Function to load from NOAA downloaded dataset to pandas dataframe (tested on MEI only)

load_noaa.py

import pandas as pd

def load_noaa(filepath, skiprows=1, skipfooter=4, na_values=-999):
    """Loads a NOAA style downloaded dataset into a tidy pandas dataframe
    
    Parameters
    ----------
    filepath : str
        Filepath of dataset
    skiprows : int, optional
        Number of rows of non-data at the top of the file. (the default is 1, 
        which works for MEIv2)
    skipfooter : int, optional
        Number of rows of non-data at the bottom of the file. (the default is 4, 
        which works for MEIv2)
    na_values : float, optional
        Missing value number (the default is -999, which works for MEIv2)

    Returns
    -------
    pandas.DataFrame
    
    """
    df = pd.read_csv(filepath, sep='\s+', 
                     skiprows=skiprows, skipfooter=skipfooter,  
                     header=None, na_values=na_values)
    df = pd.melt(df, id_vars=[0])
    
    df['date'] = pd.to_datetime(
        df.apply(
            lambda row: '{:04d}/{:02d}/01'.format(int(row.iloc[0]), int(row.iloc[1])), 
          axis=1)
    )
    df.loc[:, [ 'date', 'value']].set_index('date').sort_values('date')
    return df

if __name__=='__main__':
  filepath = "meiv2.data"
  df = load_noaa(filepath, skiprows=1, skipfooter=4)
  df.head()

Can confirm that this function does work for other NOAA datasets.

For the Atlantic oscillation dataset, you can load it using parameters such as

load_noaa("ao.long.data", skiprows=1, skipfooter=7, na_values=-9.99)