Legacy Basics Tutorial#

Overview#

Welcome to aurora-asi-lib. This tutorial will guide you though the main functions in aurora-asi-lib (imported as asilib).

First off, we need to import the necessary packages.

[3]:

from datetime import datetime, timedelta
from IPython.display import Video
import numpy as np

import matplotlib.pyplot as plt
import asilib

plt.style.use('dark_background')

print(f'asilib version: {asilib.__version__}')

asilib version: 0.12.1

First of all, you should know where the data and movies are saved to. This information is in asilib.config and can be changed with python3 -m asilib config to configure asilib.

[4]:

asilib.config

[4]:

{'ASILIB_DIR': WindowsPath('C:/Users/mshumko/Documents/research/aurora-asi-lib/asilib'),
 'ASI_DATA_DIR': WindowsPath('C:/Users/mshumko/asilib-data')}

As you can guess, asilib.config['ASILIB_DIR'] is the directory where this code resides, asilib.config['ASI_DATA_DIR'] is the directory where the data is saved to.

Plot a single image#

Since we sold you on easily downloading, loading, plotting and analyzing auroral images with asilib, lets begin with a plot of an omega band studied in:

Liu, J., Lyons, L. R., Archer, W. E., Gallardo-Lacourt, B., Nishimura, Y., Zou, Y., … Weygand, J. M. (2018). Flow shears at the poleward boundary of omega bands observed during conjunctions of Swarm and THEMIS ASI. Geophysical Research Letters, 45, 1218– 1227. https://doi.org/10.1002/2017GL076485

We can plot a fisheye lens image using asilib.plot_fisheye.

[5]:

asi_array_code = 'THEMIS'
location_code = 'ATHA'
time = datetime(2008, 3, 9, 9, 18, 0)  # You can supply a datetime object or a ISO-formatted time string.

image_time, image, ax, im = asilib.plot_fisheye(asi_array_code, location_code, time,
                    color_norm='log', color_map='auto')
plt.colorbar(im)
ax.axis('off');
../_images/tutorials_legacy_tutorial_8_0.png

It is also easy to map the ASI fisheye lens image to a geographic map using asilib.plot_map. In the code box below, the first line creates a subplot with a geographic map centered on Athabasca. If latitude or longitude bounds are not provided, asilib.plot_map defaults to a map of North America.

[6]:

ax = asilib.make_map(lon_bounds=(-127, -90), lat_bounds=(40, 65))

map_alt_km = 110
asilib.plot_map(asi_array_code, location_code, time, map_alt_km, ax=ax);
../_images/tutorials_legacy_tutorial_10_0.png

Notice that you did not need to explicitly download or load the data—asilib takes care of that for you. If you need to explicitly download data, asilib comes with four functions to download the image and skymap files.

Now what if you need to analyze the image? asilib.plot_fisheye returns the time stamp of the image image_time and the 2-d image np.array image.

[7]:

image_time, image

[7]:

(datetime.datetime(2008, 3, 9, 9, 18, 0, 51000),
 array([[2546, 2537, 2572, ..., 2554, 2616, 2536],
        [2608, 2613, 2562, ..., 2620, 2582, 2582],
        [2550, 2526, 2588, ..., 2568, 2560, 2544],
        ...,
        [2629, 2541, 2612, ..., 2553, 2546, 2525],
        [2569, 2568, 2510, ..., 2698, 2596, 2545],
        [2576, 2617, 2514, ..., 2602, 2577, 2502]], dtype=uint16))

You can also just load that image by using asilib.load_image().

[8]:

image_time, image = asilib.load_image(asi_array_code, location_code, time=time, redownload=False)

Plot a Keogram#

asilib.plot_keogram plots a keogram through the meridian, and using it is as simple as plotting a single image. If you don’t specify a map_alt, the vertical axis will be pixel number. Note: asilib.keogram returns the keogram array that is plotted by asilib.plot_keogram.

[9]:

time_range = [datetime(2008, 3, 9, 9, 0, 0), datetime(2008, 3, 9, 9, 30, 0)]
asilib.plot_keogram(asi_array_code, location_code, time_range)
plt.xlabel('Time'); plt.ylabel('Pixel number');
../_images/tutorials_legacy_tutorial_16_0.png

And if you specify a map_alt (in kilometers), the vertical axis will be geographic latitude.

