Storm Studies Using NTS2 Navigation Signals

Using data obtained from the Side-Tone-Ranging Subsystem (STR) of the Orbit Determination and Tracking System (ODATS) of the Navy NTS-2 satellite, NRL scientists were able to derive some interesting information about ionospheric storm effects [Goodman and Martin, 1983], The NTS-2 satellite, like its predecessor NTS-1, was developed by the Naval Research Laboratory as an experimental prototype of satellites in the NAVSTAR/GPS constellation. The satellite was launched in 1977 into an orbit characteristic of modern day GPS satellites. It had coherent transmissions of 335 and 1580 MHz that could be used to study ionospheric parameters. These included TEC variations and ionospheric inhomogeneities responsible for amplitude and phase scintillation. Data were obtained at four locations: NRL Chesapeake Bay Division, near Washington, DC; Panama; Australia; and Great Britain. It is noteworthy that NTS-2 had two separate navigation subsystems, one the STR-ODATS (previously mentioned), and the second a pseudo random noise (PRN) pulse-ranging system that provided the main navigation signal. It was comprised of transmissions at nominally 1575 MHz and 1227 MHz with a waveform defined by bit rates of 1 or 10 Megabits/sec. For simplicity at the time, the NRL ionospheric studies exploited the STR-ODATS ranging tones at 335 and 1580 MHz instead of the dual frequency (L-Band) PRN waveforms. The ODATS system transmitted ranging tones of 335 and 1580

MHz, as mentioned above, with the UHF and L-Band transmissions consisting of a carrier, a single side-band reference tone and ten ranging tones equi-spaced up to 6.4 MHz, from the reference tone.

Based upon the NTS-2 data, high latitude scintillation observations have been presented by Goodman et al. [1978] and Goodman [1979], Figure 3-27 shows a period of scintillation to the north of Washington DC as NTS-2 transited the oval and trough regions moving equatorward. Notice the loss of phase lock for part of the pass. This event was unusual since it exhibits a distinct trough region during daylight hours within which the slant TEC is significantly depressed, and this is correlated with scintillation occurrence near the poleward boundary. A fading range in excess of 40 dB was observed at 335 MHz near the poleward segment of the trough. It is noteworthy that on December 2nd, the A-index rose to 46 at Fredericksburg and 191 at Anchorage, while the planetary Ap index was ~ 70. The K-indices ranged from 4-6 at Fredericksburg and 3-9 at Anchorage. Clearly storm activity has moved the high latitude trough significantly equatorward.

From the Panama site, it was possible to examine the impact of storm effects on the anomaly crest. Figure 3-28 shows the ODATS output during the October 31-November 01, 1977 period. Two curves are shown, with the upper and lower curves corresponding to the L-Band and UHF signals respectively. Ignoring the data trends, for present purposes, the only significance of the data lies in the (vertical) difference between the two curves. This represents the ionospheric contribution to group path delay. It is seen that the difference is smallest near the closest point of approach of the satellite to the ground station (i.e., the CPA), and this would generally be expected. Of significance is the fact that the greatest difference appears on the N-S transit, somewhat ahead of the meridian transit that is due south. This corresponds to the northern hemispheric crest of the equatorial anomaly. The erosion of the crest is also observed as the satellite moves further southward and goes over the horizon near the meridian transit point. On its northward journey, the NTS-2 maps the same anomaly crest, albeit seven hours later.

Figure 3-29 is the superposition of UHF and L-Band phase data for a number of days. The presentation is a bit peculiar, but we have chosen to preserve the original format. The shaded area approximates the slant TEC at any given time, since it shows the time-delay difference between the UHF and L-Band signals. From the Panama site, and the given orbit, one may only observe the Northern Hemispheric crest of the Appleton Anomaly, as the southern crest is over the horizon. In any case, the northern crest is clearly observed as NTS-2 moves southward, and once again (but less distinctly) when it moves northward. From Figure 3-29, there is some evidence that Kp variations are correlated with poleward motion of the anomaly crest. This phenomenon will be revisited below.

Figure 3-27: (Top): RF phase variation (radians), time delay (ns), and the slant TEC (electrons/in2 x 101 ) for aNTS-2 pass on December 2, 1977. The vantage point is Washington, DC, and the mean ionospheric point (MEP) referenced to 400 km was transiting the Great Lakes region at midday (i.e., 1334-1508 LMT, and 1845-2008 GMT). (Bottom): Fading range at UHF is given in the approximate SI units. Here SI ~ the S4 index. The fading range exceeded 40 dB for more than 10 minutes and there were four periods of phase-lock being lost. From Goodman and Martin, [1983],

Figure 3-27: (Top): RF phase variation (radians), time delay (ns), and the slant TEC (electrons/in2 x 101 ) for aNTS-2 pass on December 2, 1977. The vantage point is Washington, DC, and the mean ionospheric point (MEP) referenced to 400 km was transiting the Great Lakes region at midday (i.e., 1334-1508 LMT, and 1845-2008 GMT). (Bottom): Fading range at UHF is given in the approximate SI units. Here SI ~ the S4 index. The fading range exceeded 40 dB for more than 10 minutes and there were four periods of phase-lock being lost. From Goodman and Martin, [1983],

