JPL Horizons Queries (`astroquery.jplhorizons`/astroquery.solarsystem.jpl.horizons)¶

Warning

The default search behavior has changed. In v0.4.3 and earlier, the default id_type was 'smallbody'. With v0.4.4, the default is None, which implements JPL Horizons’s default behavior: search major bodies first, and if no major bodies are found, then search small bodies.

Note

Due to serverside changes the jplhorizons module requires astroquery v0.4.1 or newer. Previous versions are not expected to function, please upgrade the package if you observe any issues.

Overview¶

The HorizonsClass class provides an interface to services provided by the Solar System Dynamics group at the Jet Propulation Laboratory.

Because of its relevance to Solar System science, this service can also be accessed from the topical submodule astroquery.solarsystem.jpl. The functionality of that service is identical to the one presented here.

In order to query information for a specific Solar System body, a Horizons object has to be instantiated:

>>> from astroquery.jplhorizons import Horizons
>>> obj = Horizons(id='Ceres', location='568', epochs=2458133.33546)
>>> print(obj)
JPLHorizons instance "Ceres"; location=568, epochs=[2458133.33546], id_type=None

id refers to the target identifier and is mandatory; the exact string will be used in the query to the Horizons system.

location means either the observer’s location (e.g., Horizons ephemerides query) or the body relative to which orbital elements are provided (e.g., Horizons orbital elements or vectors query); the same codes as used by Horizons are used here, which includes MPC Observatory codes. The default is location=None, which uses a geocentric location for ephemerides queries and the Sun as central body for orbital elements and state vector queries. User-defined topocentric locations for ephemerides queries can be provided, too, in the form of a dictionary. The dictionary has to be formatted as follows: {'lon': longitude in degrees (East positive, West negative), 'lat': latitude in degrees (North positive, South negative), 'elevation': elevation in km above the reference ellipsoid}. In addition, 'body' can be set to the Horizons body ID of the central body if different from Earth; by default, it is assumed that this location is on Earth if it has not been specifically set. The following example uses the coordinates of the Statue of Liberty as the observer’s location:

>>> statue_of_liberty = {'lon': -74.0466891,
...                      'lat': 40.6892534,
...                      'elevation': 0.093}
>>> obj = Horizons(id='Ceres',
...                location=statue_of_liberty,
...                epochs=2458133.33546)
>>> print(obj)
JPLHorizons instance "Ceres"; location={'lon': -74.0466891, 'lat': 40.6892534, 'elevation': 0.093}, epochs=[2458133.33546], id_type=None

epochs is either a scalar or list of Julian dates (floats or strings) in the case of discrete epochs, or, in the case of a range of epochs, a dictionary that has to include the keywords start, stop (both using the following format “YYYY-MM-DD [HH:MM:SS]”), and step (e.g., '1m' for one minute, '3h' three hours, '10d' for ten days). Note that all input epochs, both calendar dates/times and Julian Dates, refer to UTC for ephemerides queries, TDB for element queries and vector queries. By default, epochs=None, which uses the current date and time.

id_type controls how Horizons resolves the ‘id’ to match a Solar System body:

`id_type`	Query behavior
`None` (default)	Searches major bodies (planets, natural satellites, spacecraft, special cases) first, and if none are found, then searches small bodies.
`smallbody`	Limits the search to small solar system bodies (comets and asteroids).
`designation`	Limits the search to small body designations, e.g., 73P or 2014 MU69.
`name`	Limits the search to asteroid or comet names, e.g., Halley will match 1P/Halley and (2688) Halley.
`asteroid_name`	Limits the search to asteroid names, e.g., Don Quixote.
`comet_name`	Limits the search to comet names, e.g., Encke will only match comet 2P/Encke, and not (9134) Encke.

