From CedarWiki
Jump to: navigation, search

Jicamarca IS Radar



Contact Persons:

  • The general contact is Jorge (Koki) Chau
  • For drift data prior to March 1995 contact Wesley Swartz
  • For drift data after March 1995 contact Erhan Kudeki
  • For recent Faraday rotation data (Ne, Te, Ti) contact [ David Hysell]
  • For bistatic coherent E-region data (Ne) contact David Hysell

Web Pages:


  • For the Incoherent Scatter Radar: The Jicamarca Radio Observatory is a facility of the Instituto Geofisico del Peru and is operated with support from National Science Foundation Cooperative Agreements through Cornell University.
  • For the Bistatic Data (ISR and Paracas Receiver): The Jicamarca Radio Observatory and the Paracas receiver stations are facilities of the Instituto Geofisico del Peru and are operated with support from National Science Foundation Cooperative Agreements through Cornell University.

Also have JULIA (Jicamarca Unattended Long-term Investigations of the Ionosphere and Atmosphere) coherent scatter radar data of Vi=ExB F region drifts from the 150-km echoes.

Instrument/Model Description

The 50 MHz incoherent scatter radar at Jicamarca, Peru (11.95 S, 76.87 W; 520 m alt) has been operating since 1963. On day 359 of 2001 at the ground, the apex magnetic lat,lon were (0.58, -5.12) deg. The magnetic inclination and declination angles were 1.145 deg and 0.367 deg. The magnetic local time at 0 UT is ~1835 MLT.

There are three basic types of operating modes for the incoherent scatter radar:

  1. Faraday rotation to obtain electron densities (Ne) and electron (Te) and ion temperatures (Ti)
  2. the drift mode
  3. the bistatic coherent radar mode (E-region Ne above Paracas)

In the CEDAR Data Base, there are old Faraday rotation results between 1966 and 1969, and then recent results starting in 1996, although the technique was revived in the 1980's and the Database has electron densities only for June 26-29, 1984.

The drift mode, which finds perpendicular east and north (near vertical) ion drifts, has data from 1984 to the present. The analysis scheme to derive the drifts was changed starting in September of 1994 with a resulting increase in the data quality. Hard copies of time series plots of the older ion drifts are available upon request from NCAR.

The bistatic coherent radar experiment involves the 50 MHz transmitter at Jicamarca and a reciever at Paracas (13.85 S, 76.25 W), which is about 200 km south of Jicamarca. E region (~95-110 km) electron density data are available using a 4 microsecond (~0.6 km) pulse length from Mar 2004. This experiment can run concurrently with either the Faraday rotation or the drift mode.

The Drift Mode

Since 1994, incoherent scatter F region plasma drift measurements at Jicamarca have been implemented using a new signal processing approach [KINDAT=1900,1901,1910,1911] replacing the traditional pulse-to-pulse correlation method [KINST=1040,1050] after May 1995. The new method -- based on Doppler spectrum estimation and nonlinear least squares fitting to model spectra obtained from incoherent scatter theory -- improves the instrumental sensitivity remarkably under low S/N (signal-to-noise) conditions. With the new method, it has become possible to obtain very high quality drifts data at nearly all hours of the day throughout most F region heights. Altitude smoothing of the drifts data to reduce measurement noise is no longer necessary, and studies of the height variations of drifts can be preformed with much greater certainty than before Kudeki et al [1999]. Small-amplitude gravity wave oscillations have been detected at F region heights and a vertical circulation of the F region plasma has been observed in the post sunset period Kudeki and Bhattacharyya [1999].

The purpose of the East-West drift experiment is to estimate ExB plasma drift velocities perpendicular to the magnetic field from a pair of measured line of sight drift velocities. The directions of these vector components are zonal towards magnetic east (code 1241) and quasi upward, or perpendicular north to the B field (code 1252). The Jicamarca antenna uses crossed dipoles. In the following, UP and DOWN polarizations refer to the particular element used. In the experiment, reception antennas 1 and 2 which are formed by the NW and SE halves of Jicamarca array on UP polarization, are connected to receivers A and B. They receive returns from the West beam. This beam is formed by the entire array on UP polarization phased for westward beam formation. Antennas 3 and 4, formed by the NW and SE halves of Jicamarca array on DOWN polarization, connected to receivers C and D, respond to returns from the East beam formed by the entire array on DOWN polarization phased for eastward beam formation.

Each beam is about 2.5 degrees off the geomagnetic meridional plane (1.90 degrees to the west and 2.65 degrees to the east at 2001 changing slowly with time as the magnetic field over Jicamarca changes). In 2001, the West beam azimuth and elevation angles have been -64.99 deg and 86.73 deg, while the East beam angles were 50.61 deg and 88.07 deg. The beamwidth is about 0.5 degree. Sixty range gates specified by XPR control panel 0000xpan*csound are probed. A three baud barker code is used with a 6.666 millisecond Inter Pulse Period (IPP). The range resolution is 15 km. The lowest sampled height is 45 km. The altitude (code 110) is calculated using the largest offset from the zenith angle which is 3.27 degrees for the 2 beam experiment.

