Search code examples
pythonnumpyweathergribnoaa

How to iterate over latitude longitude values of a grib2 file?


I'm trying to get the uvIndex of all the lat,lng present in a grib2 file. This is the link from where I'm getting the file. The problem is I'm not able to understand the structure of the file so that I can get the data. I'm using pygrib to read the file.

Here's the code I've tried:

grbs = pygrib.open('uv.t12z.grbf01.grib2')
grb = grbs.select(name='UV index')[0]
print grb.data(23.5,55.5)

What I'm trying to achieve is either iterate over all the lat longs and print the corresponding uvIndex value or enter a lat long and get the corresponding value. Read the docs of pygrib but couldn't find any suitable command that will serve my purpose. Please help.


Solution

  • You have to iterate though GRIB file and find desirable record, then get data, like here:

    for g in grbs:
        print g.shortName, g.typeOfLevel, g.level # print info about all GRIB records and check names 
        if (g.shortName == shortName and g.typeOfLevel == typeOfLevel and g.level == level):
            tmp = np.array(g.values)
            # now work with tmp as numpy array
    

    To get lat and lon arrays use: lt, ln = g.latlons(), g - element of grbs.

    Read the examples in Section python at https://software.ecmwf.int/wiki/display/GRIB/GRIB+API+examples (pygrib use this library to read GRIB).

    The fastest way to get data from large GRIB file is to make index:

    # use attributes what you want to build index
    indx = pygrib.index(gribfile,'typeOfLevel','level','parameterName') 
    
    # important: msg is an array and may have more then one record
    # get U wind component on 10 m above ground
    msg = indx.select(level = 10, typeOfLevel = "heightAboveGround",
     parameterName = "U U-component of wind m s**-1")
    u10 = np.array(msg[0].values)
    # get V wind component on 10 m above ground
    msg = indx.select(level = 10, typeOfLevel = "heightAboveGround",
     parameterName = "V V-component of wind m s**-1")
    v10 = np.array(msg[0].values)