Special Issues in low frequency solar radio observation.
A. KERDRAON Observatoire de Paris

Dynamic range

Interferences

Calibration

Ionospheric disturbances

Special Issues at Low frequencies: Dynamic range

Bursts expected in the range 10^-2 to 10⁴SFU

Dynamic in RF channels (before correlation)

Wide bandwidth channels: interference is the main problem, the power level may be dominated by interferences ~constant over a wide frequency band.

Narrow bandwidth channels:

Assuming T_sys = noise + sky background +

For a large antenna (50 m²), dynamic range is 300 - 10⁶ K (40 db)

For a small antenna (1m² - dipole), it is ~23 db

Special Issues at Low frequencies: Dynamic (ctd.)

Power variations may be > 20 db/sec.

Implications:

High dynamic electronics needed

Automatic attenuators or very fast AGC may be be needed in narrow band channels.

Correlation is better made by 1 bit correlator +power measurement (or many bits correlator - challenging ?)

If a spectrum is computed before correlation, it will have the dynamic of narrow channels (neglecting interferences)

Note: Dynamic on images (T_b) may be much higher

Special Issues at Low frequencies: Interferences (ctd.)

Low frequencies ( < 2 GHz) are crowded:

Very powerful transmitters (TV and FM) (emitters ~100 kW)

Low power transmissions and GSM (emitters of 10 - 100 W)

Satellites

Protected (for radioastronomy) bands are very few and difficult to keep free of interferences:

151, 327, 408, 610, 1400 MHz… with bandwidth from 3 to a few ten MHz, in principle .

Satellites emissions may occur very close (149.9 MHz) to astronomy band

Antennas provide usually a poor rejection ( 10 db)

Special Issues at Low frequencies: Interferences

150 - 250 MHz band Nançay (interference survey antenna)

Wide band example

Special Issues at Low frequencies: Interferences

150 - 152 MHz band Nançay (interference survey antenna)

Narrow band examples

Special Issues at Low frequencies: Interferences(ctd.)

Strong interferences: may be the main challenge:

Power of TV is 80 db above quiet sun emission at 200 MHz (Nançay)

A stop band filter may be required for each strong emitter

The power is very high in the wide band electronic

Very high dynamic components are needed in order to avoid intermodulation (which can generate thousands of parasitic lines)

That level may be to high for digital filtering and interference excision.

Site quality is very important.

Special Issues at Low frequencies: Interferences(ctd.)

Mean level interferences ( <40db above quiet sun)

At the lower end of the band (< 300 MHz):

It is very difficult to find a 1 Mhz clean band

Interferences have a small bandwidth ( 5-20 kHz)

This imply:

Small observing bandwidth, very good filters.

Digital interference excision will require high dynamic, high resolution spectra and filters.

There are (at present) less problems above 300 MHz…

Don ’t forget internal interferences, mainly from high speed digital electronics.

Special Issues at Low frequencies: Calibration

General

Based on stability (over days and weeks)

Stability is achieved by buried electronics and transmission lines

Stability allows observation of calibrators outside sun observing time.

Calibration provides complex gains for each antenna and frequency valid for days and weeks.

Special Issues at Low Frequencies: Calibration (cdt.)

Calibrators

Very few (5 ?) strong calibrators

All of them have structure above 1000 l.

Calibrators are not polarized

No signal on X-Y correlators: corrected by rotating some antennas by 45 degrees during calibrations.

Calibrators flux decreases strongly with increasing frequency

Special Issues at Low frequencies: Calibration

Nançay present solution:

Uses only Cygnus A radio-galaxy, which gives a strong signal at 5000 l and 500 MHz

Uses a 2 components model, and fits fringes

Accuracy could be better for long baselines

Starting a new procedure based on redundancies and short baselines: antennas are calibrated step by step, and almost no source model is needed.

Special Issues at Low frequencies: Calibration

Conclusions

Phase/gain stability is mandatory

Self-cal on calibrators should be the best solution

The array design is important:

Redundancies are useful

Antennas involved only in long baselines are difficult to calibrate.

Special Issues at Low frequencies: Ionosphere

Ionosphere at 164 MHz

Very severe case (includes some distorsion)

In most cases: smaller motion and no distorsion.