The radial (line-of-sight) velocities for the two beams are also given. The zonal drift is found from the difference of the 2 beams, while the 'vertical' drift (Vperp North) is found from the sum of the beams. Most of the time two beams have been used. However, there are times where only a perpendicular to B beam is used. Here, only direction 7, the West beam is used, where the azimuth angle is -34.80 and the elevation angle is 88.44 deg such that at some height the radial velocity is perpendicular to the B field (code 1252). In this case, just the 'vertical' drift is obtained (code 1252 = code 1272). The East beam, direction 8 and the zonal drift are thus missing (-32767).

Codes 3200 and 32001 refer to reduced chisquares for the west and east beam incoherent scatter spectral fits. When spectral fitting to the incoherent scatter model fails, the values are set to -32767.

The powers of both coherent and incoherent echoes are given in terms of the log10 of the S/N ratio plus one. (Adding the one makes the log10 0 when the S/N is small.) The drift from incoherent echoes represents the ExB ion drift in the F region. Coherent echoes are caused by satellites, space debris, spread F, 150-km irregularities and equatorial electrojet. The incoherent echoes are processed first, but drifts from weak spread F coherent echoes are also from ExB drifts, although drifts from strong spread F coherent echoes are NOT ExB drifts as shown in Plates 1-3 of [2]. Therefore, drifts are found first from suitable incoherent echoes, and second from suitable coherent echoes, although the drifts from strong coherent echoes are not ExB, and DRIFTS BELOW 200 KM SHOULD NOT BE COMPLETELY TRUSTED. The signal is often saturated below 200 km, and arbitrarily split into incoherent and coherent parts. Velocities from the incoherent power can be found to great precision (small error bars), but in general it is not clear what is being measured below 200 km. Around 45 km in the MST region, the neutral coherent turbulence leads to echoes which are mis-identified as incoherent echoes, but these are not ExB drifts. Sometimes the ExB drifts are good down to 150 km, but INTERPRETATION OF THE DRIFTS BELOW 200 KM SHOULD ALWAYS BE IN PARTNERSHIP WITH THE CONTACT PEOPLE. The contact people should be notified about other data use also, in compliance with the 'Rules of the Road'.

There are two different types of data records. The first type (KINDAT=1910) contains information for every range gate sampled, up to and including the 58th range gate. Range gates 59 and 60 have not been included since returns from these range gates cannot be Barker decoded. All range gates are 15 km wide. Every data value contained in this data file spans a 5 minute interval and has been obtained by integrating the data for 5 minutes. The data are processed in the frequency domain. Returns are classified as coherent or incoherent echoes using a dual power/coherence filter and are processed differently. Least squares estimation is used for incoherent scatter measurements whereas first moment estimation is used to handle coherent echoes. A rigorous description of the incoherent scatter velocity and error estimation technique is in Kudeki et al [1999], along with details of the data processing techniques used.

In KINDAT=1911, the 'vertical' and eastward drifts are averaged over the 24 ranges between 218 km and 577 km. There is a total F-region average over the altitude averaging interval (code 115) of 360 km, as well as six averages over 4 range gates with an effective height resolution (code 115) of 60 km. Weighted averages are computed from the velocity values and error bars determined FROM INCOHERENT ECHOES ONLY in KINDAT=1910 records, thus eliminating velocities from strong spread F coherent echoes that do not represent ExB drifts. The range (code 120) and altitude (code 110) represent midpoints of the various averaging regions.

The traditional pulse-to-pulse correlation method of finding the drifts

In order to determine when the results are good, it is neccessary to look at the correlation profiles of the two beams. Values greater than about 0.3 usually indicate either spread-F or satellite contamination, even though the error bars may be extremely small. Also, if the correlation is very small (~<0.05) in one beam, then the perpendicular ion drift velocities are not reliable. Finally, if the power profiles in the two beam directions are very different, then the velocities are unreliable. The power profiles are raw un-normalized backsacttered power from each beam, but they cannot be used to deduce the electron density sine the beams are perpendicular to the Earth's magneitc field, and so the correlation times in the medium are much longer and become very strong when any field aligned irregularities are present. There can also be Faraday rotation effects that cannot be removed with this particular experiment.