In the case of ambiguities in the name resolution, a list of matching objects will be provided. In order to select an object from this list, provide the respective id number or record number as id and use id_type=None:

>>> from astroquery.jplhorizons import Horizons
>>> print(Horizons(id='Encke').ephemerides())
Traceback (most recent call last):
...
ValueError: Ambiguous target name; provide unique id:
    Record #  Epoch-yr  Primary Desig  >MATCH NAME<
    --------  --------  -------------  -------------------------
        9134            4822 P-L        Encke
    90000034    1786    2P              Encke
    90000035    1796    2P              Encke
    90000036    1805    2P              Encke
         ...     ...    ...               ...
>>> print(Horizons(id='90000034', id_type=None).ephemerides())  
targetname       datetime_str          datetime_jd    ... RA_3sigma DEC_3sigma
   ---               ---                    d         ...   arcsec    arcsec
---------- ------------------------ ----------------- ... --------- ----------
  2P/Encke 2018-Jan-17 05:06:07.709 2458135.712589224 ...        --         --

The JPL Horizons system provides ephemerides, orbital elements, and state vectors for almost all known Solar System bodies. These queries are provided through three functions: ephemerides(), elements(), and vectors().

Ephemerides¶

ephemerides() returns ephemerides for a given observer location (location) and epoch or range of epochs (epochs) in the form of an astropy table. The following example queries the ephemerides of asteroid (1) Ceres for a range of dates as seen from Mauna Kea:

>>> from astroquery.jplhorizons import Horizons
>>> obj = Horizons(id='Ceres', location='568',
...                epochs={'start':'2010-01-01', 'stop':'2010-03-01',
...                        'step':'10d'})
>>> eph = obj.ephemerides()
>>> print(eph)
     targetname       datetime_str   datetime_jd ... alpha_true  PABLon  PABLat
        ---               ---             d      ...    deg       deg     deg
 ----------------- ----------------- ----------- ... ---------- -------- ------
 1 Ceres (A801 AA) 2010-Jan-01 00:00   2455197.5 ...    12.3609 238.2494 4.5532
 1 Ceres (A801 AA) 2010-Jan-11 00:00   2455207.5 ...    14.1057 241.3339 4.2832
 1 Ceres (A801 AA) 2010-Jan-21 00:00   2455217.5 ...    15.7313 244.3394 4.0089
 1 Ceres (A801 AA) 2010-Jan-31 00:00   2455227.5 ...    17.2067 247.2518 3.7289
 1 Ceres (A801 AA) 2010-Feb-10 00:00   2455237.5 ...    18.5029 250.0576 3.4415
 1 Ceres (A801 AA) 2010-Feb-20 00:00   2455247.5 ...    19.5814 252.7383 3.1451

The following fields are available for each ephemerides query:

>>> print(eph.columns)  
<TableColumns names=('targetname','datetime_str','datetime_jd','H','G','solar_presence','flags','RA','DEC','RA_app','DEC_app','RA_rate','DEC_rate','AZ','EL','AZ_rate','EL_rate','sat_X','sat_Y','sat_PANG','siderealtime','airmass','magextinct','V','surfbright','illumination','illum_defect','sat_sep','sat_vis','ang_width','PDObsLon','PDObsLat','PDSunLon','PDSunLat','SubSol_ang','SubSol_dist','NPole_ang','NPole_dist','EclLon','EclLat','r','r_rate','delta','delta_rate','lighttime','vel_sun','vel_obs','elong','elongFlag','alpha','lunar_elong','lunar_illum','sat_alpha','sunTargetPA','velocityPA','OrbPlaneAng','constellation','TDB-UT','ObsEclLon','ObsEclLat','NPole_RA','NPole_DEC','GlxLon','GlxLat','solartime','earth_lighttime','RA_3sigma','DEC_3sigma','SMAA_3sigma','SMIA_3sigma','Theta_3sigma','Area_3sigma','RSS_3sigma','r_3sigma','r_rate_3sigma','SBand_3sigma','XBand_3sigma','DoppDelay_3sigma','true_anom','hour_angle','alpha_true','PABLon','PABLat')>

The values in these columns are the same as those defined in the Horizons Definition of Observer Table Quantities; names have been simplified in a few cases. Quantities H and G are the target’s Solar System absolute magnitude and photometric phase curve slope, respectively. In the case of comets, H and G are replaced by M1, M2, k1, k2, and phasecoeff; please refer to the Horizons documentation for definitions.

