BLIP Data File Description

This description is rudimentary but should generally suffice.  Further information (possibly more up-to-date) is given at http://www.drjack.info/twiki/bin/view/RASPop/DataFileDescription.  If you are seriously interested in using the BLIP data files and have questions about what is written here, post your questions to the on-line Blipmap Forum and I will answer them there. 

Data file names and parameters

BLIP data files are available for all BLIPMAP parameters and forecast times.  They can be downloaded using standard web html downloading tools (NOT via anonymous ftp) from the same directories in which BLIPMAP plot images are found - they have the same name as the corresponding BLIPMAP plot, except the data file's name tail is ".data.zip" instead of the image's ".png".  (However, no data file is associated with "composite" plot filenames since those plot 2 parameters - each of those parameters has its own individual data file.)  The easiest way to determine a desired data file name is to view the BLIPMAP plot you are interested in and use that URL with "png" being replaced by "data.zip". 

The data files are in compressed "zip" file format, with only one data file per zip file.  The zip file name ("outer filename") includes the parameter name, the relative day for NAM files only (e.g. "curr" for current day, "curr+1", etc.), and validation time: an example NAM data file name is "wfpm.curr.21z.data.zip"  Note that the model and region is determined by the directory within which the file is located, while the date is referenced to the "current" run date.  The archived data file name ( "inner filename") is more complex, providing the model, region, parameter, relative validation day, validation time, forecast period, initialization time, local validation day (YYMMDD), and GMT validation day and time: an example NAM archived file name is: "blipmap_nam_canv_wfpm_curr_21z_9h_12z_061012=061012_21z".  Note that the primary purpose of the archived name is to provide information about the file that was archived without needing to unzip the data file itself - it is not normally un-archived using that name, as it is more convenient to ignore the archived filename by unzipping the entire zip file (containing a single archive file) to produce an unzipped file, which can be stored by a more convenient name. 

Download is available only for registered BLIPMAP users, access being obtained through use of a DrJack "cookie" or "cookieless" URL query line arguments (see http://www.drjack.info/INFO/registration.html).  Subscriber status is required to download all but the "Upward Thermal Velocity" and "Buoyancy/Shear Ratio" parameters. 

Data file format

The unzipped ASCII (text) data file consists of:

  1. A marker line, which is simply "---" as the first three characters in the line. 
  2. A title line ("first header line"), which currently gives the parameter name, forecast validation time and day, forecast period, and model used.  For example,
          Thermal Updraft Velocity (W*) [fpm] 21Z(14pdt) FRI 26 May 2006 6hrFcst RUC
    Note: this title is intended for printing at the top of my plots, not for parsing to obtain parameter values.  The title can contain markup used by my plotting routine to generate sub/superscripts and non-alphabetic characters.  And it can change without warning.  So parameters such as the date, etc. should instead be obtained from the other header lines.  .  (BTW, any scaling factor displayed with the title units does not apply to the data values within the file - that factor is applied separately, in the plotting routine). 
  3. A grid information line ("second header line"), which gives the model and region used, gridding information (following a ' Indexs= ' identifier, the xmin,xmax,ymin,ymax indicies of the model grid covered by that subgrid), and grid projection information (following a ' Proj= ' identifier).  The "Proj" projection parameters are included so that grid point lat/lon values can be computed if needed (especially if the grid used is no longer the current grid). The projection values are: (1) Type (should always be "lambert"), (2) X direction grid spacing (m), (3) Y direction grid spacing (m) (should always equal X direction spacing), (4) Latitude of projection's first lambert cone intersection (deg), (5) Latitude of projection's second lambert cone intersection (deg) (will always equal (4) for a tangent-plane lambert projection, as all RASP projections should be), (6) Longitude of projection's vertical meridian (deg), (7) Latitude of local grid center (deg), (8) Longitude of local grid center (deg). Parameters (4), (5), and (6) are the actual projection parameters whereas the other parameters specify the specific location of the grid within that projection. Note that the center of the grid may not be at a grid point location and that each of the nested grids can have a different center!
          For example,
          Model= RUC Region= RUC-CANV Indexs= 38 82 79 136 Proj= lambert 20318 20318 25.0000 25.0000 -95.0000 37.3869 -118.6811
    Here a "Identifier= value(s)" format is used, with a single space (blank character) always being used as a delimiter (a space will never be used in an identifier or in a value).  The ordering of the identifiers is not guaranteed - e.g. a new identifier+value combination might later be inserted into the middle or end of the current structure to accomodate newly identified needs. 
  4. A data information line ("third header line"), which gives the soaring day, local validation time, GMT validation time, forecast period (in hours), model analysis (initialization) time (in GMT), abbreviated parameter identifier, parameter units, multiplier used to convert actual values to datafile values, and minimum and maximum values (with their grid locations). 
         For example,
          Day= 2006 5 26 FRI ValidLST= 14 pdt ValidZ= 21 Fcst= 6 AnalZ= 15 Param= wfpm Unit= ft/min Mult= 1 Min= 315 (38,79) Max=855 (82,136)
    This also uses the "Identifier= value(s)" format.
  5. The data values themselves, with one data row per plot row.  The first data row is for the lowest (most "southerly", or smallest y index) grid row.  Subsequent rows are for incremental increases in the y index.  Most parameters have integer data values, the "actual" value being that integer divided by the given multiplication factor. 

Latitude/longitude values

The data values can often be utilized, for example in plots, without needing to know the latitude/longitude of individual points.  But for some purposes the lat/lon values are needed.  Files of regional lat/lon values do exist - these are in the same zipped format described above for the data file format, but the lat/lon values are given in floating point (non-integer) numbers.  The lat/lon data files have names "latitude.data.zip" and "longitude.data.zip" and are found one directory up from the forecast (FCST) directory used for the parameter data files.  It is also possible to calculate the values as needed.  For the RUC model this is done using the algorithm in this Fortran file (the algorithm can also be used for NAM forecasts but different map parameter values are needed).  The header lines used for these files will not contain all the identifier/value combinations used for the data files since some are not applicable to these files, e.g. "Day" and "Time" identifiers and values are not included.