By combining the time-of-flight or LIDAR principle with a Thomson backscatter diagnostic, spatial profiles of the electron temperature and density are measured in a magnetically confined fusion plasma. This technique was realized for the first time on the JET tokamak. A ruby laser (3-J pulse energy, 300-ps pulse duration, 0.5-Hz repetition rate) together with a 700-MHz bandwidth detection and registration system yields a spatial resolution of about 12 cm. A spectrometer with six channels in the wavelength range 400–800 nm gives a dynamic range of the temperature measurements of 0.3–20 keV. The stray light problem in the backscatter geometry is overcome by spectral discrimination and gating of the photomultipliers. A ruby filter in the spectral channel containing the laser wavelength allows calibration of the vignetting along the line of sight by means of Raman scattering, enabling the measurement of density profiles. The low level of background signal due to the short integration time for a single spatial point yields low statistical errors (ΔTe /Te ≊6%, Δne /ne ≊4% at Te =6 keV, ne =3×1019 m−3 ). Goodness-of-fit tests indicate that the systematic errors are within the same limits. The system is described and examples of measurements are given.

SUMMARYThis paper describes a representation of the distribution of sensible and latent heat from the surface through the atmospheric boundary layer which has been formulated for use in a 10-level primitive equation model atmosphere. The transfer process is represented in two parts : (i) the transfer of energy across the Earth's surface into the lowermost 100 mb layer of the model atmosphere; and (ii) the subsequent redistribution of this energy through two or more such layers by small-scale convection. The fluxes of energy across the surface are calculated using empirical ' bulk aerodynamic ' relationships. In land regions consideration of the energy balance at the surface is also necessary, and diurnal variations of radiation are taken into account. The redistribution of energy by small-scale convection is represented by convective adjustments which ensure that a certain neutral lapse rate of temperature is never exceeded. Some results of the incorporation of these effects into the 10-level model are described. INTRODCCTIONCarlson and Ludlam (1965) have described the important role played by a layer of small-scale convection near the Earth's surface in the general scheme of midlatitude tropospheric motion. It is by means of the distribution of energy from the surface through such a layer that the large-scale flow regains both the energy lost by radiation to space and the water lost in precipitation. A n adequate representation of this process is therefore an essential requirement for any numerical model attempting to simulate the long-term global circulation of the atmosphere.Such a representation is perhaps less important when forecasts of one or two days are envisaged, as is the case with the 70-level primitive equation model of the atmosphere described by Bushby and Timpson (1967), Benwell and Timpson (1968), and Benwell and Bushby (1970). The incorporation of a representation of surface energy transfers into this 10-level model is, nevertheless, desirable for two reasons. Firstly, in some situations, as when cold air blows over a much warmer sea, the surface energy transfers can be so rapid (Craddock 1951, Manabe 1957) that they significantly influence the development of the large-scale flow over periods of one or two days. Eveninothersituations, the accuracy of details in the forecasts computed with the model should be improved if surface exchanges are represented. Secondly, the representation of surface energy transfers is necessary in order that the model may be able to predict the associated distributions of convective rainfall.The representation of surface energy exchanges which has been formulated for use in the 10-level model treats the transfer process in two parts : (i) the transfer of energy across the Earth's surface; and (ii) the distribution of this energy through the atmospheric boundary layer. In the sense that the term is used here, the ' atmospheric boundary layer ' ranges in vertical extent from a few metres to several kilometres, according to the varying vertical extent of motions initiat...

An economical explicit integration scheme for numerical weather prediction models is described. A splitting technique is used, in which the horizontal advection terms in the governing equations are integrated with a timestep limited by the wind speed, whilst the terms which describe gravity‐inertia oscillations are integrated in a succession of shorter steps. A two‐level numerical scheme with small phase speed errors is used for the advection stage and a forward‐backward method for the gravity‐inertia terms. The split explicit scheme has been applied to the Meteorological Office operational 10‐level model, and to a similar sigma coordinate model, to compute forecasts for periods up to six days ahead. The quality of the numerical forecasts obtained is not reduced when the timesteps used are close to the limits set by linear computational stability criteria. The use of such timesteps leads to a substantial computational economy relative to previously available integration schemes. For the northern hemisphere version of the 10‐level model the computing time required is one‐third of that for a split semi‐implicit scheme and one‐sixth of that for the original explicit Lax‐Wendroff scheme. Verification statistics for split explicit forecasts indicate improved accuracy when compared with those for earlier operational models.

The LIDAR Thomson Scattering System on the JET tokamak is described. Backscattering geometry and the time of flight method are employed to measure profiles along the major radius in the equatorial plane. First electron temperature profiles with a spatial resolution of better than 15 cm obtained with this new diagnostic technique are presented.

SUMMARYIt is demonstrated that, for linear advection in one space dimension, a simple modification to the wellknown Lax-Wendroff integration scheme leads to a substantial reduction in phase speed errors. The modification is designed in such a way that no additional restriction is placed on the timestep used in an integration. The improved performance is comparable or superior to that of previously proposed advection schemes, yet it is achieved without recourse to expensive or cumbersome computations.The new scheme is readily extended to two space dimensions and has been used in numerical weather prediction and sea wave modelling in the Meteorological office. Anexample is given of an improved numerical forecast of the 500 mb height field which was obtained by applying the new scheme in the operational 10-level numerical weather prediction model.