Optional parameters of ephemerides() correspond to optional features of the Horizons system: airmass_lessthan sets an upper limit to airmass, solar_elongation enables the definition of a solar elongation range, max_hour_angle sets a cutoff of the hour angle, skip_daylight=True rejects epochs during daylight, rate_cutoff rejects targets with sky motion rates higher than provided (in units of arcsec/h), refraction accounts for refraction in the computation of the ephemerides (disabled by default), and refsystem defines the coordinate reference system used (ICRF by default). For comets, the options closest_apparition and no_fragments are available, which selects the closest apparition in time and limits fragment matching (73P-B would only match 73P-B), respectively. Note that these options should only be used for comets and will crash the query for other object types. Extra precision in the queried properties can be requested using the extra_precision option. Furthermore, get_query_payload=True skips the query and only returns the query payload, whereas get_raw_response=True returns the raw query response instead of the astropy table.

ephemerides() queries by default all available quantities from the JPL Horizons servers. This might take a while. If you are only interested in a subset of the available quantities, you can query only those. The corresponding optional parameter to be set is quantities. This parameter uses the same numerical codes as JPL Horizons defined in the JPL Horizons User Manual Definition of Observer Table Quantities. For instance, if you only want to query astrometric RA and Dec, you can use quantities=1; if you only want the heliocentric and geocentric distances, you can use quantities='19,20' (note that in this case a string with comma-separated codes has to be provided).

Orbital elements¶

elements() returns orbital elements relative to some Solar System body (location, referred to as “CENTER” in Horizons) and for a given epoch or a range of epochs (epochs) in the form of an astropy table. The following example queries the osculating elements of asteroid (433) Eros for a given date relative to the Sun:

>>> from astroquery.jplhorizons import Horizons
>>> obj = Horizons(id='433', location='500@10',
...                epochs=2458133.33546)
>>> el = obj.elements()
>>> print(el)
     targetname      datetime_jd  ...         Q                 P
        ---               d       ...         AU                d
 ------------------ ------------- ... ----------------- -----------------
 433 Eros (A898 PA) 2458133.33546 ... 1.782442696867877 642.9387350660577

The following fields are queried:

>>> print(el.columns)  
<TableColumns names=('targetname','datetime_jd','datetime_str','H','G','e','q','incl','Omega','w','Tp_jd','n','M','nu','a','Q','P')>

Optional parameters of elements() include refsystem, which defines the coordinate reference system used (ICRF by default), refplane which defines the reference plane of the orbital elements queried, and tp_type, which switches between a relative and absolute representation of the time of perihelion passage. For comets, the options closest_apparition and no_fragments are available, which select the closest apparition in time and reject fragments, respectively. Note that these options should only be used for comets and will crash the query for other object types. Also available are get_query_payload=True, which skips the query and only returns the query payload, and get_raw_response=True, which returns the raw query response instead of the astropy table.

Vectors¶

vectors() returns the state vector of the target body in cartesian coordinates relative to some Solar System body (location, referred to as “CENTER” in Horizons) and for a given epoch or a range of epochs (epochs) in the form of an astropy table. The following example queries the state vector of asteroid 2012 TC4 as seen from Goldstone for a range of epochs:

>>> from astroquery.jplhorizons import Horizons
>>> obj = Horizons(id='2012 TC4', location='257',
...                epochs={'start':'2017-10-01', 'stop':'2017-10-02',
...                        'step':'10m'})
>>> vec = obj.vectors()
>>> print(vec)
 targetname    datetime_jd    ...        range              range_rate
    ---             d         ...          AU                 AU / d
 ---------- ----------------- ... ------------------- ---------------------
 (2012 TC4)         2458027.5 ... 0.04293321014985829 -0.004080187136570278
 (2012 TC4) 2458027.506944444 ... 0.04290487450966027 -0.004080407283352606
 (2012 TC4) 2458027.513888889 ... 0.04287653879860514 -0.004080207494158422
 (2012 TC4) 2458027.520833333 ... 0.04284820593301306 -0.004079587874430119
        ...               ... ...                 ...                   ...
 (2012 TC4) 2458028.479166667 ... 0.03916439496172948 -0.004061365183433081
 (2012 TC4) 2458028.486111111 ... 0.03913618624590608 -0.004062675762467314
 (2012 TC4) 2458028.493055556 ... 0.03910796987365932 -0.004063569840032833
 (2012 TC4)         2458028.5 ... 0.03907974874463358 -0.004064045454467904
Length = 145 rows

