The ROCSAT-1 IPEI Preliminary Results: Low-Latitude Ionospheric Plasma and Flow Variations

We analyze the ROCSAT-1 IPEI data collected between March andJune 1999 to study the statistical features of the ion vertical drifts at equatorialand tropical latitudes.The dependencies of ion vertical drifts on localtime,longitude and geomagnetic field configuration,as well as geomagneticactivity are examined.The variations of the equatorial vertical driftsnear the dawn and dusk terminators are of particular interest.From thispreliminary study,we have shown that the overall local-time characteristicsof the quiet-time equatorial vertical drift patterns derived from IPEIare in good agreement with those observed by other satellites and ground-based instruments.More importantly,several new results due to the unique35° orbital inclination of ROCSAT-1 and the 100% duty-cycle operation ofIPEI are found.These include:(a)enhanced upward ion drifts to a criticallevel of 30-60m/s at post-sunset hours strongly correlate with the occurrenceof rising bubbles in the pre-midnight local time sector;(b)large(＞300m/s)downward ion drifts are most often found near sunrise and atlongitudes where the geomagnetic field has greatest variations;(c)the statisticaldrift patterns strongly depend on the hemispheres at the equatorialanomaly latitudes.This north-south asymmetry may result from seasonaleffects and/or from differences in geomagnetic field configuration.


INTRODUCTION
ROCSAT-1 was launched into a 600 km circular orbit with a 35° inclination at 1934 EST, January 26, 1999 from Cape Canaveral, Florida. The Ionospheric Plasma and Electrodynam ics Instrument. (IPEI) is one of the three payloads onboard ROCSAT-1. The IPEI payload consists of four sensors: an ion trap (I T) to measure the ion concentration; a pair of drift meters (V DM and HDM) to measure the cross-track velocity components perpendicular to the satel lite velocity, and a retarding potential analyzer (RP A) to derive the ion composition, tempera ture and the ram velocity. The !PEI hardware, system architecture, and the operation proce dure, as well as the scientific objectives have been published elsewhere in the literature (Chang et al., 1999;Yeh et al., 1999a).
The in-situ measurement of low-latitude ionospheric plasma parameters has been, at best, intermittent since the completion of the AE-E and Hinotori missions in the early 80s. The ROCSAT-1 !PEI payload, operated on a 100% duty cycle, will certainly quench the scientific thirst for new data as the sun-spot cycle returns to maximum activity. The fast longitudinal 788 TAO, Vol. JO, No. 4, December 1999 coverages by the low-inclined ROCSAT-1 orbit will provide us many exciting new observa tions. In this first report of the IPEI preliminary results, we shall present examples of longitu dinal variations of the vertical flow in the dawn sector, together with a couple of the new observations of the Equatorial Spread F event on the transitional scale (10 to 100 m).