The Faraday Rotation Mode

The Faraday experiment was performed early in the history of the observatory but was discontinued until the mid 1980s, when the computers of the time became sufficiently powerful to tabulate the necessary lag products in real time. It was found that the quality of the experiment could be improved considerably by accounting for and subtracting experimental biases introduced by polarization crosstalk, imperfect quadrature detection, and other subtle effects. These corrections are now routinely made to the data. In the mid 1990s, numerous flaws in the Jicamarca antenna array caused by corrosion were repaired, and the quality of the Faraday data improved still further. In late 1997, the experiment was upgaded with the introduction of a modernized data acquisition system at Jicamarca. Data are now collected with more lags than before and with a full-time duty cycle.

The typical online incoherent integration time for the Faraday experiment is 1 minute. The data made available here have been further integrated during post-processing for 15 minutes. Data with finer time resolution may be obtained from the contact person, David Hysell (

An estimate of the absolute electron density is derived from the measured polarization of the scattered signal. This estimate is used to calibrate the total scattered power so as to yield a robust estimate of the uncorrected electron density at altitudes between approximately 100-1000 km. The calibration is performed by minimizing the RMS discrepancy between the power profile and the Faraday profile. The range of altitudes used to perform the calibration is determined dynamically and changes with geophysical conditions. The power profiles are calibrated but are not corrected for unequal electron and ion temperatures. This is because the data sometimes represent coherent echoes from plasma irregularities, for which such a correction makes no physical sense. Users are free to make the correction themselves in portions of the data uncontaminated by coherent echoes. The formula is:
Ne(corrected) = Ne(uncorrected)*0.5*(1+alpha**2+Te/Ti)*(1+alpha**2)
where alpha**2=7.654e+5 Te/(Ne(corrected)*lambda**2) with Ne's in m-3, Te, Ti, in K, lambda (radar wavelength) in m. For Jicamarca, lambda = 6.0 m precisely. Note that alpha needs the corrected Ne, but using the uncorrected Ne here is a relatively good first approximation.

The autocorrelation function of the scattered signal is also measured. Because of the strong clutter introduced by plasma irregularities in the equatorial electrojet, a double-pulse pattern is used to assemble measurements of the ACF one lag at a time. The length of each pulse is typically 15 km, and the lags of the ACF typically range from 0 to 2 ms in uniform steps of 0.2 ms. Such a technique minimizes the effect of the electrojet clutter, which enters through sidelobes in the antenna pattern. However the accumulation of lag product estimates takes place relatively slowly. It is therefore often impractical to fit the ACFs for more than two parameters simultaneously. Between 0600-2000 LT (1100-0100 UT), the electron temperature Te (code 560) and the ion temperature Ti (code 550) are the fit parameters. Outside of this window, the electron and ion temperatures are assumed to be equal and the parameters fit are Te=Ti and the percentage concentration of H+ (code 660). Note that the failure to allow for minor constituents in the fitting of daytime data produces artificially high ion temperatures at altitudes above the F peak.

Because of the effect of electron Coulomb collisions quantified by Aponte et al (2002), measured electron temperatures are anomalously small when the radar wavevector is directed very close to perpendicular to B. For these experiments, the antenna beam was directed approximately 4.5 degrees off perpendicular. This antenna position appears to mitigate the problem while preserving a useful antenna radiation pattern. The 1960s Faraday rotation temperature ratios Te/Ti, were arbitrarily multiplied by a constant so that Te was always greater than Ti. Complete correction of the electron temperature for Coulomb collisions can be done with all data for which the complete spectra still exist, so some data will be corrected.

The Bistatic Coherent Radar Mode

Since March 2004, data from the Bistatic Coherent Radar Experiment at Jicamarca Radio Observatory have been derived. The transmitter is located at Jicamarca (11.98 deg S, 76.87 deg W) and the receiver is located at Paracas (13.85 deg S, 76.25 deg W) about 200 km south. Because the receiver is not at Jicamarca, a new kinst (11) was assigned to these data. The bistatic coherent radar mode can run concurrently with either the Faraday rotation or the drift mode. Electron densities in the E region (~95-110km) can be found, mostly in the daytime because the electron density (signal strength) is largest then.

Due to the presence of fluctuations above the thermal level, the conventional incoherent scatter radar technique cannot be applied to probe the electrojet region. The radar experiment, therefore, relies on coherent scattering from the electrojet medium instabilities. The theoretical foundation for the radar experiment is the quasilongitudinal approximation of wave propagation in a cold magnetoplasma. The theory relates mathematically electron density to phase angle difference (Faraday angle) between the left and right circularly polarized waves. The significance of the bistatic radar geometry are:

  1. a substantial amount of Faraday angle can be gathered
  2. Bragg's scattering vector is normal to the geomagnetic field, which is a condition for coherent scattering from field aligned irregularities.

Electron density is extracted from the equation shown below by finite differencing the Faraday angle with respect to altitude. Interference from power data have been removed and the phase has been cleaned before the electron density profile is computed.

