Follow storm track
##################

In some occasion it is practical to follow a storm when retrieving satellite information. **wavy** can do so which is shown in the following with an example on the hurricane track of Katrina until landfall.

Open python in the wavy conda environment:

.. code-block:: bash
   
   $ conda activate wavy
   $ python

Define a track file that could look like the following csv file. It has to include a column called "lat" and one called "lon":

.. code-block:: vim

   date,lat,lon,Pmb,Ukt,CAT
   2005-08-23 18,23.05,-75.01,1008,30,TD
   2005-08-24 00,23.38,-75.64,1007,30,TD
   2005-08-24 06,23.76,-76.20,1007,30,TD
   2005-08-24 12,24.46,-76.50,1006,35,TS
   2005-08-24 18,25.36,-76.88,1003,40,TS
   2005-08-25 00,25.80,-77.70,1000,45,TS
   2005-08-25 06,26.06,-78.36,997,50,TS
   2005-08-25 12,26.18,-78.95,994,55,TS
   2005-08-25 18,26.15,-79.56,988,60,TS
   2005-08-25 22,26.00,-80.10,984,70,H1
   2005-08-26 00,25.88,-80.28,983,70,H1
   2005-08-26 06,25.35,-81.27,987,65,H1
   2005-08-26 12,25.05,-81.96,979,75,H1
   2005-08-26 18,24.87,-82.61,968,85,H2
   2005-08-27 00,24.56,-83.29,959,90,H2
   2005-08-27 06,24.37,-83.98,950,95,H2
   2005-08-27 12,24.36,-84.68,942,100,H3
   2005-08-27 18,24.49,-85.28,948,100,H3
   2005-08-28 00,24.77,-85.89,941,100,H3
   2005-08-28 06,25.15,-86.69,930,125,H4
   2005-08-28 12,25.68,-87.70,909,145,H5
   2005-08-28 18,26.25,-88.61,902,150,H5
   2005-08-29 00,27.16,-89.20,905,140,H5
   2005-08-29 06,28.15,-89.62,913,125,H4
   2005-08-29 11,29.27,-89.61,920,110,H3
   2005-08-29 12,29.50,-89.59,923,110,H3
   2005-08-29 14,30.18,-89.57,928,105,H3

This track is retrieved from *https://coast.noaa.gov/hurricanes*. Save this example under *Katrina_track.csv*.

Download the respective satellite files:

.. code-block:: bash

   ./wavyDownload.py -sat multi -sd 2005082312 -ed 2005083000 -product cci_L3

Now start you python script:

.. code-block:: python3

    # imports
    >>> from wavy.satellite_module import satellite_class as sc
    >>> from wavy.insitu_module import poi_class as pc
    >>> from wavy.utils import parse_date
    >>> import pandas as pd

    >>> f = pd.read_csv('/path/to/your/wavy/tests/data/track/Katrina_track.csv')
    >>> # convert dates to datetime
    >>> dt = [parse_date(d) for d in f.date]
    >>> lons = f.lon.values
    >>> lats = f.lat.values

    >>> # define poi dictionary for track
    >>> poi_dict = {'time':dt,'lons':lons,'lats':lats}

    >>> # initialize poi class object
    >>> pco = pc(poi_dict, nID='Katrina', name='Katrina', varalias='Hs')

    >>> # retrievals
    >>> sco = sc(twin=180, distlim=200, name='multi',
    ...          nID='CCIv1_L3', region='global',
    ...          varalias = 'Hs', # default
    ...          poi=pco)
    >>> sco = sco.populate(path='/path/to/your/wavy/tests/data/L3/multi/')

    >>> # quicklook including track by passing poi
    >>> sco.quicklook(m=True,poi=pco)


This script gathers all footprints within a 200km radius and a +-180min time window given time and location of tracked storm. Executing yields the following figure:

.. image:: ./docs_fig_stormtrack.png
   :scale: 80