Bielefeld GLOW Science Projects
In the last three years, the single LOFAR stations in Germany have been equipped with a reliable, sensitive and highly capable observing system, which has allowed weekly monitoring of a large number of pulsars. Currently 114 pulsars (of which nearly 20 are recycled) are being monitored on a weekly basis; and have been monitored for anywhere between one and three years. This vast amount of data enables a host of experiments at unprecedented levels of precision, as discussed below. So far, most of the GLOW data analysis has been performed as part of undergraduate thesis projects and none of this work has been published yet, but several publications are in preparation, based on the thesis work done to date.
- Measurements of the structure of the ionised interstellar medium (IISM): an initial analysis of the GLOW data from the first two years of observing, including data on 80 different pulsars, has demonstrated that our GLOW data provide unprecedentedly precise measurements of interstellar dispersion (and thereby the integrated interstellar electron density or "Dispersion Measure", DM):
Histogram of median values for DM measurements of GLOW data, for 80 pulsars. The sample contained 11 MSPs (in dark red) and 69 slow pulsars (in light red). The median DM precision we get for the MSPs is indicated with the vertical dark red line, at ~2x 10^{-4} pc/cm^3, which is a factor of a few more precise than the most sensitive measurements published in literature (You et al., 2007; Keith et al., 2013). The entire population of pulsars has a median DM precision of slightly better than 2x10^{-3}pc/cm^3, which is orders of magnitude better than the two comparable samples from literature (Hobbs et al., 2004 and Petroff et al., 2013). With this unprecedented precision on this rather sizeable source sample (the sample has since been expanded to 114 pulsars), the structure of the IISM can be probed like never before. (Source: Donner B.Sc. thesis, 2014.) - Measurements of the smallest density scales in the IISM: for some of the brightest pulsars in our sample, day-long observations have demonstrated the potential to probe IISM density variations on scales that are much smaller than any previously observed. This allows a more detailed view of the IISM, which may prove to be a missing link between large-scale structure typically observed with DM monitoring (see the previous bullet point) and small-scale structure typically investigated in secondary spectra (see the next bullet point). The following figure shows an example of intra-day DM variability observed with the Effelsberg LOFAR station.
Interstellar dispersion (as quantified by the dispersion measure, DM) towards pulsar B0329+54, during a day-long observation at the Effelsberg LOFAR station, DE601. Each point in this plot represents 20 minutes of data on this brightest pulsar in the sky. All points are independent so any trends are caused by physical structure in the IISM. Of particular notice is the steep gradient at the start of the day (a drop of ~1.5x 10^{-3} pc/cm^3 in about 5 hours). The large scatter of the points may either be caused by imperfections in our data analysis (potentially related to moding behaviour of the pulsar), or by physical variations on shorter timescales still -- or a combination of these. While modes did occur in the second half of this data set, the first five hours were entirely mode-free. A deeper investigation into short-term DM variability of this pulsar is currently being undertaken. (Source: Lachetta B.Sc. thesis, 2014.) - Measurements of scintillation arcs: on the smallest scales, inhomogeneities in the IISM cause faint interference patterns known as "scintillation arcs". These are features that can be detected in the two-dimensional Fourier transform of the dynamic spectrum, which is a power-spectrum of the pulsar, plotted as a function of time and frequency. These arcs have previously been studied at the most sensitive telescopes in the world (primarily Arecibo), typically at wavelengths of a few hundred MHz. We have been able to detect such arcs in a growing number of pulsars in GLOW data. Detection of these arcs across a wide range of frequencies allows further detailed investigations into the propagation and refractive effects that cause the interference; and the structures that cause it. Specifically, the low-frequency GLOW data has sensitivity to material at far larger angular separations than the higher-frequency arecibo data and may therefore be used to probe a much larger part of the Galaxy, provided these arcs can be detected in a sufficient number of pulsars.
- Pulsar scintillation parameters have long been one of the more numerous inputs to models of interstellar gas and turbulence distributions (e.g. Cordes & Lazio, 2002). Relatively little research has been done on the variability of the scintillation parameters, however, and how that relates to the turbulent IISM's properties. Our GLOW data provides unprecedented precision due to its high fractional bandwidth and the small scintillation bandwidth at LOFAR frequencies, allowing many scintles to be detected at any time. Furthermore, the flexible GLOW hardware allows almost arbitrary frequency resolution, allowing scintles to readily be resolved. An example of our sensitivity to pulsar scintillation is shown in the following figure.
Dynamic spectrum (pulsed intensity as a function of time and frequency) for PSR B0809+74 over a 24-hour long observation. Not only can the scintles (the small "islands") be very clearly detected -- in large numbers -- but furthermore can the frequency scaling of the scintillation parameters (scintillation bandwidth and timescale) be measured directly from these data. Monitoring of these parameters allows new investigations of the structure and turbulence in the IISM. (Source: Jung B.Sc. thesis, 2014.) - Heliospheric studies: Our GLOW data are not only highly sensitive to the interstellar plasma, but also to the interplanetary plasma. Given our high sensitivity to interstellar dispersion (see above) and to Faraday rotation (see Section 6b), our data can be used to probe the density and magnetic field of the Solar wind. Furthermore, given the high cadence of our observations (we normally perform weekly observations), the resolution in our Solar mapping is also unprecedented. Figure BB indicates our potential. Following the initial investigation (part of which is shown in the figure below), this project was taken to be the centre piece in an application for a Humboldt fellowship of our new post-doc, Caterina Tiburzi.
Excess dispersion towards PSR J0034-0534 as a function of the angle between the pulsar and the Sun. Shown are the integrated electron density between the Earth and the pulsar, _after_ subtraction of a model for the Solar wind, for three consecutive passages of the pulsar behind the Sun. Clearly the standard spherical model of the pulsar wind does not fit well for this pulsar; and in 2015 some hints of smaller-scale structure appear to be visible. Note, however, that the scale of these deviations from the model (a few 10^{-4} pc/cm^3) is small enough that this would go entirely unnoticed in any more traditional (i.e. higher observing frequency, smaller fractional bandwidth) observations of this pulsar. (Source: von Kamen B.Sc. thesis, 2015.) - Pulsar Moding: Because of the large amount of telescope time available, exceptional projects are possible which would be too prohibitive in their observing needs to be accepted on more traditional telescopes. Specifically we have commenced the investigation of pulsar moding, with a focus on PSR B0329+54. So far, two B.Sc. thesis have covered this topic, fed by a number of (multi-)day-long observation(s). First, we investigated the moding timescale, i.e. the amount of time it takes for this pulsar to switch from a normal state into a moded state. Our results were inconclusive due to small-number statistics, but preferred a gradual switch to a sudden switch, counter to standard theories. A second investigation applied a series of different analysis tools to attempt and create an automatic algorithm for mode-identification, allowing moding investigations to be undertaken on a larger amount of data more easily. Specifically, the Principal Component Analysis (PCM) presented by Oslowski et al. (2013) turned out to be highly useful at automatically separating sub-integrations with moded pulses from those with unmoded pulses. We are now applying this analysis to all available data on PSR B0329+54 in order to derive highly precise moding statistics for this pulsar at LOFAR frequencies.