IPEI DATA PRESENTATIONS
ROCSAT-1 is a three-axis stabilized spacecraft with the z-axis pointing toward the Earth's center, the y-axis pointing to the negative direction of the orbital angular momentum, and the x-axis along the velocity direction in a circular orbit. The normal of the IPEI sensor plane is directed along the x-axis with the two DM sensors aligned in the horizontal plane. Figure 1 depicts the relationship between the spacecraft coordinate system and the Earth-centered iner tia (ECI) coordinate system. The IPEI data plotted in the Quick-Look display format for public browsing is in the spacecraft coordinate system. Figure 2 shows  The offset of V is somewhat more complicated to obtain. First we notice that V contains the projection of a 6'orotation flow velocity which has a value given by where roE!l is the Earth's angular velocity and r is the radial position of ROCSAT-1 from the Earth's center.La ndi represent, respectively, the ROCSAT-1 geocentric latitude and orbital inclination (i =35° ). The positive sign in Equation (1)       The plcit of V, data in Fig. 3 by itself is very interesting. First of all, we notice that the amplitudes of the variations in V , are quite large and they are not due to the offset in V ,· Apart from many outliers, the dominant part (dark part) of the variation in V, from dawn to midnight is rather featureless. The prereversal enhancement of the post sunset effect has been masked out and become obscure. However, when the seasonal effect is separated out, the prereversal enhancement does stand out from the average (Yeh et al., 1999b).
The variation of V around dawn in Fig. 3 is quite variable. When many consecutive z ROCSAT-1 orbits are examined, the variation of V, around dawn shows a dramatic longitudinal and latitudinal/local-time dependence. This is demonstrated as follows. The orbit of ROCSAT-1 is fixed in an inertia coordinate system except that the oblate Earth together with the Sun's motion around the Earth will cause the orbit to displace westward 0.46 hour per day in local time. Thus, for practical purposes, every data point in the 15 orbits of ROCSAT-1 per day can be assumed to be fixed at a certain local time and latitude. However, the longitudinal location along the orbit varies as the Earth rotates under the ROCSAT-1 orbit so that the IPEI data can be used to study the longitudinal variations of the ionospheric plasma and flow in one day. The local-time effect is then derived after many days' data are examined.
We select March 23, 1999 to start such a study. This day is a geomagnetically quiet day with Kp = 1-, 1-, 1-, 2-, 2, 2, 2, 2+ for each 3-hour slot. The first 360 minutes of data are shown in Fig. 4. The plot utilizes the Quick-Look display data except that the offset in V is removed z and the corotational value in V is added as the dot-dashed line in the second panel for reference. The V,, V Y , o+ temperat� re and Q+ concentration percentage (newly added data) have undergone 16-second running averages to reduce scatter. The general feature of diurnal variations in ion density N, and o+ temperature can be easily recognized. The ion density changes from 104/cm3 at midnight to 106/cm3 around local noon and again around 2000 LT in the equatorial anomaly region. The equatorial spread F events of density depletion near the equator are conspicuously noticeable. The Q+ temperature drops to 1000 K in the night and rises to 2000 K during the daytime.
The diurnal variations in the ion density and temperature seem to be more or less univer sal and appear to be longitude dependent, as reported in the literature (Anderson, 1981;Su et al., 1996). On the other hand, the drastic variations in V ,. V Y and o+ concentration percentage around the dawn sector is very prominent, as seen in Fig. 4 .. Around 0500 LT at dawn, the VDM detects a large downward flow in V, at -400 mis when ROCSAT-1 is at latitude 30° S and longitude 30° E, a location near the South Atlantic Anomaly region. V also deviates y strongly from the corotational reference value. The o+ temperature shows a large fluctuation and the Q+ concentration percentage reaches the minimum dip of about 25%. The other ion species, most likely H+, rises to 75%, as the data implies. These drastic changes subside UT .. ,.,   shortly after sunrise as ROCSAT-1 moves away from the Anomaly region.

03:la
The longitudinal dependence of these variations can be reconfirmed from further study of the IPEI data 12 hours later, when the ROCSAT-1 orbit drifts in longitude to about 180° away from longitude 30° E. Figure 5 displays data from such an observation. Notice that the local dawn terminator remains at latitude 30° S, but falls on longitude 220° E. The diurnal varia tions in ion density and o+ temperature are much flatter but retain similar patterns, as seen in F ig. 4. However, the variations in V "' V Y and o+ concentration percentage around 05 00 LT, the dawn terminator seem to be diminishing.
ROCSAT-1 drifts westward 0.46 hour per day in local time so that 26 days after March 23, 1999 the dawn terminator of ROCSAT-1 falls on the northern hemisphere. Figure 6  ROCSAT-1 emerges from the nightside ionosphere in the southern hemisphere, the cause could be related to a large H+ flow from the protonsphere to ionosphere (Hanson and Ortenburger, 1961) or inter-hemispheric transport phenomenon. Some detailed statistical study of the hemispheric asymmetry on the vertical flow is reported by Yeh et al. (1999b).
Incidentally, a surge spike in the downward flow of V z observed at sunrise, for example at 05 14 LT (0046 UT) in Fig . 4, has been ruled out as was caused by the ROCSAT-1 spacecraft potential changes at sunrise. Since the spike is also related to the spike in the decrease of the Q+ percentage, it is very likely to be related to the H+ flow surge from the higher altitude region of the protonsphere where the Sun's light arrives earlier than the lower part of the protonsphere ionosphere boundary where ROCSA T-1 is orbiting.