[10]:

map_alt_km = 110
asilib.plot_keogram(asi_array_code, location_code, time_range, map_alt=map_alt_km)
plt.xlabel('Time'); plt.ylabel('Geographic latitude [deg]');
../_images/tutorials_legacy_tutorial_18_0.png

And lastly, if you specify a map_alt and set aacgm=True, the keogram’s vertical axis will be magnetic latitude estimated using the aacgmv2 Python package that implements the Altitude-adjusted corrected geomagnetic coordinates defined in Shepherd 2014.

[11]:

map_alt_km = 110
asilib.plot_keogram(asi_array_code, location_code, time_range, map_alt=map_alt_km, aacgm=True)
plt.xlabel('Time'); plt.ylabel('Magnetic latitude [deg]');
../_images/tutorials_legacy_tutorial_20_0.png

Simple enough? You can choose any other altitude from the FULL_MAP_ALTITUDE key in the skymap calibration data (described in the following section). If you pick a wrong altitude, asilib will give you a helpful error.

[12]:

try:
    asilib.plot_keogram(asi_array_code, location_code, time_range, map_alt=100)
except AssertionError as err:
    print('AssetionError:', err)

AssetionError: 100 is not in the skymap altitudes: [ 90. 110. 150.]

These aurora-emission altitudes are in the THEMIS ASI skymap files; They are different for the REGO ASIs.

Skymap calibration data#

You may wonder how the image’s pixel values were mapped to geographic latitude—this is where the skymap files come in. asilib also makes this easy and you only need to give it the ASI array code, location code, and the time (so the correct skymap file is loaded).

[13]:

skymap = asilib.load_skymap(asi_array_code, location_code, time)
skymap.keys()

[13]:

dict_keys(['GENERATION_INFO', 'SITE_UID', 'IMAGER_UID', 'PROJECT_UID', 'IMAGER_UNIX_TIME', 'SITE_MAP_LATITUDE', 'SITE_MAP_LONGITUDE', 'SITE_MAP_ALTITUDE', 'FULL_ROW', 'FULL_COLUMN', 'FULL_IGNORE', 'FULL_SUBTRACT', 'FULL_MULTIPLY', 'FULL_ELEVATION', 'FULL_AZIMUTH', 'FULL_MAP_ALTITUDE', 'FULL_MAP_LATITUDE', 'FULL_MAP_LONGITUDE', 'FULL_BIN', 'BIN_ROW', 'BIN_COLUMN', 'BIN_ELEVATION', 'BIN_AZIMUTH', 'BIN_MAP_ALTITUDE', 'BIN_MAP_LATITUDE', 'BIN_MAP_LONGITUDE', 'SKYMAP_PATH'])

The skymap data that asilib uses for keograms is the FULL_MAP_LATITUDE. Also, to confirm that you loaded the correct skymap file, the skymap dictionary contains a SKYMAP_PATH key that points to the local skymap file.

[14]:

skymap['SKYMAP_PATH']

[14]:

WindowsPath('C:/Users/mshumko/asilib-data/themis/skymap/atha/themis_skymap_atha_20070301-20090522_vXX.sav')

The possible skymap mapping altitudes (in meters) that you can use in asilib.plot_keogram and asilib.plot_map.

[15]:

skymap['FULL_MAP_ALTITUDE']

[15]:

array([ 90000., 110000., 150000.], dtype=float32)

Mapping multiple all-sky images#

You can plot one image from multiple ASI locations using a for-loop. In the following example, we will replicate Fig. 2b from:

Donovan, E., Liu, W., Liang, J., Spanswick, E., Voronkov, I., Connors, M., … & Rae, I. J. (2008). Simultaneous THEMIS in situ and auroral observations of a small substorm. Geophysical Research Letters, 35(17).