The following fields are queried:

>>> print(vec.columns)  
<TableColumns names=('targetname','datetime_jd','datetime_str','H','G','x','y','z','vx','vy','vz','lighttime','range','range_rate')>

Similar to the other HorizonsClass functions, optional parameters of vectors() are get_query_payload=True, which skips the query and only returns the query payload, and get_raw_response=True, which returns the raw query response instead of the astropy table. For comets, the options closest_apparation and no_fragments are available, which select the closest apparition in time and reject fragments, respectively. Note that these options should only be used for comets and will crash the query for other object types. Options aberrations and delta_T provide different choices for aberration corrections as well as a measure for time-varying differences between TDB and UT time-scales, respectively.

How to Use the Query Tables¶

astropy table objects created by the query functions are extremely versatile and easy to use. Since all query functions return the same type of table, they can all be used in the same way.

We provide some examples to illustrate how to use them based on the following JPL Horizons ephemerides query of near-Earth asteroid (3552) Don Quixote since its year of Discovery:

>>> from astroquery.jplhorizons import Horizons
>>> obj = Horizons(id='3552', location='568',
...                epochs={'start':'2010-01-01', 'stop':'2019-12-31',
...                        'step':'1y'})
>>> eph = obj.ephemerides()

As we have seen before, we can display a truncated version of table eph by simply using

>>> print(eph)  
        targetname            datetime_str   ...  PABLon   PABLat
           ---                    ---        ...   deg      deg
-------------------------- ----------------- ... -------- --------
Don Quixote (1983 SA) 2010-Jan-01 00:00 ...   8.0371  18.9349
Don Quixote (1983 SA) 2011-Jan-01 00:00 ...  85.4082  34.5611
Don Quixote (1983 SA) 2012-Jan-01 00:00 ... 109.2959  30.3834
Don Quixote (1983 SA) 2013-Jan-01 00:00 ... 123.0777   26.136
Don Quixote (1983 SA) 2014-Jan-01 00:00 ... 133.9392  21.8962
Don Quixote (1983 SA) 2015-Jan-01 00:00 ... 144.2701  17.1908
Don Quixote (1983 SA) 2016-Jan-01 00:00 ... 156.1007  11.1447
Don Quixote (1983 SA) 2017-Jan-01 00:00 ... 174.0245   1.3487
Don Quixote (1983 SA) 2018-Jan-01 00:00 ... 228.9956 -21.6723
Don Quixote (1983 SA) 2019-Jan-01 00:00 ...  45.1979  32.3885

Please note the formatting of this table, which is done automatically. Above the dashes in the first two lines, you have the column name and its unit. Every column is assigned a unit from astropy units. We will learn later how to use these units.

Columns¶

We can get at list of all the columns in this table with:

>>> print(eph.columns)     
<TableColumns names=('targetname','datetime_str','datetime_jd','H','G','solar_presence','flags','RA','DEC','RA_app','DEC_app','RA_rate','DEC_rate','AZ','EL','AZ_rate','EL_rate','sat_X','sat_Y','sat_PANG','siderealtime','airmass','magextinct','V','surfbright','illumination','illum_defect','sat_sep','sat_vis','ang_width','PDObsLon','PDObsLat','PDSunLon','PDSunLat','SubSol_ang','SubSol_dist','NPole_ang','NPole_dist','EclLon','EclLat','r','r_rate','delta','delta_rate','lighttime','vel_sun','vel_obs','elong','elongFlag','alpha','lunar_elong','lunar_illum','sat_alpha','sunTargetPA','velocityPA','OrbPlaneAng','constellation','TDB-UT','ObsEclLon','ObsEclLat','NPole_RA','NPole_DEC','GlxLon','GlxLat','solartime','earth_lighttime','RA_3sigma','DEC_3sigma','SMAA_3sigma','SMIA_3sigma','Theta_3sigma','Area_3sigma','RSS_3sigma','r_3sigma','r_rate_3sigma','SBand_3sigma','XBand_3sigma','DoppDelay_3sigma','true_anom','hour_angle','alpha_true','PABLon','PABLat')>

