3.1. Storm Enhanced Density in Australia
 The MWA differential TEC maps will provide important information about the characteristics of SED in the Southern Hemisphere. The SED phenomenon is responsible for some of the largest space weather effects in the midlatitudes [Coster and Foster, 2007]. SED structures [Foster, 1993] are associated with severe gradients in TEC, and these gradients can lead to large errors in the measured range, or in differential range, used for precision applications such as geodesy, surveying, and navigation of inland waterways. Some of the largest TEC gradients on Earth, in the range of 50 TEC units per degree have been observed over the Northeast and Northwest continental U.S. SED effects can persist for several hours and they are a significant concern for reliable differential GPS (DGPS) applications and for satellite based augmentation systems (SBAS) such as the Federal Aviation Authority's Wide Area Augmentation System in North America. For the navigation and surveying communities, there is a need to identify the presence of SED effects and assess the expected impact on GPS users. Yet the distribution and evolution of SED is neither well understood nor adequately observed in a global sense, making SED effects hard to predict and ameliorate [Kintner et al., 2008].
 Foster and Rideout performed a detailed study of the magnetic conjugacy of SED for several storms. They observed that most, but not all, features of SED exhibited elements of magnetic conjugacy and appeared simultaneously in both hemispheres. For example, the entry of the SED plume into the polar cap near noon is observed in both hemispheres. They interpreted this conjugacy as an indication that storm-time electric fields are at least partially responsible for the generation of SED. A feature of SED that does not appear to be conjugate is the magnitude of TEC enhancement at the base of the SED plume. The amount of TEC enhancement appears to exhibit localized and longitude-dependent features, perhaps due to the offset of the geographic and geomagnetic poles in the American sector [Coster et al., 2007]. The offset of the geographic and geomagnetic poles in the Southern Hemisphere places the MWA at a special longitude for investigating ionospheric space weather disturbances, as described in section 1. Many of the consequences for the SED development in the American sector should also be observed in the Australian sector over the MWA. Studying the effects of SED in this longitude sector will address questions concerning the role of the offset of the poles.
 To date, there has been limited evaluation of SED effects in Australia. This is primarily because there was a relative paucity of GPS receivers in the Australian continent during the last solar maximum (1998–2003). Yizengaw et al. studied the space weather effects observed during the 31 March 2001 storm using data from 6 GPS receivers in Australia (plus 4 additional receivers). They reported that a tomographic reconstruction (based on GPS TEC) indicated equatorward propagating finger-like ionization features with 600 km height extension and a width of 300–400 km. Other SED events have been observed over Australia and were described bySkone and Coster .
 We report here on observations collected by the MWA GPS receiver during a recent geomagnetic event, one of the first major geomagnetic storms of the new solar cycle, on 26–27 September 2011. Differential TEC maps from the MWA will not be produced by the MWA until the full 128 tile system is working, so there are no corresponding ionospheric observations from the MWA during this time period. We show here the GPS vertical total electron content (TEC) results collected on the two storm days. Figure 2 plots the vertical TEC estimated by each satellite in view during this time period at the MWA site. The satellite estimates of TEC are color coded by magnetic latitude. Red values indicate higher magnetic latitudes (as high as −38 degrees) while blue values represent lower values (as low as −58 degrees). On September 26, 2011, the F10.7 cm solar flux index reached a value of 148, significantly higher than it has been in recent years and the Kp index, an indicator of geomagnetic storm strength, reached a high value of 6.3. The peak of the storm occurred between 16 and 22 UT on the 26th, but this occurred during local night at the MWA when TEC is nominally low. Prior to 13:00 UT on the 26th, the geomagnetic activity was fairly quiet.
 A little after local noon on the 27th (approximately 4 UT), still during the storm, the peak daytime values of TEC at the MWA reached nearly 60 TEC units. At this time, the geomagnetic activity was still disturbed, with the Kp = 5-. The excursion in TEC between the northern and southern magnetic latitudes is considerably larger than the day before (up to 20 TEC units), and is indicative of fairly large TEC gradients, which may be difficult to reproduce accurately enough with the differential TEC maps used for the MWA radio astronomical calibration solutions. We anticipate observing more storm time conditions as we move into the coming solar maximum.
3.2. Traveling Ionospheric Disturbances (TIDs)
 The MWA differential TEC maps will also provide important information about the characteristics of traveling ionospheric disturbances (TIDs) in the Southern Hemisphere. TIDs have been detected using the differential 73.8 MHz phase observations of the radio source Virgo A at the Very Large Array (VLA) in New Mexico [Perley and Bust, 2002]. At the MWA, the differential TEC maps will essentially be able to image the movement of TIDs across the array. TIDs are manifestations of middle-scale ionospheric irregularities arising as response to acoustic-gravity waves. Though TIDs are frequently observed at high and middle latitudes, their activity and amplitudes vary widely depending on latitude, longitude, local time, season, and solar cycle (e.g.,Kotake et al., 2006; Hernández-Pajares et al., 2006]. Large variations in TID characteristics and propagation directions are not surprising as they can be generated by very distinct classes of processes, including auroral sources, solar terminator passage, geomagnetic storms, hurricanes and tornadoes. Understanding the sources, energetics, and scale sizes of energy coupling from the lower atmosphere to the upper atmosphere, along with resulting ionospheric effects, is critical to a detailed knowledge of overall upper atmospheric energy balance. Current community investigations in this area include gravity wave excitation mechanisms [Fritts and Alexander, 2003] and implications of observed gravity wave geographical and temporal variability [Allen and Vincent, 1995].
