Instruments:rfp

From CedarWiki
Revision as of 17:12, 18 April 2008 by Emery (Talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

Resolute Bay Fabry-Perot

Contents


Contact Persons

Web Sites

Acknowledgments

The Resolute Bay Canada Fabry-Perot Spectrometer is operated by the High Altitude Observatory of the National Center for Atmospheric Research with support from the Fabry-Perot Interferometer at Resolute Bay, Canada (74.73N, -94.89E, 87 m ASL) has been operated since 2003 by the High Altitude Observatory of the National Center For Atmospheric Research. The apex magnetic coordinates of Resolute Bay at 250 km height in 2004 were (83.1, -39.0) with a magnetic declination of -24.4 deg, an inclination of 88.0 deg, and 0 UT at 16.31 MLT.

Description of Instrument

The dispersing element of the interferometer is an air-spaced, 10 cm diameter effective clear-aperture Fabry-Perot etalon. It has a fixed gap of 2.0 cm. The etalon is housed in a temperature controlled chamber (Wu et al. 2004). The instrument detector is a back illuminated CCD camera with 1024x1024 pixels and 1x1 inch size. The CCD temperature is set at -55 C. The readout noise of the CCD is 4 electrons.

The interferometer has an 8-position filter wheel. The wavelengths used at Resolute Bay are:

  1. The red line (630.0 nm, kindat=17011) of atomic oxygen (OI) with a typical emission height peak in the range 210 to 300 km.
  2. The green line (557.7 nm, kindat=17012) of atomic oxygen (OI), with a typical emission height peak range near 94-98 km.
  3. The [OH] line (892.0 nm, kindat=17013) of the nightglow excited hydroxyl [OH*] with an emission peak between about 87 and 91 km.

Only the 630.0 nm emission data are in the CEDAR base. The other emission data will be submitted in a later date. They are available upon request.

The observing sequence is a zenith observation followed by the 4 cardinal directions of W, E S and N at an elevation of 45 degrees above the horizon. For the red line, each direction is integrated for 5 min for a total observing cycle of 25 min. The time listed in the records is the start time of each 5 min integration. The end time in the prologue is the start time of the last time plus 5 minutes.

The 'zero wind position' (quality code 4172) is determined on hourly basis. Observations were eliminated in the hourly blocks of the zero wind if there were -NaN or *** (too large) values. Individual points were eliminated if the line-of-sight (los) or wind error bars (codes -800, -1410, -1420, -1431) were bigger than 200 m/s, if the horizontal wind magnitudes (1410, 1420) were larger than 1000 m/s, if the temperature (code 813) was negative or its error bar larger than 500K, if the brightness (code 2506) was less than 15,000 or its error bar greater than 5,000. Very low relative temperatures (code 813) are found in solar minimum around local midnight when the brightness is less than ~30,000, although the winds appear to be fine, so relative temperatures less than 550 were set to missing in these low brightness conditions. The filter focus stuck in the summer of 2007 so the fringes were more blurred, possibly resulting in higher temperatures in the 2007-2008 season.

Nightly medians were found using the hourly blocks of the los zero wind (code 4172). If the hourly block was non-zero and was different from the nightly median by more than 200 m/s, the block was eliminated if the solar zenith angle (code 180) was less than 100 degrees. Otherwise, the hourly block los zero wind value was replaced with the nightly median value and the los (800), horizontal (1410, 1420) and vertical (1431) winds were re-calculated and a flag (code 4174) was set to 1. Nightly median, mean, and standard deviation values of the los zero wind (code 4172), the relative neutral temperature (code 813), and the vertical wind (code 1431) were re-calculated, excluding any changed times. Then if the difference between the relative neutral temperature and its nightly median was more than 400 K, or the difference from the nightly mean was more than 3.5 times the standard deviation, the data were eliminated for that time. If the difference of the los zero wind from the nightly mean was larger than the maximum of 65 m/s or 2.9 times the standard deviation, the los zero wind (code 4172) was replaced by the nightly median and the los etc winds were re-calculated and the flag (code 4174) was set to 1. If this resulted in winds that were more divergent in the NS or EW directions, then the los zero wind change was not made, while for other periods when the mean and block difference was 2.5 times the standard deviation, the change in the los zero wind was made. If the individual vertical wind was different from the nightly median vertical wind by more than 100 m/s or was different from the nightly mean by more than 3.5 times the standard deviation, the data for that time were deleted. Most deletions occurred at low solar zenith angles or during high cloud cover conditions. Users should beware of values at the beginning or end of the nights where the solar zenith angle (code 180) is less than ~105 degrees.

The automatic checks for bad data were expanded in April 2008 so the 222 nights between Oct 2003 and Apr 2005 were re-analyzed. Of 14790 observations, 193+79=272 or 1.8% (previously 271) were removed/changed. The initial analysis was redone for 2005024,054 to reduce anomalous Tn. After 2004070, all negative background values were set to missing.