We can address each column individually by indexing it using its name as provided in this list. For instance, we can get all RAs for Don Quixote by using

>>> print(eph['RA'])     
   RA
   deg
---------
50204
77158
85659
60021
44947
58967
32129
6992
11974
1066

This column is formatted like the entire table; it has a column name and a unit. We can select several columns at a time, for instance RA and DEC for each epoch

>>> print(eph['datetime_str', 'RA', 'DEC'])    
    datetime_str       RA      DEC
        ---           deg      deg
 ----------------- --------- --------
 2010-Jan-01 00:00 345.50204 13.43621
 2011-Jan-01 00:00  78.77158 61.48831
 2012-Jan-01 00:00 119.85659 54.21955
 2013-Jan-01 00:00 136.60021  45.8241
 2014-Jan-01 00:00 147.44947 37.79876
 2015-Jan-01 00:00 156.58967 29.23058
 2016-Jan-01 00:00 166.32129 18.48174
 2017-Jan-01 00:00  180.6992  1.20453
 2018-Jan-01 00:00 232.11974 -37.9554
 2019-Jan-01 00:00   16.1066 45.50296

We can use the same representation to do math with these columns. For instance, let’s calculate the total rate of the object by summing ‘RA_rate’ and ‘DEC_rate’ in quadrature:

>>> import numpy as np
>>> print(np.sqrt(eph['RA_rate']**2 + eph['DEC_rate']**2))
     dRA*cosD
------------------
18728612153883
337249029653798
520859656742434
679843758686584
775809055378625
874886005626538
183281978025435
295600636473737
84824546372009
952470898018017

Please note that the column name is wrong (copied from the name of the first column used), and that the unit is lost.

Units¶

Columns have units assigned to them. For instance, the RA column has the unit deg assigned to it, i.e., degrees. More complex units are available, too, e.g., the RA_rate column is expressed in arcsec / h - arcseconds per hour:

>>> print(eph['RA_rate'])  
 RA_rate
arcsec / h
----------
  72.35438
  -23.8239
  -20.7151
  -15.5509
   -12.107
  -9.32616
  -5.80004
  3.115854
  85.22719
  19.02548

The unit of this column can be easily converted to any other unit describing the same dimensions. For instance, we can turn RA_rate into arcsec / s:

>>> eph['RA_rate'].convert_unit_to('arcsec/s')
>>> print(eph['RA_rate'])
       RA_rate
      arcsec / s
----------------------
   0.02009843888888889
-0.0066177499999999995
 -0.005754194444444445
 -0.004319694444444445
-0.0033630555555555553
-0.0025905999999999998
-0.0016111222222222222
 0.0008655150000000001
  0.023674219444444443
  0.005284855555555556

Please refer to the astropy table and astropy units documentations for more information.

Hints and Tricks¶

Checking the original JPL Horizons output¶

Once either of the query methods has been called, the retrieved raw response is stored in the attribute raw_response. Inspecting this response can help to understand issues with your query, or you can process the results differently.

