental determination of f(v,, vu) from radially and tangentially observed spectra. The latter made it possible to determine the toroidal current density. The scattering spectra, averaged over 30 laser pulses observed in both the radial and the tangential direction, show clearly two symmetrical dips near the top of the Gaussian spectrum. The magnitude of the dips is relatively small, only 7% of the amplitude of the Gaussian. The partial electron density responsible for the non-thermal features was found to scale proportionally to the local density and to exhibit a specific development when the plasma is grossly distorted by fast current pulses or minor disruptions. The individual spectra show a spread in photon yield around the wavelength of these dips which is much larger than can be explained by photon statistics. If a spherically symmetric velocity distribution is assumed, the dips in the averaged spectrum (over 30 pulses) can only be explained by strong distortions of the electron velocity distribution (at least 50%). For the individual spectra, which show even larger distortions, the explanation cannot be found by considering only the velocity space. The observations can possibly be explained by strong spatial anisotropy, for example filamentation of the plasma current.