From CedarWiki
Jump to: navigation, search

Arequipa Fabry-Perot


Contact Persons

Web Sites


Arequipa Automatic Airglow Observatory is jointly operated by the University of Pittsburgh and Clemson University with support from the National Science Foundation (NSF). Space and on-site technical support are provided by the National Aeronautics and Space Administration (NASA).

Data Description

The Fabry-Perot interferometer of the Arequipa Automatic Airglow Observatory (A3O) is operated jointly by the University of Pittsburgh and Clemson University under grants from the NSF Atmospheric Sciences Aeronomy division, with space and on-site technical support provided by the NASA Laser Satellite Tracking Station near Arequipa, Peru. The station is located at latitude 16.47 S and longitude 71.49 W at an elevation of 2489 m. The apex magnetic latitude is about -12, with a magnetic dip latitude of about 3, an inclination angle of -6 to -7, and a declination between +1 and -2 between 1983 and 1999. The mean local solar time is behind UT by 4 hours and 46 minutes (-71.49/15 = -4.766 = -4h 46m). The FPI uses the atomic oxygen filter at 630.0 nm, where the approximate emission height is assumed to be about 250 km.

FPI observations are adversly affected by clouds, moonlight, and haze conditions that lead to weaker airglow emissions. Results from the Boston University imager were used to deduce cloud cover from 1996-1999, with observations on very cloudy nights eliminated from the data set. In other years, poor data quality due to clouds was estimated from the data and those nights were removed. Also observations were not taken for the 5 nights around full moon. Solar minimum data sets are of lesser quality, because the airglow signals are considerably weaker. Further, active magnetic conditions lead to a higher ionosphere and consequently a weaker airglow.

A quality code (4064) is assigned for each night of data where the value:

  1. (=A) Nights with good quality results,
  2. (=B) Partial nights with good data or whole nights that yielded mediocre results (whether because of active magnetic conditions, weak solar minimum signals, or haze),
  3. (=C) Nights with just a few data points of mediocre quality, and
  4. (=D) Nights significantly degraded by clouds or for other reasons.

Nights classified as 4(D) have already been removed from this data set. In the plots below, these nightly quality codes are at the bottom day axis where 1=A, 2=B, 3=C, and ?=no designation.

Spatial coverage of the thermosphere's neutral wind field is facilitated by use of a dual mirror pointing head that allows any point in the sky to be selected for observation. Typically, a 6 position sequence is used in the repeated order N, E, Z, S, W, and Z. Most observation cycles involve twice as many zenith measurements, since they are important in the determination of the Doppler zero velocity reference. The order of observations is chosen to minimize the time between the two meridional and the two zonal observations. The elevation angle is 30 degrees. Starting in early 1997, the azimuth angles were rotated 10 degrees counterclockwise from the cardinal azimuths (i.e., to azimuth angles of - 350, 80, 170, and 260 degrees for "N, E, S and W") to facilitate bistatic measurements with an FPI in Chile. The midpoint time and azimuth and elevation directions for the original observing directions are given. The N,E and S,W pairs of wind components in the rotated coordinate system after correction for the conversion of line-of-sight (l-o-s) speeds to horizontal speeds were transformed to determine the geographic meridional and zonal wind components, which are listed in these tables. A positive value for the zonal wind component or for the meridional wind component indicates a flow direction that is to the geographic east or to the geographic north. To obtain the horizontal wind components from the l-o-s velocities, a multiplier of 1/cos(30) has been used, on the assumption that the vertical velocity is quite small. The errors deduced in the non-linear, least squares analysis represent an estimate of the standard deviation [Meriwether et al., 1997]. Horizontal wind errors were about 15 m/s for 1996 and 1997, improving to about 10 m/s for 1998 and 1999.

Several modifications were made to the instrument to reduce the measurement uncertainties. In October 1993, the RCA 31034 phototube was replaced by a Hammatsu R943-02 phototube which has a larger photocathode area. With this, we were able to improve the field-widened FPI's sensitivity by increasing the transfer efficiency of the radiation passing through the multiple order aperture plate to the photocathode. Then, in April 1996, a narrower pass-band 630 nm filter was installed in front of the FPI etalon, making possible the elimination of the OH contamination at 627.79 nm that had decreased the temperature determinations in 1994 and 1995, especially at the beginning and end of the night.

