Solar Flux And Geomagnetic Index — Robotics/Celestial Phenomena

Solar Flux and Geomagnetic Index

Robotics / Celestial Phenomena

The solar radio flux F10.7 and the geomagnetic indices Ap for a UTC instant, read out of the space-weather table – the three drivers the NRLMSISE-00 atmosphere takes, in the shapes it takes them. With u the instant in days since 1 October 1957 and d = floor(u):

• F107 = F10.7 of the previous day, F(d − 1)
• F107a = the 81-day centred average of F10.7 at d
• aph = the daily Ap of day d, the 3-hour ap of the current slot floor(8u) and of the three before it, and the means of the eight 3-hour values 12–33 and 36–57 hours before

The table (the CelesTrak consolidated file as MATLAB R2026a ships it, aeroSpaceWeatherData.mat) has daily rows from 1 October 1957 to 16 March 2025, then monthly predicted rows to 1 October 2041 that carry F10.7 and no geomagnetic data. F10.7 is interpolated linearly between rows, y₁ + (x − x₁)(y₂ − y₁)/(x₂ − x₁), which reads every daily row exactly. Outside the table the extrapolation settings decide:

• F10.7, before 1 October 1957 and from 1 November 2041 (the end of the last row's month): the first or last row (None - clip), a Constant, or the Least squares fit 128.2351780622538 + 54.3285872213434·cos(0.0015460708364·t + 0.2429462096495·sin(0.0015563188188·t)), t in days since 1 October 1957. Between the last row and 1 November 2041 the last row holds. There the 81-day average is no longer the table's: it is the mean of F10.7 over d − 40 to d + 40, summed in order.
• Ap, outside the daily rows (the monthly ones included): the nearest daily row (None - clip, keeping the 3-hour slot, so before the table the first day's eight values repeat) or a Constant, slot by slot.

Ports

• year – the UTC year, a scalar [1,1]; a fraction is dropped.
• doy – the day of year, 1 for 1 January, a scalar [1,1]. A fraction counts as part of a day, and a value past the year's end carries into the next.
• sec – the UTC seconds of the day, a scalar [1,1]; 86400 or more carries into the next day.
• F107a – the 81-day centred average of F10.7, in solar flux units (10⁻²² W m⁻² Hz⁻¹), [1,1].
• F107 – the previous day's F10.7, in the same units, [1,1].
• aph – the seven magnetic indices above, [7,1]: the NRLMSISE-00 Atmosphere block's aph input.

Parameters

• F10.7 Extrapolation Method – how F10.7 and its average are carried outside the table:
  • None - clip (default) – the first row before it, the last row after it.
  • Constant – the F10.7 Extrapolation Value.
  • Least squares fit – the trigonometric fit above.
• F10.7 Extrapolation Value – the constant for Constant, a scalar of zero or more (default 150).
• Magnetic Index Extrapolation Method – how the indices are carried outside the daily rows:
  • None - clip (default) – the nearest daily row, slot for slot.
  • Constant – the Magnetic Index Extrapolation Value.
• Magnetic Index Extrapolation Value – the constant for Constant, a scalar of zero or more (default 4).
• Sampling Time (s) – zero or less inherits the solver's rate; a positive value runs the block at that period.

Code export

Seven targets: Python, MATLAB, Java, Rust, C, C++ and PLC Structured Text. Each carries the whole table – about 270000 numbers, the 3-hour indices as integers – and the lookup, with the extrapolation settings baked in, so every target answers what the block's own simulation answers. Java receives the table as text parsed once, because array literals that long exceed a Java method's size limit.

The three HDL targets (VHDL, Verilog, SystemVerilog) are not offered – an export to one of them stops and names this block – because the UTC seconds (to 86400) cannot sit on a Q16.16 port, whose range ends at ±32768.

Simulink bridge

Import and export, mapped to Aerospace Blockset's aerolibcelestial/Solar Flux and Geomagnetic Index. Four configuration values cross, one for one: F10.7 Extrapolation Method → F107ExtrapMethod, F10.7 Extrapolation Value → F107ExtrapValue, Magnetic Index Extrapolation Method → MagneticIndexExtrapMethod and Magnetic Index Extrapolation Value → MagneticIndexExtrapValue. Simulink's third index method, IGRF, is not offered here, and an import that names it is reported. Two Simulink parameters are always implied: SpaceWeatherDataFile is aeroSpaceWeatherData.mat, the table this block carries, and InputSpec is Year, day, sec, since its Julian date setting gives the Simulink block one input rather than three and no configuration here can change the ports. The Simulink block defines no SampleTime parameter, measured, so the rate stays on the ICore side. The global Sampling Time (s) → SampleTime pair therefore does not cross.

Notes

• Algebraic and stateless: the outputs depend on this sample's instant alone.
• Not linear, so the block carries no state space and model reduction correctly reports it as unmergeable.
• The table is a snapshot, republished daily by CelesTrak; dates past it read the extrapolation until the block's table is refreshed.
• Verified against R2026a's block: the arithmetic of every region – the daily rows, the monthly interpolation, both ends and all three flux settings – agrees bit for bit, the least-squares cosine to within a unit in the last place.