[16]:

# ASI parameters
time = datetime(2007, 3, 13, 5, 8, 45)
asi_array_code = 'THEMIS'
location_codes = ['FSIM', 'ATHA', 'TPAS', 'SNKQ']
map_alt = 110
min_elevation = 2

fig, bx = plt.subplots(figsize=(8, 5))
bx = asilib.make_map(lon_bounds=(-160, -52), lat_bounds=(40, 82), ax=bx)

for location_code in location_codes:
    asilib.plot_map(asi_array_code, location_code, time, map_alt, ax=bx,
        min_elevation=min_elevation)

bx.set_title('Donovan et al. 2008 | First breakup of an auroral arc');

Made directory at C:\Users\mshumko\asilib-data\themis\skymap\fsim
Downloading themis_skymap_fsim_20070126_vXX.sav: |####################| 100%
Downloading themis_skymap_fsim_20070201-20081229_vXX.sav: |####################| 100%
Downloading themis_skymap_fsim_20090123-20090503_vXX.sav: |####################| 100%
Downloading themis_skymap_fsim_20090912_vXX.sav: |####################| 100%
Downloading themis_skymap_fsim_20110202_vXX.sav: |####################| 100%
Downloading themis_skymap_fsim_20120226_vXX.sav: |####################| 100%
Downloading themis_skymap_fsim_20130103_vXX.sav: |####################| 100%
Downloading themis_skymap_fsim_20150915_vXX.sav: |####################| 100%
Downloading themis_skymap_fsim_20170906_vXX.sav: |####################| 100%
Downloading themis_skymap_fsim_20190227_vXX.sav: |####################| 100%
Downloading themis_skymap_fsim_20190303_v02.sav: |####################| 100%
Downloading themis_skymap_fsim_20191031_v02.sav: |####################| 100%
Downloading themis_skymap_fsim_20200313_v02.sav: |####################| 100%
Downloading themis_skymap_fsim_20201111_v02.sav: |####################| 100%
Downloading themis_skymap_fsim_20210307_v02.sav: |####################| 100%
Downloading themis_skymap_fsim_20211205_v02.sav: |####################| 100%
Downloading themis_skymap_fsim_20220304_v02.sav: |####################| 100%
Made directory at C:\Users\mshumko\asilib-data\themis\skymap\tpas
Downloading themis_skymap_tpas_20070214_vXX.sav: |####################| 100%
Downloading themis_skymap_tpas_20070217_vXX.sav: |####################| 100%
Downloading themis_skymap_tpas_20080116-20080118_vXX.sav: |####################| 100%
Downloading themis_skymap_tpas_20070803-20090514_vXX.sav: |####################| 100%
Downloading themis_skymap_tpas_20090914_vXX.sav: |####################| 100%
Downloading themis_skymap_tpas_20100909_vXX.sav: |####################| 100%
Downloading themis_skymap_tpas_20110207_vXX.sav: |####################| 100%
Downloading themis_skymap_tpas_20120226_vXX.sav: |####################| 100%
Downloading themis_skymap_tpas_20130105_vXX.sav: |####################| 100%
Downloading themis_skymap_tpas_20151206_vXX.sav: |####################| 100%
Downloading themis_skymap_tpas_20170926_vXX.sav: |####################| 100%
Downloading themis_skymap_tpas_20190204_vXX.sav: |####################| 100%
Downloading themis_skymap_tpas_20200302_v02.sav: |####################| 100%
Downloading themis_skymap_tpas_20200918_v02.sav: |####################| 100%
Downloading themis_skymap_tpas_20210304_v02.sav: |####################| 100%
Downloading themis_skymap_tpas_20220304_v02.sav: |####################| 100%

c:\Users\mshumko\.conda\envs\asilib_test\lib\site-packages\scipy\io\idl.py:281: UserWarning: Not able to verify number of bytes from header
  warnings.warn("Not able to verify number of bytes from header")

Made directory at C:\Users\mshumko\asilib-data\themis\skymap\snkq
Downloading themis_skymap_snkq_20070306_vXX.sav: |####################| 100%
Downloading themis_skymap_snkq_20070307_vXX.sav: |####################| 100%
Downloading themis_skymap_snkq_20070313_vXX.sav: |####################| 100%
Downloading themis_skymap_snkq_20071219_vXX.sav: |####################| 100%
Downloading themis_skymap_snkq_20070919-20090521_vXX.sav: |####################| 100%
Downloading themis_skymap_snkq_20090913_vXX.sav: |####################| 100%
Downloading themis_skymap_snkq_20100910_vXX.sav: |####################| 100%
Downloading themis_skymap_snkq_20110207_vXX.sav: |####################| 100%
Downloading themis_skymap_snkq_20130105_vXX.sav: |####################| 100%
Downloading themis_skymap_snkq_20150516_vXX.sav: |####################| 100%
Downloading themis_skymap_snkq_20170926_vXX.sav: |####################| 100%
Downloading themis_skymap_snkq_20180919_vXX.sav: |####################| 100%
Downloading themis_skymap_snkq_20190204_v02.sav: |####################| 100%
Downloading themis_skymap_snkq_20200301_v02.sav: |####################| 100%
Downloading themis_skymap_snkq_20200920_v02.sav: |####################| 100%
Downloading themis_skymap_snkq_20210304_v02.sav: |####################| 100%
Downloading themis_skymap_snkq_20210901_v02.sav: |####################| 100%
Downloading themis_skymap_snkq_20220304_v02.sav: |####################| 100%
../_images/tutorials_legacy_tutorial_31_3.png