309 nights between Oct 2005 and Jan 2008 were added. Of 23095 observations, 919+100=1019 or 4.4% of the total were removed or changed. More observations were taken at low solar zenith angles, so more observations were removed than during earlier periods.

The observations are made during the northern winter season (Sep-Apr). Because of the narrow filters used, operation of the instrument is not affected by moonlight. All the observations obtained are included since the thermospheric temperature observations are statistically indistinguishable in the absence or presence of clouds (Smith and Hernandez, 1995). This does not apply to the wind and brightness observations, which are strongly affected by the presence of clouds, and should not be used if the cloud cover is more than 50%. The cloud cover (code 442 in tenths of total cloud amount covering the sky) observations are obtained from Environment Canada and are described in an associated header.

Reduction

Doppler shifts (winds) are determined from the displacement of the line profile relative to the hourly 'zero wind position' observations (code 4172) along the line-of-sight (los). The los winds before the hourly average zero wind removal (code 4173) are converted to regular los winds (positive away, code 800) by subtracting the average los zero wind position (code 4172). Horizontal winds are calculated as the los winds over the cosine of 45 degrees, the elevation angle (code 140), positive to the east and north (codes 1410 and 1420).

The temperatures are not calibrated to absolute values and can only be used for relative variations (code 813).

Summarizing, the reported measurements are horizontal and vertical winds, relative temperatures and relative emission rates. The 'errors' given in the data are uncertainties of measurement, that is the statistically determined effect that noise in the measurement will cause in the final result. This noise is inherent to the signal, since photons obey Bose-Einstein statistics. These uncertainties are 1 sigma uncertainty of the deduced horizontal wind, temperature and emission rate.

Sky Cover

Sky cover observations are included with Fabry-Perot (FPI) data because clouds can adversely affect ground based observations of the thermosphere. The observations are made during the northern winter season (Sep-Apr). Because of the narrow filters used, operation of the instrument is not affected by moonlight. All the observations obtained are included since the thermospheric temperature observations are statistically indistinguishable in the absence or presence of clouds (Smith and Hernandez, 1995). This does not apply to the wind and brightness observations, which are strongly affected by the presence of clouds. Overcast conditions show no observable radial Doppler shift (very low winds in all directions), while clear conditions show a consistent wind pattern in opposite look directions. For in-between conditions (which occur most of the time), the instrument estimate of sky cover may not match that of a ground observer.

Sky conditions are routinely observed every hour at nearby Resolute Bay by Environment Canada. Sky conditions are assumed to persist between observations, which are taken at the beginning of each hour. When no observations are made, the cloud cover is set to -32767 (missing). The original sky data contain quantitative estimates of the total cloud opacity and the total cloud amount in tenths, and the same estimates of cloud opacity and cloud amount for the lowest three cloud levels. This information is proprietary, so only the total cloud amount in tenths of the sky covered is retained for the UT periods of the observations and is located in code 442. In addition, the sky cover in tenths is available for only the winter periods of September-April for 2003-2007. These were made from the original data set from Environment Canada, where the local standard time (LT) was changed to universal time (UT=LT+6 hours) and linked in this file at:

  1. Cloud Cover Data
  2. Cloud Cover Plots

To summarize, sky cover is a subjective evaluation of the skyward visibility which conservatively characterizes actual FPI viewing conditions at Resolute Bay. FPI winds are most trustworthy under clearest conditions (0-3 tenths or 0-30%), perhaps still okay under partly cloudy conditions (up to 5 tenths or 50%), and unreliable when cloudier (more than 5 tenths).

References for the instrument and data processing procedures

Smith, R. W. and G. Hernandez, Upper thermospheric temperatures at South Pole, Adv. Space. Res., 16(5), 31-39, 1995.
Wu, Q., R. D. Gablehouse, S. C. Solomon, T. L. Killeen, and Chiao-Yao She, A new NCAR Fabry-Perot Interferometer for upper atmospheric Research, Proc. SPIE, 5660, 218-227, 2004.

Data Files for Resolute Bay cloud cover in tenths

Summary Plots for Resolute Bay cloud cover in tenths

These are summary plots of the total cloud cover in tenths from Environment Canada for Resolute Bay.

Summary Plots for Resolute Bay Fabry-Perot in OI (630.0 nm) (peak emission ~210-300 km)

Summary plots of the cloud cover in tenths, the 4 cardinal directions of the horizontal winds, the vertical winds, relative emission and temperature. The distance between opposite look directions at 45 degrees elevation angle is about 500 km for a 250 km emission layer. The relative emission and relative neutral temperature show values for the vertical look direction, and the S direction. Cloud cover observations are usually every hour and are assumed to persist during that hour.


-Revised 19 Apr 2008 by Barbara Emery