Likely to occur at low site angle

Special Issues at Low Frequencies: Ionosphere

Generalities

Density inhomogeneities due to Travelling Ionospheric Disturbances (TIDs) may affect radio observations at dam to dm wavelength.

TIDs most often due to gravity waves, sometimes to other phenomena (including magnetosphere).

Effects are proportionnal to f^-2

Gravity waves are neutral atmosphere phenomenon , which couples through collisions to electrons and ions

Special Issues at Low Frequencies: Ionosphere

Observations

Many techniques, including radio interferometry

Large database in Nançay (C. Mercier), sun and other sources, day and night

Very large database in Los Alamos, made with a 9 antennas VLBI array and satellites beacons (A. Jacobson)

Some observations at VLA (cosmic radio sources)

Special Issues at Low frequencies: Ionosphere

Ionospheric gravity wave schema

~horizontal wave plan

~horizontal energy propagation

~horizontal atmosphere motions

Amplitude increases with decreasing density

Special Issues at Low Frequencies: Ionosphere

Properties of gravity waves:

Auroral origin:

T ~1h, l_horiz ~1000 km, Vf > 400 m/sec southward, usually at night.

Meteorological origin:

T = 15-60 mn, l_horiz = 150-250 km, Vf ~100 m/sec, usually southward

Mostly during the day, quickly damped over ~1000 km.

Origin (local) is strongly site dependant

TEC gradient 2. 10¹¹ m^-3 typically

Electronic density changes (1-10%) occur in F2 layer( ~250 km)

Special Issues at Low Frequencies: Ionosphere

Gravity effects on observations

Due to first and second derivative of TEC

Very dependent on geometry: maximum when line of sight is along wave fronts, small when perpendicular to wavefronts, or //B.

Typical image motion +/- 1 arcmin at 169 MHz, typical timescale: 30 mn.

At least 10x in worst conditions (+ some image distorsion)

10x less at nightime.

Focusing may occur at low frequencies, exceptionnal cases up to 50 MHz at low elevation.

Special Issues at Low Frequencies: Ionosphere

Conclusions:

In order to minimize most effects, it is important to avoid low elevation angles (high latitudes, sunrise and sunset)

Remaining effects:

Distortion should be corrected by self-cal (but who can self-cal quiet sun emission ?).

Image shifts should be detected by images correlation, or observation of stable small source (Noise Storms), and corrected within less than 1 beamwidth.

For transient bursts (ex. Type II/III/IV/V bursts) in the absence of Noise Storms, there is no trivial position correction.


	Dynamic range
	Interferences
	Calibration
	Ionospheric disturbances


Bursts expected in the range 10^-2 to 10⁴SFU
Dynamic in RF channels (before correlation)
	Wide bandwidth channels: interference is the main problem, the power level may be dominated by interferences ~constant over a wide frequency band.

	Narrow bandwidth channels:
		Assuming T_sys = noise + sky background +

		For a large antenna (50 m²), dynamic range is 300 - 10⁶ K (40 db)
		For a small antenna (1m² - dipole), it is ~23 db


Power variations may be > 20 db/sec.
Implications:
	High dynamic electronics needed
		Automatic attenuators or very fast AGC may be be needed in narrow band channels.
		Correlation is better made by 1 bit correlator +power measurement (or many bits correlator - challenging ?)
	If a spectrum is computed before correlation, it will have the dynamic of narrow channels (neglecting interferences)
Note: Dynamic on images (T_b) may be much higher


	Low frequencies ( < 2 GHz) are crowded:
		Very powerful transmitters (TV and FM) (emitters ~100 kW)
		Low power transmissions and GSM (emitters of 10 - 100 W)
		Satellites
		Protected (for radioastronomy) bands are very few and difficult to keep free of interferences:
		151, 327, 408, 610, 1400 MHz… with bandwidth from 3 to a few ten MHz, in principle .
		Satellites emissions may occur very close (149.9 MHz) to astronomy band
		Antennas provide usually a poor rejection ( 10 db)


	150 - 250 MHz band Nançay (interference survey antenna)
	Wide band example


	150 - 152 MHz band Nançay (interference survey antenna)
	Narrow band examples


Strong interferences: may be the main challenge:

	Power of TV is 80 db above quiet sun emission at 200 MHz (Nançay)
		A stop band filter may be required for each strong emitter
		The power is very high in the wide band electronic
		Very high dynamic components are needed in order to avoid intermodulation (which can generate thousands of parasitic lines)
		That level may be to high for digital filtering and interference excision.