Working with multiple images#

The asilib.load_image function can also load (and automatically download) image data given a time duration specified by time_range. In this section, we will plot a montage of fisheye lens images that show the equatorward movement of a STEVE aurora studied in

Gallardo-Lacourt, B., Nishimura, Y., Donovan, E., Gillies, D. M., Perry, G. W., Archer, W. E., et al. (2018). A statistical analysis of STEVE. Journal of Geophysical Research: Space Physics, 123, 9893– 9905. https://doi.org/10.1029/2018JA025368

[17]:

asi_array_code = 'REGO'
location_code = 'LUCK'
time_range = [datetime(2017, 9, 27, 7, 15), datetime(2017, 9, 27, 8, 15)]

image_times, images = asilib.load_image(asi_array_code, location_code, time_range=time_range, redownload=False)

[18]:

image_times.shape

[18]:

(1200,)

[19]:

images.shape  # First axis is time

[19]:

(1200, 512, 512)

Now lets pick 5 images from that time interval and show the movement of STEVE.

[20]:

n_plots = 5

delta_time_s = int((time_range[1]-time_range[0]).total_seconds()/n_plots)
montage_times = [time_range[0]+timedelta(seconds=i*delta_time_s) for i in range(n_plots)]

fig, cx = plt.subplots(1, n_plots, figsize=(15, 8))

for montage_time, cx_i in zip(montage_times, cx):
    asilib.plot_fisheye(asi_array_code, location_code, montage_time, ax=cx_i)
    cx_i.axis('off')
../_images/tutorials_legacy_tutorial_37_0.png

Make a movie#

Let’s now make a simple fisheye lens movie of a substorm using asilib.animate_fisheye.

[21]:

asi_array_code = 'THEMIS'
location_code = 'FSMI'
time_range = [datetime(2015, 3, 26, 6, 7, 0), datetime(2015, 3, 26, 6, 30, 0)]

# loglevel is to suppress the verbose ffmpeg output.
asilib.animate_fisheye(asi_array_code, location_code, time_range, overwrite=True, ffmpeg_output_params={'loglevel':'quiet'})
plt.close()  # To show a clean output in this tutorial---it is often unnecessary.

# When you run this, you should see the video below in your asilib-data/movies directory.
Video('https://github.com/mshumko/aurora-asi-lib/blob/e6147ff1a3309c602c1aa48711ebc8a90a7863e1/docs/_static/'
    '20150326_060700_062957_themis_fsmi.mp4?raw=true')

Created a C:\Users\mshumko\asilib-data\animations\images\20150326_060700_themis_fsmi_fisheye directory

[21]:

Animating images into the ionosphere is also straightforward.

[22]:

asi_array_code = 'THEMIS'
location_code = 'FSMI'
time_range = [datetime(2015, 3, 26, 6, 7, 0), datetime(2015, 3, 26, 6, 30, 0)]

# We need the skymap only to center the map on the projected image.
skymap = asilib.load_skymap(asi_array_code, location_code, time_range[0])
lat_bounds = (skymap['SITE_MAP_LATITUDE']-7, skymap['SITE_MAP_LATITUDE']+7)
lon_bounds = (skymap['SITE_MAP_LONGITUDE']-10, skymap['SITE_MAP_LONGITUDE']+10)

ax = asilib.make_map(lon_bounds=lon_bounds, lat_bounds=lat_bounds)
plt.subplots_adjust(top=0.99, bottom=0.05, left=0.05, right=0.99)
asilib.animate_map(asi_array_code, location_code, time_range, 110, overwrite=True, ax=ax,
    ffmpeg_output_params={'loglevel':'quiet'})

plt.close()  # To show a clean output in this tutorial---it is often unnecessary.

