Incorporate SIS data

[jgieseler]

The Suprathermal Ion Spectrograph (SIS) is the fourth sensor of the EPD suite, and not incorporate yet.

Function get_available_soar_files already is capable of checking for SIS CDF files (level 2 and low-latency) available at SOAR:

>>> from solo_epd_loader import *
>>> get_available_soar_files(20210101, 20210101, 'sis', 'l2')
['solo_L2_epd-sis-a-rates-medium_20210101_V04.cdf', 'solo_L2_epd-sis-a-rates-slow_20210101_V04.cdf', 'solo_L2_epd-sis-b-rates-medium_20210101_V04.cdf', 'solo_L2_epd-sis-b-rates-slow_20210101_V04.cdf']
>>> get_available_soar_files(20210101, 20210101, 'sis', 'll')
['solo_LL02_epd-sis-a-rates_20210101T000046-20210102T000045_V03C.cdf', 'solo_LL02_epd-sis-b-rates_20210101T000046-20210102T000045_V03C.cdf']

Next steps to take:

• Modify functions _epd_l2_download and _epd_ll_download to work with sensor='sis', so that downloads from SOAR are working
• Write new function to actually load the SIS CDF data files, analog to _read_step_cdf and _read_epd_cdf (latter is for EPT and HET). Problem is that SIS data product is structured a bit differently.
• Include sensor=SIS option in epd_load (easy).

[jgieseler]

Maybe it's a better approach to write a new function _read_sis_cdf() that utilizes SunPy's Fido (resp. sunpy_soar) to download the data:

import sunpy_soar
from sunpy.net import Fido
import sunpy.net.attrs as a
from sunpy.timeseries import TimeSeries

# Create search attributes
instrument = a.Instrument("EPD")
time = a.Time("2021-02-01", "2021-02-02")
level = a.Level(2)

# Search
results = Fido.search(instrument & time & level)
print(results)

This lists all the available data products, which then could directly be accessed:

product = a.soar.Product("EPD-SIS-A-RATES-MEDIUM")

# Do search
result = Fido.search(instrument & time & level & product)
print(result)

# Download files
files = Fido.fetch(result)
print(files)

So for that, one would need to hard-code all the interesting data product strings.

Afterwards, the reading could be really easy if the cdf files behave nicely (i.e., there are no 3d data):

solo_sis = TimeSeries(files, concatenate=True)
df_solo_sis = solo_sis.to_dataframe()
return df_solo_sis

[jgieseler]

So, a full working function would already be the following (for one specific data product):

import sunpy_soar
from sunpy.net import Fido
import sunpy.net.attrs as a
from sunpy.timeseries import TimeSeries

def read_sis(startdate, enddate, instrument="EPD", level=2, sis_product="EPD-SIS-A-RATES-MEDIUM"):
    # Create search attributes
    instrument = a.Instrument(instrument)
    time = a.Time(startdate, enddate)
    level = a.Level(level)
    product = a.soar.Product(sis_product)

    # Do search
    result = Fido.search(instrument & time & level & product)
  
    # Download files
    files = Fido.fetch(result)
  
    solo_sis = TimeSeries(files, concatenate=True)
    df_solo_sis = solo_sis.to_dataframe()
    return df_solo_sis

df = read_sis(startdate="2021-02-01", enddate= "2021-02-02", instrument="EPD", level=2, sis_product="EPD-SIS-A-RATES-MEDIUM")