FLOWS IN EQUATORIAL SPREAD F
Another prominent feature noticed in Figs. 4 and 5 is the frequent occurrence of the bubbles and blobs in equatorial spread F (ESP) events. The expanded plots of data from 1440 to 15 00 UT for March 23, 1999 are shown in Fig. 9. The ESF events have been observed by spacecraft for many years (see e.g., Hanson et al., 1973;Dyson et al., 1974;Oya et al., 1986;Kil and Heelis, 1998). However, the !PEI data shown in Fig. 9 can provide a better understanding of the microscopic features of an ESP event with the simultaneous measurements of V , V , N, Q+ temperature and Q+ concentration percentage. We shall first study the flow patte:.O i� Fig. 9       for the gross feature of the ESF events.

ROCSAT-1/IPEI
In the ion depletion region where low ion density (bubbles) occurs, such as that which started at 1444 UT when V is seen to flow upward. On the other hand, V is southward until z y 1448 UT but turns northward afterwards. However, when the occurrence of ESF events is related to the geomagnetic location of the ROCSA T-1 orbit, we conclude that the V Y flow is poleward along the field line at locations both before and after 1448 UT. The accompanying er temperature fluctuates and so does the o+ concentration percentage. The detailed ion mo tions inside the bubbles have been reported by Yeh et al. (1999c).

S. POWER SPECTRUM OF FLUCTUATIONS INSIDE A BUBBLE
The 1024 Hz sampling rate for density, V and V fluctuations inside a bubble will enable ' y us to study, for the first time from a spacecraft observation, the dynamics of bubble structure on the transitional scale range (10 -100 m). An example of such observation is shown in Fig.  10 for data from 0325 to 0327 UT on March 28, 1999. This example is chosen from when the IPEI operated at Fast Mode to sample data at 1024 Hz during the passage of bubble events.
Following a procedure similar to Kil and Heelis (1998) The power spectral density in the fourth panel in Fig. 11 has been broken into three sec tions, as discussed by Kelley et al. (1982) -medium scale (10-1000 km), intermediate scale (0.1 -10 J<.m), and transitional scale (10 -100 m) -to fit Pocfll separately to obtain the spectral indices, n 1 , n2, and n 3 • The values of n1 and n2 have been reported from previous spacecraft observation ( Kil and Heelis, 1998). The n3 value has only been obtained by rocket observation (Kelley et al., 1982). The value n3=-4 seen in the fourth panel is a new result and is similar to those reported by the rocket observations in a slightly lower ionospheric region. However, the fluctuations in 8 V, and 8 V on the transitional scale have not been reportedfrom any space bome instruments. Detailed �nalysis of the fluctuation in the ESP events has been reported by Su et al. (1999).

CONCLUSIONS AND DISCUSSION
The low-inclination of the ROCSAT-1 orbit enables us to examine the local-time/ latitu dinal dependence of the longititudinal variation of the low-latitude ionospheric plasma elec- trodynamics. In this first report of the preliminary results, strong hemispheric asymmetry of the enhanced downward flow at dawn has been observed. Although the mechanism for the flow of 300 meters per second or more cou. ld be related to the protons (as inferred from IPEI/ RPA data) rushing down along the field line to the ionosphere at dawn, detailed analysis is still needed to support this.
Furthennore, this report is to fulfill the design goal of the IPEI payload to be the first spacecraft to observe the equatorial spread F (ESP) event on the transitional scale (10 -100 m). The spectral index of the power spectral density for the density fluctuation in the ESF event is found to be about -4, which is very close to the rocket observation made in the early 80s.
The purpose of this preliminary report is to demonstrate the fine quality of IPEI data such that they can be readily used for new ionospheric study inferred from data published in the Quick-Look display plots in the Website. This report shall encourage scientists to utilize the IPEI data.