# When you run this, you should see the video below in your asilib-data/movies directory.
Video('https://github.com/mshumko/aurora-asi-lib/blob/main/docs/_static/20150326_060700_062957_themis_fsmi_map.mp4?raw=true')

Created a C:\Users\mshumko\asilib-data\animations\images\20150326_060700_themis_fsmi_map directory

[22]:

In the next section, we make a more sophisticated movie using asilib.animate_fisheye_generator.

Satellite conjunction#

As a last example, here we show how asilib can be used to make a conjunction movie. Be forewarned, this example is long. We will use ASI data from the THEMIS-RANK imager.

[23]:

asi_array_code = 'THEMIS'
location_code = 'RANK'
time_range = (datetime(2017, 9, 15, 2, 32, 0), datetime(2017, 9, 15, 2, 35, 0))

Now we define an orbit path of a low Earth orbiting satellite. In this context, lla stands for the (latitude, longitude, altitude) coordinates. That we place near the THEMIS-RANK imager, hence we use load_skymap to extract the imager’s location in Canada.

[24]:

# Load the skymap calibration data. This is necessary to create the satellite track overhead.
skymap_dict = asilib.load_skymap(asi_array_code, location_code, time_range[0])

# Create the fake satellite track coordinates: latitude, longitude, altitude (LLA).
# This is a north-south satellite track oriented to the east of the THEMIS/RANK
# imager.
n = int((time_range[1] - time_range[0]).total_seconds() / 3)  # 3 second cadence.
lats = np.linspace(skymap_dict["SITE_MAP_LATITUDE"] + 5, skymap_dict["SITE_MAP_LATITUDE"] - 5, n)
lons = (skymap_dict["SITE_MAP_LONGITUDE"] - 0.5) * np.ones(n)
alts = 110 * np.ones(n)
lla = np.array([lats, lons, alts]).T

Made directory at C:\Users\mshumko\asilib-data\themis\skymap\rank
Downloading themis_skymap_rank_20061101-20071203_vXX.sav: |####################| 100%
Downloading themis_skymap_rank_20071212_vXX.sav: |####################| 100%
Downloading themis_skymap_rank_20071218_vXX.sav: |####################| 100%
Downloading themis_skymap_rank_20071207-20090428_vXX.sav: |####################| 100%
Downloading themis_skymap_rank_20090911_vXX.sav: |####################| 100%
Downloading themis_skymap_rank_20100909_vXX.sav: |####################| 100%
Downloading themis_skymap_rank_20110202_vXX.sav: |####################| 100%
Downloading themis_skymap_rank_20120225_vXX.sav: |####################| 100%
Downloading themis_skymap_rank_20130107_vXX.sav: |####################| 100%
Downloading themis_skymap_rank_20150825_vXX.sav: |####################| 100%
Downloading themis_skymap_rank_20170915_v02.sav: |####################| 100%
Downloading themis_skymap_rank_20171003_v02.sav: |####################| 100%
Downloading themis_skymap_rank_20190206_v02.sav: |####################| 100%
Downloading themis_skymap_rank_20190827_v02.sav: |####################| 100%
Downloading themis_skymap_rank_20200227_v02.sav: |####################| 100%
Downloading themis_skymap_rank_20200919_v02.sav: |####################| 100%
Downloading themis_skymap_rank_20210309_v02.sav: |####################| 100%
Downloading themis_skymap_rank_20210829_v02.sav: |####################| 100%
Downloading themis_skymap_rank_20220304_v02.sav: |####################| 100%

With the satellite track, we first need to map its location to the imager’s elevation and azimuth (azel) using asilib.lla2azel. Note: if you need to find the satellite’s footprint, you can use asilib.lla2footprint to map along the magnetic field line before you pass the footprint LLA into asilib.lla2azel. However, asilib.lla2footprint requires the IRBEM library.

[25]:

sat_azel, sat_azel_pixels = asilib.lla2azel(asi_array_code, location_code, time_range[0], lla)

Here we create the conjunction movie. We can’t easily split up the steps into notebook cells, but I’ve add STEPS 1-4 for the main steps:

  • STEP 1: create the subplots. ax[0] will plot the ASI images, while ax[1] will plot the ASI intensity time series along the satellite path. We initialize the asilib.animate_fisheye_generator function which allows us to manipulate the individual images before they are stitched together into a movie.