For all query types, the query URI (the URI is what you would put into the URL field of your web browser) that is used to request the data from the JPL Horizons server can be obtained from the HorizonsClass object after a query has been performed (before the query only None would be returned):

>>> print(obj.uri)  
https://ssd.jpl.nasa.gov/api/horizons.api?format=text&EPHEM_TYPE=OBSERVER&QUANTITIES=%271%2C2%2C3%2C4%2C5%2C6%2C7%2C8%2C9%2C10%2C11%2C12%2C13%2C14%2C15%2C16%2C17%2C18%2C19%2C20%2C21%2C22%2C23%2C24%2C25%2C26%2C27%2C28%2C29%2C30%2C31%2C32%2C33%2C34%2C35%2C36%2C37%2C38%2C39%2C40%2C41%2C42%2C43%27&COMMAND=%223552%22&SOLAR_ELONG=%220%2C180%22&LHA_CUTOFF=0&CSV_FORMAT=YES&CAL_FORMAT=BOTH&ANG_FORMAT=DEG&APPARENT=AIRLESS&REF_SYSTEM=ICRF&EXTRA_PREC=NO&CENTER=%27568%27&START_TIME=%222010-01-01%22&STOP_TIME=%222019-12-31%22&STEP_SIZE=%221y%22&SKIP_DAYLT=NO

If your query failed, it might be useful for you to put the URI into a web browser to get more information why it failed. Please note that uri is an attribute of HorizonsClass and not the results table.

Date Formats¶

JPL Horizons puts somewhat strict guidelines on the date formats: individual epochs have to be provided as Julian dates, whereas epoch ranges have to be provided as ISO dates (YYYY-MM-DD HH-MM UT). If you have your epoch dates in one of these formats but you need the other format, make use of astropy.time.Time for the conversion. An example is provided here:

>>> from astropy.time import Time
>>> mydate_fromiso = Time('2018-07-23 15:55:23')  # pass date as string
>>> print(mydate_fromiso.jd)  # convert Time object to Julian date
2458323.163460648
>>> mydate_fromjd = Time(2458323.163460648, format='jd')
>>> print(mydate_fromjd.iso) # convert Time object to ISO
2018-07-23 15:55:23.000

astropy.time.Time allows you to convert dates across a wide range of formats. Please note that when reading in Julian dates, you have to specify the date format as 'jd', as number passed to Time is ambiguous.

Keep Queries Short¶

Keep in mind that queries are sent as URIs to the Horizons server. If you query a large number of epochs (in the form of a list), this list might be truncated as URIs are typically expected to be shorter than 2,000 symbols and your results might be compromised. If your query URI is longer than this limit, a warning is given. In that case, please try using a range of dates instead of a list of individual dates.

Reference Frames¶

The coordinate reference frame for Horizons output is controlled by the refplane and refsystem keyword arguments. See the Horizons documentation for details. Some output reference frames are included in astropy’s coordinates:

Method	`location`	`refplane`	`refsystem`	astropy frame
`.vectors()`	`'@0'`	`'ecliptic'`	N/A	`'custombarycentricecliptic'`
`.vectors()`	`'@0'`	`'earth'`	N/A	`'icrs'`
`.vectors()`	`'@10'`	`'ecliptic'`	N/A	`'heliocentriceclipticiau76'`

For example, get the barycentric coordinates of Jupiter as an astropy SkyCoord object:

>>> from astropy.coordinates import SkyCoord
>>> from astropy.time import Time
>>> from astroquery.jplhorizons import Horizons
>>> epoch = Time('2021-01-01')
>>> q = Horizons('599', location='@0', epochs=epoch.tdb.jd)
>>> tab = q.vectors(refplane='earth')
>>> c = SkyCoord(tab['x'].quantity, tab['y'].quantity, tab['z'].quantity,
...              representation_type='cartesian', frame='icrs',
...              obstime=epoch)
>>> print(c)
<SkyCoord (ICRS): (x, y, z) in AU
    [(3.03483263, -3.72503309, -1.67054586)]>

Acknowledgements¶

This submodule makes use of the JPL Horizons system.

The development of this submodule is in part funded through NASA PDART Grant No. 80NSSC18K0987 to the sbpy project.

Reference/API¶

astroquery.jplhorizons Package¶

JPLHorizons¶

author: Michael Mommert (mommermiscience@gmail.com)

Classes¶

HorizonsClass([id, location, epochs, id_type])

A class for querying the JPL Horizons service.

Conf()

Configuration parameters for astroquery.jplhorizons.

Navigation

JPL Horizons Queries (astroquery.jplhorizons/astroquery.solarsystem.jpl.horizons)¶