Site quality is very important.


Mean level interferences ( <40db above quiet sun)
	At the lower end of the band (< 300 MHz):
		It is very difficult to find a 1 Mhz clean band
		Interferences have a small bandwidth ( 5-20 kHz)
	This imply:
		Small observing bandwidth, very good filters.
		Digital interference excision will require high dynamic, high resolution spectra and filters.
	There are (at present) less problems above 300 MHz…

Don ’t forget internal interferences, mainly from high speed digital electronics.


General

	Based on stability (over days and weeks)
		Stability is achieved by buried electronics and transmission lines
	Stability allows observation of calibrators outside sun observing time.
	Calibration provides complex gains for each antenna and frequency valid for days and weeks.


Calibrators
	Very few (5 ?) strong calibrators
	All of them have structure above 1000 l.
	Calibrators are not polarized
		No signal on X-Y correlators: corrected by rotating some antennas by 45 degrees during calibrations.
	Calibrators flux decreases strongly with increasing frequency


	Nançay present solution:

		Uses only Cygnus A radio-galaxy, which gives a strong signal at 5000 l and 500 MHz
		Uses a 2 components model, and fits fringes
		Accuracy could be better for long baselines
		Starting a new procedure based on redundancies and short baselines: antennas are calibrated step by step, and almost no source model is needed.


Conclusions

	Phase/gain stability is mandatory
	Self-cal on calibrators should be the best solution
	The array design is important:
		Redundancies are useful
		Antennas involved only in long baselines are difficult to calibrate.


	Ionosphere at 164 MHz
	Very severe case (includes some distorsion)
	In most cases: smaller motion and no distorsion.
	Likely to occur at low site angle


	Generalities

		Density inhomogeneities due to Travelling Ionospheric Disturbances (TIDs) may affect radio observations at dam to dm wavelength.
		TIDs most often due to gravity waves, sometimes to other phenomena (including magnetosphere).
		Effects are proportionnal to f^-2
		Gravity waves are neutral atmosphere phenomenon , which couples through collisions to electrons and ions


	Observations

		Many techniques, including radio interferometry
		Large database in Nançay (C. Mercier), sun and other sources, day and night
		Very large database in Los Alamos, made with a 9 antennas VLBI array and satellites beacons (A. Jacobson)
		Some observations at VLA (cosmic radio sources)


	Ionospheric gravity wave schema

		~horizontal wave plan
		~horizontal energy propagation
		~horizontal atmosphere motions
		Amplitude increases with decreasing density


Properties of gravity waves:
	Auroral origin:
		T ~1h, l_horiz ~1000 km, Vf > 400 m/sec southward, usually at night.
	Meteorological origin:
		T = 15-60 mn, l_horiz = 150-250 km, Vf ~100 m/sec, usually southward
		Mostly during the day, quickly damped over ~1000 km.
		Origin (local) is strongly site dependant
		TEC gradient 2. 10¹¹ m^-3 typically
		Electronic density changes (1-10%) occur in F2 layer( ~250 km)


Gravity effects on observations

	Due to first and second derivative of TEC
	Very dependent on geometry: maximum when line of sight is along wave fronts, small when perpendicular to wavefronts, or //B.

	Typical image motion +/- 1 arcmin at 169 MHz, typical timescale: 30 mn.
		At least 10x in worst conditions (+ some image distorsion)
		10x less at nightime.
	Focusing may occur at low frequencies, exceptionnal cases up to 50 MHz at low elevation.


Conclusions:
	In order to minimize most effects, it is important to avoid low elevation angles (high latitudes, sunrise and sunset)
	Remaining effects:
		Distortion should be corrected by self-cal (but who can self-cal quiet sun emission ?).
		Image shifts should be detected by images correlation, or observation of stable small source (Noise Storms), and corrected within less than 1 beamwidth.
		For transient bursts (ex. Type II/III/IV/V bursts) in the absence of Noise Storms, there is no trivial position correction.