The formula that relates electron density to Faraday angle is:
Where d(thetha) is the differential phase angle, f is the radar frequency in MHz, B is the geomagnetic field in Gauss, dh is differential altitude in meters, gamma is the angle between the ray and the geomagnetic field, phi is the zenith angle, and Ne is electron density.

The applicability of the theory to the E-region equatorial ionosphere was tested in a trial bistatic radar experiment in September of 2000. The current bistatic experiment is an improvement of the original version in order to maximize sensitivity and thereby enhance signal-to-noise ratio. Improvement of the following components of the bistatic radar system have been made: transmitter and receiver were swapped (in the original (2000) radar experiment the transmitter was at Paracas and the receiver was in Jicamarca), antenna arrays were enlarged, and pulse mode were modified (the present bistatic experiment transmits Barker codes).

For more information, consult Hysell and Chau (2001) and Ratcliffe (1959). There are summary plots for bistatic data at Jicamarca at, in addition to the plots below.

The Summary Plots

The summary plots are quite different for the different modes used at Jicamarca. Between 1966 and 1969 there are 2-day image plots of the electron density, and the electron and ion temperature. These are plotted as a function of height and time where the time bin is 40 minutes and the height bin is 10 km for the electron density and 40 km for the temperatures. The electron density is calibrated and corrected for temperature effects. These plots are resumed in 1997, but here the electron density is not corrected for temperature effects, and the uncalibrated density is also available. Starting in 1984, the drift mode was used extensively, so the summary plots are of the ion drift in the perpendicular east and north (near vertical) directions in 4 separate 2-day plots per page. The summary plots are divided into Faraday rotation mode and drift mode. Other web sites for Jicamarca plots are listed below.


[/commun/cedartuts/2002/aponte-02-Te-Ti.pdf N. Aponte, M. P. Sulzer and S. A. Gonzalez, "Correction of the Jicamarca Te/Ti ratio problem: Verifying the effect of electron Coulomb collisions on the incoherent scatter spectrum"], 2002 CEDAR Workshop, June 16-21, Longmont, Colorado.
Farley, D. T., Faraday rotation measurements using incoherent scatter, Radio Science, 4, 143, 1969a
Farley, D. T. , Incoherent scatter correlation function measurements, Radio Science, 4, 935, 1969b.
Hysell, D.L. and J. L. Chau, Inferring E region electron density profiles at Jicamarca from Faradar rotation of coherent scatter, Journal of Geophy. Research, 106 (A12), 30,371-30,380, 2001
Kudeki, E., S. Bhattacharyya, and R. F. Woodman, A new approach in incoherent scatter F region ExB drift measurements at Jicamarca, J. Geophys. Res., 104, 28145-28162, 1999.
Kudeki, E. and S. Bhattacharyya, Postsunset vortex in equatorial F-region plasma drifts and implications for bottomside spread-F, J. Geophys. Res., 104(A12), 28,163-28,170, 1999.
Kelley, M.C., "The Earth's Ionosphere - Plasma Physics and Electrodynamics", Academic Press, Inc, 1989.
Pingree, J. E., Incoherent scatter measurements and inferred energy fluxes in the equatorial F region ionosphere, Ph.D. dissertation, Cornell Univ., Ithaca, NY, 1990.
Ratcliffe, J. A., Magneto-ionic theory and its applications to the ionosphere, Cambridge University Press, New York, 1959

Old Faraday Rotation Summary Plots for Jicamarca IS Radar (1966-1969)

These data came from a tape deposited with the World Data Center in Boulder and converted to CEDAR Database format by Roy Barnes. The Te/Ti ratio was arbitrarily multiplied by some (unknown?) constant so that Te/Ti>1 in the original tape.

Revived Faraday Rotation Summary Plots for Jicamarca IS Radar (1980s)

New Faraday Rotation Summary Plots for Jicamarca IS Radar (1996-present)

Old Drift Summary Plots for Jicamarca IS Radar (1984-1995)

These are plots of the drifts ~300 km (295-305km).

New Revised Drift Summary Plots for Jicamarca IS Radar (1994-present)

Eight periods in 1994-1996 were replaced. The F-region averge drifts (218-578 km) only included incoherent echoes (ExB), while the drifts at ~300 km (295-305 km) include the coherent spread F echoes. Thus times when there are drifts at ~300 km but not in the average are usually during times of spread F. Starting in September 2003, the summary plots were only of the drifts around 300 km (295-305 km), since the F-region drifts are similar but much cleaner and without any coherent echoes.

Summary Plots for Jicamarca-Paracas Bistatic Coherent Radar (2004-present)

The electron density above the reciever at Paracas between 90 and 120 km is plotted over 2-day intervals in bins of 5 min and 1 km. The data in 2005 for Oct 11 and Nov 1 were withheld from the database for further processing.

Other I.S. Radars

Daily Listing for IS/HF Radars

-Revised 11 Jun 2007 by