 A number of recent papers using differential GPS techniques [Tsugawa et al., 2006, 2007; Shiokawa et al., 2005; Saito et al., 2002; Otsuka et al., 2004; Afraimovich et al., 2000; Hernández-Pajares et al., 2006] have contributed to an increased understanding of TIDs. Tsugawa et al. recently showed that the daytime TIDs in the U.S. propagate southeastward right after dawn until around mid-afternoon, and then change direction to southwestward in the late afternoon. These TIDs had 10–60 min period and wavelengths of 300–1000 km. Peak-to-peak amplitudes were larger than 0.5 TECU, and the amplitude of these daytime medium-scale TIDs (MSTIDs) seemed to increase with equatorward travel. Knowledge of the temporal and spatial characteristics of TIDs will be used to help identify possible source mechanisms. Comparison of TID structure to scintillation statistics at VHF and L-band can also identify those mechanisms most effective at producing small-scale irregularities in the ionosphere.Figure 3 shows a small TID (amplitude of 0.1–0.2 TEC) measured by differencing TEC values estimated for one GPS satellite (svn 13) by two receivers separated 30 km apart near the MWA site. Our goal is to characterize TIDs using GPS observations at the site, and then to compare them with MWA observations.
Figure 3. Example of a traveling ionospheric disturbance measure August 8, 2008 between two GPS receivers spaced approximately 30 km apart near the MWA site.
Download figure to PowerPoint
3.3. Faraday Rotation
 The 128-tile MWA will be able to demonstrate that MWA-like instruments can observe and track coronal mass ejections (CMEs), shocks and other plasma structures from points close to their inception near or on the solar surface, and in addition, will be able to characterize their magnetic field (J. Bowman et al., Science with the Murchison Widefield Array, manuscript in preparation, 2012). This latter characterization – knowledge of the direction of the magnetic field within the solar wind – has significant implications for the prediction of space weather effects on the Earth. Currently no instrument exists capable of measuring the direction of the magnetic field within the solar wind from near the surface of the sun to the L1 libration point where the ACE satellite is located. Knowledge of the direction of the magnetic field within the solar wind is important for understanding the geo-effectiveness of a particular CME or solar event. This information is available from the ACE satellite, however this only provides an approximate one-hour advance warning of impending geomagnetic activity. MWA-like instruments offer the promise of tracking the solar magnetic field from the corona into interplanetary space by measuring the Faraday Rotation, or Rotation Measure (RM), of linear-polarized radio sources.
 Faraday rotation depends on the wavelength (λ) of the propagating electromagnetic wave, and is due to the fact that left and right-hand circularly polarized waves propagate at different speeds in the direction of the magnetic field. As any linearly polarized wave can be described as a superposition of both a left and right-hand circular polarized wave, the end result after propagating a certain distance in the direction of the magnetic field, is a rotation of the linear polarization by the amountβ:
where RM is the rotation measure, and λ is the wavelength. The rotation measure, RM, is defined by:
is the density of electrons at each point s along the path;
is the component of the magnetic field in the direction of propagation at each point s along the path;
is the charge of an electron;
is the speed of light in a vacuum;
is the mass of an electron;
is the vacuum permittivity.
 The integral in equation (3) is taken over the entire path from the source to the observer. This means that the FR effect measured by the MWA will be a combination of FR from several regions: the ionosphere, the plasmasphere, magnetosphere, the solar wind, and the heliosphere. The predicted Faraday Rotation from these regions is discussed in detail by D. Oberoi and C. J. Lonsdale (Media responsible for Faraday Rotation: A review, submitted to Radio Science, 2012). However, the main point is that to obtain an accurate method of estimating the FR due to the heliosphere, the contributions to FR by all other media must be removed. The work we describe here has the goal of enabling the removal of the ionospheric component of FR, which is one of the largest, if not the largest, contributor to the FR.
 The MWA is fully polarization capable, with two orthogonal linear polarizations available at each antenna. In recent experimental campaigns at the MWA site, observations have recorded the effects of ionospheric Faraday rotation at VHF during beacon satellite overflights. The material presented here documents the progress made toward capturing and analyzing the Defense Meteorological Satellite Program (DMSP) F15 radio beacon signal, the only known active linearly polarized radio beacon in orbit and accessible to the MWA.
 Since the late 1970s, satellite communication platforms have moved away from linearly polarized transmissions, such as used by the Applications Technology Satellite (ATS) geosynchronous series, to circular polarization. However, one important quasi-linearly polarized radiator remains operational in orbit. The U.S. Department of Defense maintains DMSP as a set of satellites in nearly circular orbits at approximately 840 km altitude in two sun-synchronous planes [Strom and Iwanaga, 2005]. One of these satellites, DMSP F15, was launched in December 1999 and is locked to approximately 0900 / 2100 h local time. Of relevance for MWA purposes, the beacon antenna is monopole oriented in the ram orbital direction, transmitting a nominally linearly polarized coherent 1-watt CW beacon tone at 150.012 MHz.