The 630 nm observations for each night were preceded by a series of laser calibrations using a HeNe frequency stabilized laser. The temperature results from the analysis for the 630 nm sky measurements were corrected to remove the instrumental broadening contribution by subtracting an average of the observed temperatures determined from the laser calibrations. This correction generally ranged from 10 to 50 K and represents the variability of the instrumental function from one night to the next. The observations for 1996-1999 with kindat 17002 were replaced with kindat 17003 to correct the temperatures. A few other corrections in the winds were made at this time also, but most winds and emissions are identical to the values in 17002.

In 2002, we plan to install a bare CCD camera detector that is expected to improve the sensitivity of these measurements by a factor of 15. After testing the new CCD detector in 630 nm measurements, the observations will be extended to determinations of 731.615 nm OH winds and temperatures at the centroid altitude of 87 km. We plan to scan the thermal structure of the upper thermosphere and exosphere region between 250 and 800 km by observing the 732 nm emission of O+ during twilight when the terminator is moving through this altitude region. Switching between the OI 630 nm and the O+ 732 nm (or OH 731.615 nm) wavelengths will be implemented by the use of an automated filter changer.

Calibration measurements at 632.8 nm are taken with a stabilized He-Ne laser every 5 to 10 minutes throughout the night. These are compared with the set of zenith observations at 630.0 nm to create a "zero velocity reference" on the assumption of a near-zero vertical wind when averaged over the night. There is an overall instrumental drift during the night equivalent to 100-200 m/s due to unavoidable room cooling as the temperature drops outside. To compensate for this we subtract from each zenith line position the line position of the 632.8 nm interpolated to the zenith measurement time to obtain an offset value. The average of these values for the whole night is the reference offset between the 632.8 nm laser line and the nightglow 630.0 nm line position for zero velocity. This overall offset value is applied to each measurement to find the Doppler shift and therefore the l-o-s velocity for that measurement. Positive differences are red shifts (motion away from the observer).

Between 1983 and 1990 are monthly quiet-time averages of the horizontal neutral winds measured by the Fabry-Perot interferometer (FPI) at Arequipa, Peru (16.5 S, 71.5 W). Bins are every half hour, where the time given is the midpoint of the bin. The errors reported for the monthly means are not measurement error bars, but are the standard error of the mean monthly values, which are generally much larger than the measurement error bars. The first plots below are of the monthly average horizontal neutral winds provided by Manfred Biondi ( The plots are labelled January through December, where each month is 24 hours in local time and the midpoint of the month is local midnight for the monthly average.

Measurements and plots between 1996 and 1999 are nightly observations of the horizontal winds, the vertical winds, the neutral temperatures, and the real emission rates in Rayleighs. Between 1996 and 1999, the Boston University CEDAR imager was also at Arequipa and was used to calibrate the FPI's relative intensities to yield absolute emission values in Rayleighs. The errors associated with the cross calibration are estimated to be about +/-20-25%. The 1 sigma uncertainty estimate given for each parameter (neutral wind, temperature, and intensity) results from the data analysis processing explained in detail by Meriwether et al [1997].

The south direction was chosen in addition to the zenith to plot the temperature and emission. The intensity values to the south of Arequipa are significantly higher because the Observatory is looking toward the Appleton anomaly, where the ionosphere has higher electron density due to the fountain effect - the uplifting of plasma from the magnetic equatorial region. Another reason that signal is higher in the south is the Van Rijn gain represented by horizon measurements.

References for the instrument and data processing procedures

Manfred A. Biondi, Dwight P. Sipler, and Maurice Weinschenker, Multiple aperture exit plate for field-widening a Fabry-Perot interferometer, Appl. Opt., 24, 232-236, 1985.
J. W. Meriwether, M. A. Biondi, F. A. Herrero, C. G. Fesen and D. C. Hallenback, Optical interferometric studies of the nighttime equatorial thermosphere: Enhanced temperatures and zonal wind gradients, J. Geophys. Res., 102, 20,041-20,058, 1997.
D. P. Sipler, M. A. Biondi, and R. G. Roble, F-region neutral winds and temperatures at equatorial latitudes: Measured and predicted behavior during geomagnetically quiet conditions, Planet. Space Sci., 31, 53-66, 1983.

Monthly Summary Plots for Arequipa Fabry-Perot in OI (6300 A)

Nightly Summary Plots for Arequipa Fabry-Perot in OI (6300 A)

-Revised 11 April 2002 by Barbara Emery