The data this page describes are here. The format is very similar to the ECMWF format, with the following exceptions: height above the ground is included, var, flx and sfc files are provided, there are fewer vertical levels, and the extraction code runs on the midnight model run (i.e. at 00h, and data are extracted for the model forecasts +12h to +36h).
The data for each Cloudnet site consists of three ASCII files per day, with names of the form cloudnet_AAA_NNN_YYYYMMDD, where YYYYMMDD is the date, NNN is the name of the site and AAA is either var (variables on model levels), flx (fluxes between levels) and sfc (some surface fluxes).
Each file has a four line header, with each line beginning with "#". The first three are self-explanatory. The fourth line contains the variable names for each of the columns of data that follow. The first four columns are common to all files: idat and itime are the initial date and time of the forecast from which the data was retrieved. vdate and vtime are the verifying date and time of the data meaning date and time which it is valid for. Hence the difference to the initial date and time indicates the forecast range used. The next column (lev) is the model level of the data. The columns that follow these are specific to the file type:
These files contain the model variables on the 41 model levels. Level 1 is the top full level of the model, level 41 is the lowest full level. The order of the variables is:
These files contain the radiative, turbulent, and precipitation fluxes throughout the atmosphere. The fluxes are averages over the last hour.
The order of the variables is:
These files contain some surface/single-level fields and fluxes.The level number is set to 42.
The order is: