The MU (Middle and Upper atmosphere) radar (34.8N, 136.1E) is a 46.5-MHz, pulse-modulated, monostatic, Doppler radar with an active phased-array antenna which consists of 475 Yagis. This MST radar is also designed to be able to observe the weak incoherent scatter (IS) from free electrons of the ionosphere. Measurements are made at 4 beam directions simultaneously.

Initially (before December 1996), a 1-hour observation cycle was applied with the first 15 minutes of each cycle devoted to single-pulse measurements (for electron density), and the last 45 minutes to two-pulse (for ion drift velocity) or four-pulse (for ion/electron temperature) measurements. Measurements have usually been taken about 3 days per month starting on September 17, 1986. On the first 2 days, drift measurements are combined with electron density measurements, and on the last 1 day temperature measurements are combined with electron density measurements.

Since December 1996, a new scheme has been applied to improve the simulaneity between electron density and ion drift (or temperature): 1-minute single-pulse measurements are alternated by 3-minute multipulse measurements. Each year, about 9 IS experiments have been performed with the radar, 6 of which are for electron and drift combined measurements and 3 for electron and temperature combined measurements. Each experiment period is continuous 4 days, making the total number of observation hours in a year for each combination the same as before December 1996.

Power Profile

Electron density data in this database are obtained as hourly averages. Before December 1996, these averages come from the 15 min power profile measurements; after that they are from measurements over 1 hour period centering the hour.

Data above around 600 km usually become unusable owing to weak signal strength. Data above 1000 km are used to determine the noise level. Data below about 200 km may be contaminated by meteor echoes or coherent clutter echoes. The data for Codes 4101-4104 are for the different directional beams of the radar. All are for an elevation angle of 70 degrees. The azimuth angles are 355, 85, 175, 265 degrees for KINDATs of 8001, 8002, 8003, and 8004, respectively. The range averaging is over 16 ranges or 76.8 km. The data in each beam can be compared to see the tilt in the ionosphere, and other features. Gravity wave analyses are usually done with data using a longer pulse length.

The measured power profile has to be corrected and calibrated to get the true electron density profile. We first perform the plasma temperature effect corrections. A model of electron temperature over ion temperature ratio, Tr, is used. The model is based on the MU radar temperature measurements over 1986 to July 2003. The temperature data were divided into 8 bins: 2 solar activity bins and 4 seasonal bins, and averages were obtained. Then by using foF2 obtained by an ionosonde at Kokubunji, Japan (34N, 139E), we find the calibration factor and apply it to get the absolute value of electron density. An interpolation is used to get an appropricate foF2 at the radar centeral integration time.

The 4 MU radar beams are examined, and at each altitude, the median is found, which is the average of the 2 central values. Nighttime data during the experiments (2-4 days) are averaged for both the Kokobunji and the median MU radar data between the local times of 21 and 3. A single calibration factor of the ratio of the average Nmax(ionosonde) to the average Nmax(4 beam median MU radar) mentioned above is then obtained.

Paracode 4101-4104 are the electron density for the 4 beams, corrected for Tr effect, and calibrated with that single calibration factor. Paracode 4105 are results of Tr correction and foF2 calibrations on the median profiles of the 4 beams, and can be considered as the vertical profiles over the radar.

Other than the single calibration factor which is based on nigttime observations, we can also obtain hourly calibration factor when hourly foF2 data are available. Paracode 520 is the result after correcting Tr effects and calibration with hourly foF2 on the median profile of the 4 beams.

Plasma Temperatures

The MU radar plasma temperature are often obtained with a 4-pulse experiment. With a subpulse width of 96-us, the radar provides a 14.4-km range resolution and 6 points on the ACF evenly spaced from 192 to 1152-us lag. The radar receiver samples are collected at 32-us intervals and integrated over 10 range gates to give a single data record, producing a 45-km height resolution. The total length of the transmitted wave form of 1248-us is followed by the system guard time of 74-us, resulting in a lowest possible observable range of 198.2 km.

The ACFs obtained from the 4 directions are used to yield a mean temperature above the radar throug comparison with a library of theoretical ACFs. The ions are assumed to be O+ except in the range below 270 km altitude, where a specific height-dependent mixture of O+, NO+, and O2+ ions is assumed (Oliver, JATP, 37, 1065, 1975).

Ion and electron temperatures (Paracods 550 and 560) in this database are obtained as hourly averages. Before December 1996, these averages come from the 45 min spectrum measurements of temperatures in the last 3 quarters of each hour; after that they are from measurements over 1 hour period centering the half hour time.

Ion Drift

Ion drift velocity is measured with a two-pulse experiment, consisting of two 256 microsecond pulses separated by a 256 microsecond pulse gap. This waveform provides a range resolution of 38 km. This experiment provides a measurement of one lag on the signal autocorrelation function (ACF), at 512 microseconds, from which the line of sight velocity can be derived. Measurements are carried out simultaneously at 4 azimuths (north, east, south, and west), all at 20 degress from zenith. The four line-of-sight velocities are combined to determine a vector drift in magnetic coordinates: perpendicullar-east, perpendicular north, and parallel-upward.

"Transmitter frequency chirp" offset occurs due to the a slight droop in transmitter voltage and a resultant phase lag between two transmitted pulses. This offset was detected, through laboratory test, as -22.8m/s for the line-of-sight velocity. Assuming that the nighttime average of vertical ion drift velocity is zero, we can obtain another offset value which gives -39 m/s shift to the vertical drift velocity (or -36.6m/s shift to the line-of-sight velocity). Ion drifts in this database have been corrected for the -39m/s vertical velocity offset.

The ion drifts in this database are hourly averages over the height range 220-450km.

The vertical ion velocity can be determined from the average of the beams to the north and south (Code 4090/4110) and from the average of the beams to the east and west (Code 4091/4111). The difference of these two calculations of the vertical ion velocity is some measure of the error. The average of these two calculations is the actual vertical ion velocity (Code 1230/4112) used to determine the perpendicular northward and parallel components (Codes 1250 and 1260).

Before December 1996, these hourly values come from the 45 min velocity measurements in the last 3 quarters of each hour; after that they are from measurements over 1 hour period centering the half hour time.

Meridional Winds

Derived parameter magnetic meridional wind (code 1460) are calculated using "layer-wind" method (Oliver et al., Radio Sci., 33, 941, 1998). This wind is essentially a weighted average of the wind in the height range 220-450km. The weighting factor is very close to the singal strength.

The parallel velocity, used for wind computations, is such a weigthed average on the line-of-sight velocities. The diffusion velocity, to be removed from the parallel drift in order to get the wind, is also an average of invidual diffusion velocities in the same height range of 220-450km. They are calculated using electron and ion temperatures from the IRI95 model, neutral composition and temperature from the MSIS86 model, the collision frequency from Pesnell et al. (GRL, 20, 1343, 1993), electron density from the MU radar power profile measurements.

The MU radar belongs to and is operated by the Radio Science Center for Space and Atmosphere (RASC) of Kyoto University, Japan.


S. Fukao et al., The MU radar with an active phased array system, 1, Antenna and power amplifiers, Radio Sci., vol 20, pp 1155-1168, 1985,

S. Fukao et al., The MU radar with an active phased array system, 2, In-house equipment, Radio Sci., vol 20, pp 1169-1176, 1985,

T. Sato et al., Ionospheric incoherent scatter measurements with the middle and upper atmosphere radar: Techniques and capability, Radio Sci., vol 24, pp 85-98, 1989.

W. L. Oliver, et al., Ionospheric incoherent scatter measurements with the MU radar: observations of F-region electrodynamics, J. Geomag. Geoelectr., 40, 963-985, 1988.

W. L. Oliver et al., Measurements of ionospheric and thermospheric temperatures and densities with the middle and upper atmosphere radar, J. Geophs. Res., 96, 12,827-17,838, 1991.

Contact Information

     Prof. S. Fukao
     Radio Science Center for Space and Atmosphere, Kyoto University, 
     Gokanosyo, Uji, Kyoto 611, JAPAN 

     Phone:   +81-774-33-5343 
     Telefax: +81-774-32-8341

     Prof. W. L. Oliver
     Electrical and Computer Engineering, Boston University
     8 Saint Mary's Street, Boston MA 02215, USA

(Thanks to S. Kawamura's help, this description was prepared by Shun-Rong Zhang. For questions relevant to the MU radar database installed on the Madrigal system, please contact S.-R. Zhang at