This document describes format of the custom data files which can uploaded to the VirES for Swarm service and visualized together with the Swarm, models and auxiliary data offered by the service.
The uploaded data file is expected to contain a multi-variate time series of observations (records). Each of the records shall have a time-stamp and geo-location (ITRF latitude, longitude and radius) and any number of additional named variables.
The ITRF latitude and longitude are mandatory. The radius is optional, but when not provided, the geomagnetic models and magnetic coordinates requiring exact position cannot be evaluated. It is, therefore, recommended to always provide the complete set of coordinates, including the radius.
The additional variables can be of a scalar or vector numerical type.
The vector variables will be decomposed to its scalar components.
While any input variable is allowed, there are special variables which are interpreted similarly as the equivalent Swarm product variables, e.g., magnetic model residuals can be calculated for
The time-stamps do not need to be ordered.
The input data are internally converted in the CDF data format. The timestamps are internally converted to
CDF_EPOCH (UTC, 1ms precision).
Details of the data formats are described in the following sections.
Each separate CDF file variable is expected to have the same amount of records as the time-stamp. If not, the variable is ignored.
The CDF variables are described in the following table.
|Timestamp||yes||Time of observation||UTC||1||
|Latitude||yes||Position in ITRF – Geocentric latitude||deg||1||
|Longitude||yes||Position in ITRF – Geocentric longitude||deg||1||
|Radius||no||Position in ITRF – Radius (required to calculate QD-coordinates, MLT, and magnetic models)||m||1||
|F||no||Magnetic field intensity (required to calculate model residuals)||nT||1||
|B_NEC||no||Magnetic field vector, NEC frame (required to calculate model residuals)||nT||3||
|any||no||arbitrary custom variable||any||any||any CDF number data type|
The CSV file uses comma as a delimiter and it is required to have a header (first line) defining the names of the records' fields. Each record is is required to have the same number of values as the header.
Each records is required to have a time-stamp defined as
MJD2000variable) defined as a decimal number of days since
The Time-stamps' UTC offsets are accepted and interpreted (internally converted to UTC). Time-stamps without a UTC offset are interpreted as UTC times, e.g.,
2019-06-12T09:35:27.123Z are therefore the same times.
MJD200 are present, the
Timestamp is used as the record time-stamp.
Vector fields are encoded as semicolon
; separated list of values enclosed by curly brackets
In the special case of the
B_NEC vector variable, the
B_NEC vector is automatically composed from the
B_C scalar component if
B_NEC is missing and all three
B_C variables are present in the CSV input data.
||Time of observation||UTC||1||RFC-3339|
||Time of observation||MJD2000||1||float|
|Latitude||yes||Position in ITRF – Geocentric latitude||deg||1||float|
|Longitude||yes||Position in ITRF – Geocentric longitude||deg||1||float|
|Radius||no||Position in ITRF – Radius (required to calculate QD-coordinates, MLT, and magnetic models)||m||1||float|
|F||no||Magnetic field intensity (required to calculate model residuals)||nT||1||float|
|B_NEC||no||Magnetic field vector, NEC frame (required to calculate model residuals), automatically composed from
|B_N||no||Magnetic field vector's northing component||nT||1||float|
|B_E||no||Magnetic field vector's easting component||nT||1||float|
|B_C||no||Magnetic field vector's radial component (center oriented)||nT||1||float|
|any||no||arbitrary custom variable||any||any||integer or float|
The float means a decimal representation of the double-precision floating-point number, e.g.,