 Observations of a few high elevation transits of the DMSP F15 satellite have been carried out during 2009 to 2011 expeditions to the MWA site. We report here on a high elevation DMSP F15 pass collected on August 05, 2009 from 10:30–10:45 UT. Complex baseband signals from both linear polarizations on MWA tiles 3 and 11 were used to compute Stokes parameters [I, Q, U, V] and subsequently to derive fractional polarization and related angular information. We restrict our attention to the 410–460 s time window, corresponding to the central part of each tile's main antenna lobe where reasonable qualitative comparisons are possible. This time period is indicated between dashed lines in both Figures 4 and 5. Figure 4 shows the polarization fraction, degree of polarization ellipticity β, and the polarization vector's position angle χ [Goldstein and Collett, 2003]. For the general case of elliptical polarization, parameter β defines the degree of ellipticity of polarization (β = ±π/4 implies circular polarization and β = 0 or ±π/2 implies linear polarization) and χ is the angle between the major axis of the ellipse and the X polarization basis vector.
Figure 4. (top, middle, bottom) The fractional polarization of the signal, degree of ellipticity (β) where 45° is circular and 0° is linear polarization, and the position angle (χ), where this is measured between the major axis of the ellipse and the X polarization basis vector. These data come from Tiles 3 and 11. The transit (the highest elevation in the pass) is at ∼435 s.
Download figure to PowerPoint
Figure 5. (top and bottom) Estimated Rotation Measure (RM) and Faraday rotation. The RM predicted by the model described in the text shows a steady increase from ∼0.1 to about ∼2.2.
Download figure to PowerPoint
 Figure 4 indicates that a nearly completely polarized signal was measured during the tile main lobe period. This is consistent with the signal being highly coherent and dominated by the satellite beacon tone. β and χ follow smooth trends throughout this period as expected due to the changing relative satellite position. We note that the significant DMSP F15 beacon ellipticity measured at the MWA (nonzero values of β) is the result of projection effects between the satellite ram direction and the local tile vertical, combined with effects related to the placement of the monopole transmitting antenna on the conductive spacecraft body. Although not shown here, the total received power (Stokes parameter I) shows an amplitude oscillation of approximately 1.2 to 1.5 dB with a period between 2 and 2.5 s. This signature propagates into variations in the derived polarization parameters shown in Figure 4, and has been independently observed on multiple occasions from separate NRAO Green Bank beacon observations tracking DMSP F15 with a 43 m antenna [Anderson et al., 2011; P. Erickson, private communication, 2011]. Potential causes include slight uncorrected orbital oscillations or amplitude variations in the master beacon oscillator onboard the satellite.
 The Faraday rotation effect causes the linearly polarized fraction of the DMSP F15 beacon radiation to rotate its plane of polarization. The total rotation is a spatially and temporally dependent function of the electron density profile and the terrestrial magnetic field strength and orientation along the line of sight between the MWA tile and the satellite platform.
 To estimate these effects, at each point during the pass, we use DMSP F15 orbital information to calculate azimuth and elevation from the MWA station. The total Faraday rotation along the line of sight from the MWA to the DMSP F15 satellite is subsequently calculated using a simplified version of the Appleton-Hartree propagation equation, with time dependent parameters provided by the MIT Haystack Observatory Madrigal atmospheric science distributed database and computation engine (http://madrigal.haystack.mit.edu). For these calculations, Madrigal estimates terrestrial magnetic field strength and direction using the International Geomagnetic Reference Field (IGRF) 11th generation [International Association of Geomagnetism and Aeronomy Working Group, 2010], while the electron density variation as a function of altitude is taken from the International Reference Ionosphere's IRI-2007 model [Bilitza and Reinisch, 2008] for the MWA location and time of day. The resulting Rotation measure (RM) and Faraday rotation at 150.012 MHz for the DMSP F15 pass are shown in Figure 5, where the Faraday rotation angle = λ2 × RM. The rotation rate is seen to be faster for the later (northern) portion of the pass, although the range and hence the integration length is symmetric about its minimum value at the transit point. At the start of the pass, the Earth's magnetic field has a very significant tilt with respect to the low elevation line of sight to the satellite (magnetic aspect angle = 70 deg). As the satellite moves northward on its near polar orbit, the Earth's line of sight to the satellite becomes more closely aligned with the local magnetic field and therefore provides a larger weight in the Faraday rotation integration.
 The prediction of increasing Faraday rotation as a function of time should manifest itself as an increase in the measured position angle χ. The predicted Faraday rotation angle changes by ∼40° (Figure 5, bottom), compared with the measured polarization angle χ change of ∼45° (−50° to −5°; Figure 4, bottom) in the same period. The two are in good agreement, and show similar curvature in their respective trends. We conclude that observed trends in the variation of χ are consistent with expected signatures of ionospheric Faraday rotation.