Figure 3-28: STR-ODATS data from NTS-2 signals on October 31- November 1, 1977. The receiver site was at Panama. The difference between the L-band (top curve) and the UHF (lower curve) signals can be translated into the slant TEC. Quasi-periodic fluctuations in TEC are observed equatorward of the anomaly crest. Only the northern hemisphere crest is observed on the N-S and the S-N transits. From Goodman and Martin [1983],

Figure 3-28: STR-ODATS data from NTS-2 signals on October 31- November 1, 1977. The receiver site was at Panama. The difference between the L-band (top curve) and the UHF (lower curve) signals can be translated into the slant TEC. Quasi-periodic fluctuations in TEC are observed equatorward of the anomaly crest. Only the northern hemisphere crest is observed on the N-S and the S-N transits. From Goodman and Martin [1983],

Figure 3-29: NTS-2 data at UHF and L-Band between 11-28-1977 and 12-06-1977. The vertical separation between the 335 and 1580 MHz signals is proportional to the group path delay difference between the two signals (i.e., slant TEC). From Goodman and Martin, [1983].

3.10.4 The Halloween 2003 Storm

The space weather aspects of the Halloween storm period were covered in Section 2.3.8, where the "upstream" aspects were emphasized (i.e., solar emissions, the IMF influences, and the magnetic activity response.) In this short section we will discuss the ionospheric disturbances that occurred during the period (i.e., "downstream aspects) and will compare that with some predictions from the STORM model. Some of the telecommunication system effects are described in Section 4.5.2. Figure 3-30 gives the STORM model prediction for the Northern Hemisphere for the 2-day period from October 31 to November 01, 2003. During the stormy period, a lot of the interesting sites that collected foF2 data under normal circumstances exhibited vulgarized data, but. Figure 3-31 and Figure 3-32 show some results for Sondrestrom (77°ML) and Eglin AFB (~40°ML) respectively. A climatological prediction offoF2 for the sites is also given. We see from the STORM prediction (i.e., Figure 3-30) that there is a marked difference between the high latitude and low latitude dependence of the predicted foF2 "multiplier". An enhancement is predicted for sites in the lower CONUS and a marked diminution is predicted for the upper CONUS. A comparison of the observations with the (transient) STORM predictions as well as the climatological predictions points out the difficulty associated with any prediction methods. We know that climatológica! "predictions" cannot account for storms. The STORM model develops its predictions based upon an historical record of storms, and its output is basically an "average" prediction of stormtime effects. But there is no average storm; each storm has its own eccentricities. The Halloween storm was no exception. More work is need in this area.

Figure 3-30: STORM mode! predictions during the Halloween storm period. The plots represent the "multiplier" for the median value offoF2 based upon URSI-88 coefficients. See Section 3.10.2. The numbers on the RHS of each graph are the last recorded values, and not the maximum values. By permission of NOAA-SKC, Department of Commerce, Boulder, CO.

Figure 3-30: STORM mode! predictions during the Halloween storm period. The plots represent the "multiplier" for the median value offoF2 based upon URSI-88 coefficients. See Section 3.10.2. The numbers on the RHS of each graph are the last recorded values, and not the maximum values. By permission of NOAA-SKC, Department of Commerce, Boulder, CO.

Sondresiroiii Oct 312003

Sondrestrom Nov 01,2003

1200

t nffwy

ttiii&ai aoo

1200 aoo

Time (UTC)

1200

Model Prediction

1200 aoo

Time (UTC)

1200

Figure 3-31: foF2 data from Sondrestrom on Oct. 31-Nov. 01. 2003. A climatological prediction is shown using CCIR coefficients. Raw data was provided by NOAA-SEC.

Eg! in Oct 29,2003

15 10

0:00

1200 aoo rime (UTC)

1200

Eglin Oct 38,2083

15 10

1200 aoo

Time (UTC)

1200

Eglin Oct 31,2063

Model Prediction

15 10

aoo 1200 aoo 1200

Time (UTC)

1200 aoo

Figure 3-32:foF2 data from Eglin AFB for Oct. 29-31. A climatological model prediction of the foF2 is also shown using CCIR coefficients. Raw data was provided by NOAA-SEC.

In some instances, it has been shown [Coster, 2004] that the Equatorial Anomaly (EA) peaks move poleward during geomagnetic storms, and GPS-TEC data seems to support that. This is a very important result, and is consistent with some earlier work of Goodman and Martin [1983] described in Section 3.10.3. It suggests that poleward movement of the anomaly crest can sometimes be coincident with the well-known equatorward expansion of the auroral oval. This implies that during some geomagnetic storms the region we normally call "midlatitudes" may be contracted significantly. From a propagation perspective this introduces some important system considerations.

Figure 3-33 is the depiction of an SED event during the Halloween 2003 storm. We will continue the impact of storms and SEDs in Chapter 4.

0 0

Post a comment