  • STEP 2: We retrieve the imager time stamps and images using movie_generator (technically it is a coroutine). We need this to calculate the mean ASI intensity along the spacecraft track.

  • STEP 3: We use asilib.equal_area to calculate the pixels inside a 20x20 km area at 110 km around the satellite’s footprint (inferred from LLA). asilib.equal_area returns a mask: 1s inside the area, and NaNs outside it. We use it with image_data.images and np.nanmean to calculate the mean ASI intensity in all frames simultaneously.

  • STEP 4: Loop over every image in time_range. asilib.animate_fisheye_generator will plot the current frame in ax[0] on top of which we superpose the full satellite track and its current location. We also use plt.contour to plot the 20x20 km area at 110 km in yellow. Lastly, in ax[1] we plot the mean ASI time series in the area box around the satelite’s footprint, as well as a vertical line to guide your eye to the corresponding image in ax[0].

[26]:

# STEP 1
fig, ax = plt.subplots(
    2, 1, figsize=(7, 10), gridspec_kw={'height_ratios': [4, 1]}, constrained_layout=True
)
movie_generator = asilib.animate_fisheye_generator(
    asi_array_code, location_code, time_range, azel_contours=True, overwrite=True, ax=ax[0],
    ffmpeg_output_params={'loglevel':'quiet'}
)

# STEP 2
# Use the generator to get the images and time stamps to estimate mean the ASI
# brightness along the satellite path and in a (10x10 km) box.
image_data = movie_generator.send('data')

# STEP 3
# Calculate what pixels are in a box_km around the satellite, and convolve it
# with the images to pick out the ASI intensity in that box.
area_box_mask = asilib.equal_area(
    asi_array_code, location_code, time_range[0], lla, box_km=(20, 20)
)
asi_brightness = np.nanmean(image_data.images * area_box_mask, axis=(1, 2))
area_box_mask[np.isnan(area_box_mask)] = 0  # To play nice with plt.contour()

# STEP 4
for i, (time, image, _, im) in enumerate(movie_generator):
    # Note that because we are drawing moving data: ASI image in ax[0] and
    # the ASI time series + a vertical bar at the image time in ax[1], we need
    # to redraw everything at every iteration.

    # Clear ax[1] (ax[0] cleared by asilib.animate_fisheye_generator())
    ax[1].clear()
    # Plot the entire satellite track
    ax[0].plot(sat_azel_pixels[:, 0], sat_azel_pixels[:, 1], 'red')
    ax[0].contour(area_box_mask[i, :, :], levels=[0.99], colors=['yellow'])
    # Plot the current satellite position.
    ax[0].scatter(sat_azel_pixels[i, 0], sat_azel_pixels[i, 1], c='red', marker='o', s=50)

    # Plot the time series of the mean ASI intensity along the satellite path
    ax[1].plot(image_data.time, asi_brightness)
    ax[1].axvline(time, c='b')  # At the current image time.

    # Annotate the location_code and satellite info in the top-left corner.
    location_code_str = (
        f'{asi_array_code}/{location_code} '
        f'LLA=({skymap_dict["SITE_MAP_LATITUDE"]:.2f}, '
        f'{skymap_dict["SITE_MAP_LONGITUDE"]:.2f}, {skymap_dict["SITE_MAP_ALTITUDE"]:.2f})'
    )
    satellite_str = f'Satellite LLA=({lla[i, 0]:.2f}, {lla[i, 1]:.2f}, {lla[i, 2]:.2f})'
    ax[0].text(
        0,
        1,
        location_code_str + '\n' + satellite_str,
        va='top',
        color='red',
    )
    ax[1].set(xlabel='Time', ylabel='Mean ASI intensity [counts]')

print(f'Movie saved in {asilib.config["ASI_DATA_DIR"] / "movies"}')
plt.close()

Created a C:\Users\mshumko\asilib-data\animations\images\20170915_023200_themis_rank_fisheye directory
Movie saved in C:\Users\mshumko\asilib-data\movies

[27]:

# When you run this, you should see the video below in your asilib-data/movies directory.
Video('https://github.com/mshumko/aurora-asi-lib/blob/main/docs/_static/20170915_023200_023457_themis_rank.mp4?raw=true')

[27]: