These datasets represent meteorological information prepared by KNMI (Peter van Velthoven) for the CARIBIC project. The meteorological analysis of the CARIBIC measurement flights is based on ECWMF first guess fields (6-hour forecasts) with a resolution of 1x1 degree on ECMWF model levels. First guess fields from ECMWF were used until 13 September 2000, because ECMWF used a 3D-var assimilation scheme until then. For 3D-var, first guess fields are somewhat smoother than analyses. After 13 September 2002 ECMWF used 4D-var assimilation and its analyses were used, as first guess fields were no longer available. With 4D-var the analyses are much smoother than in 3D-var. Data format: ECMWF model data interpolated linearly in longitude, latitude, log(pressure) and time to the aircraft location. The following parameters are interpolated: Temperature (K), Potential vorticity (PVU), Potential temperature (K), Equivalent potential temperature (K), Specific humidity (10e-3 g/kg), Eastward wind component U (m/s), Northward wind component V (m/s), Vertical wind component W=dp/dt (1e-3 mbar/s, 0.1 Pa/s), Wind speed (m/s), Wind direction (deg), Water vapour volume mixing ratio, H2O (ppmv), Relative Humidity (%). Special comments to this flight: None.
The quality of ECMWF analyses is monitored and documented at ECMWF (http://www.ecmwf.int). The following expresses a subjective view by Peter van Velt...
Description
The quality of ECMWF analyses is monitored and documented at ECMWF (http://www.ecmwf.int). The following expresses a subjective view by Peter van Velthoven: The accuracy of model parameters that are also measured can easily be checked. See e.g. http://www.knmi.nl/~/velthove/CARIBIC. The difference between modelled and measured temperatures and potential temperatures is usually less than 2 K, mostly less than 1 K. The horizontal wind speeds are relatively less accurate than temperatures. Measured and modelled wind speeds can deviate up to about 10 m/s. The modelled wind directions are usually quite accurate (within 20 degrees) except when the wind speed is very low, so that the direction is badly defined. The relative and specific humidities are not very accurate because the radiosonde humidity measurements are quite inaccurate still. Errors in RH of 30 % are no exception. 1x1deg, 60 hybrid sigma levels, 6 h Interpol. to flight position every min.