Filename
stringlengths 22
64
| Paragraph
stringlengths 8
5.57k
|
|---|---|
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | where the filter FDOP is linearly interpolated for DOP values in between 0.4 and 0.6. We further filter signals based on ellipticity, since we are only interested in rectilinear body wave arrivals. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | which is then applied to the data. By selecting appropriate time and frequency windows for the P and S-wave arrivals, the polarization analysis presented is used to provide back azimuths for a large number of LF family marsquakes. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | The most prominent noise source are wind gusts, which excite long period energy with predominantly horizontal polarization (Stutzmann et al., 2021). The aforementioned 2.4 Hz resonance is not strongly excited by local wind, but by marsquakes and thus visible for most quakes that have energy above the noise around 2 Hz. It is near-vertically polarized (Hobiger et al., 2021), with a roughly 180◦ azimuth (Dahmen et al., 2021). The 2.4 Hz azimuth is independent of event polarization, but its exact azimuth seems to be methodology-dependent (Dahmen et al., 2021). In addition to the 2.4 Hz resonance, there are a number of additional spectral peaks in the data. Below 9 Hz, there are resonances that are related to the measurement system (termed 1 Hz tick noise (Zweifel et al., 2021) and its harmonic overtones) and to the lander itself (permanent lander modes at 1.6 Hz, 3.3 Hz, 4.1 Hz, 6.8 Hz, 8.6 Hz; transient lander modes at many other frequencies throughout the mission, Dahmen et al. (2021)). Since these spectral peaks have their own stable azimuth, it is important to be aware of these and select frequency bands where they are not present. Though visible, the lander modes and 2.4 Hz resonance do not affect the analysis of the data since we focus on signal below 1 Hz in this study. The tick noise is removed in this analysis by averaging over a small frequency band around 1 Hz, so the influence of the 1 Hz tick noise is minimised. A further very significant source of polarized noise are the transient pulses that proliferate in the data termed glitches. They generally appear below 1 Hz with 25-30 s duration and have a strong and linearly horizontal polarization (Scholz et al., 2020). They are a major source of noise when analysing the polarization of a seismic signal, since even weak glitches show stronger polarization than seismic events. It is therefore important to select time windows that avoid glitches. We analyse the back azimuth in a select frequency band for each event, based on the event amplitude and P and S-wave inclination. Further, they are chosen such that they minimize the effect of any low frequency glitches and avoid known spectral peaks. By default, we also select a time window around the P and S picks identified by MQS from -5s to +10s. A noise time window is provided by MQS for each event. We calculate the kernel density estimation (KDE, Scott, 1992) across the specified time windows in the frequency band. The KDE curve maximum of the P-wave is the basis of the back azimuth determination. The full width at half maximum (FWHM) of the peak is calculated and provides a consistent uncertainty estimation. The KDE peaks of the S-wave are used to corroborate the results obtained from the P-wave. However, the S-wave is mainly assessed in the more complex 3-D analysis. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | where HP is the 2-D histogram of the P window in the selected frequency range for inclination and azimuth (θ, φ). We add a ’water level’ to HP to prevent it from dominating the result, since a large part of the inclination-azimuth space of HP is zero. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | We first test our method on a set of synthetic marsquakes. For a best-case comparison with real marsquake data, we use noise recorded from a quiet evening without marsquakes and add a synthetic event to the data. This noise data includes the typical signals (e.g. glitches, tick noise), that are thus handled the same as real marsquakes. Sol 245 (a sol is a Martian day) has low noise conditions and several hours of quiet data available. We use the Mars model InSight KKS21GP (St¨ahler et al., 2021) to generate synthetic waveform data (see Data section). We use a frequency band between 0.3–1.0 Hz for the determination of the back azimuth, as this catches the main energy of the synthetic event while avoiding noise contamination at low frequencies. Since this is a 1-D model, the position of source and receiver is not relevant and only distances, magnitude, and source depth are relevant factors. The source is placed to the east of the receiver at 50 km depth, making the true back azimuth (BAZtrue) 90◦. In the example presented in this section, the distance is 40◦. We use a focal mechanism of strike = 280◦, dip = 79◦, and rake = -79◦, similar of that estimated for marsquake S02035b by Brinkman et al. (2021). The moment magnitude Mw is set to 3.5, which is expected for medium to high quality events. The observed marsquake data differ from these synthetic seismograms mainly by strong scattering, which decreases the amount of polarization in the body wave coda and suppresses surface waves. For the initial arrival window of body waves, these parameters should make this synthetic event comparable to high quality marsquakes. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | geographic directions. The gray arrow marks the direction of BAZtrue, the blue arrow marks the direction of the polarization-derived back azimuth (BAZpol). |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | The event is clearly visible at frequencies below 1 Hz. The maximum amplitude is -185 dB for the P and -175 dB for the S-wave. The noise window has amplitudes between -200 to -185 dB, with higher amplitudes at lower frequencies. In the preferred target frequency band we use to determine the back azimuth, between 0.3–1.0 Hz, the noise window has amplitudes more than 20 dB - 1 order of magnitude - below that of the signal windows. The SNR of the P window between 0.3–1.0 Hz is 6.2. When taking a closer look at the time-frequency plots in rows 2–3, parts of the signal have been filtered out. At 18:52:30 UTC, there is a small glitch visible in the noise window with east-west polarization. The P and S window are strongly polarized up to 1 Hz. The P-wave polarization persists for around 1 minute after the arrival of the signal. The S-wave polarization is shorter and more focused around the arrival of the wave. The back azimuth estimated from the KDE of the P-wave is 90◦, equal to BAZtrue, with an uncertainty of 83–98◦. The KDE maximum of the S-wave is around 200◦, roughly orthogonal to the P-wave. The noise window has no clear peak or preferred azimuth. As expected for an event at this distance with near vertically incident body wave arrivals, the P-wave has a steep inclination of 50–80◦, which contrasts with the S-wave with an inclination close to 0◦. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | In summary: Figure 6(b) shows that there is a clear difference in polarization between the pre-event noise window, the P-wave arrival, and the S-wave arrival. The properties are as expected: the P-wave has a high inclination, rectilinear signal in direction of BAZtrue; the S-wave has a low inclination, rectilinear signal with shifted azimuth. The noise window is not strongly polarized and has no preferred azimuth as described by Stutzmann et al. (2021) - though some noise features of the InSight dataset are known exceptions, including glitches. The back azimuth of the event estimated from the P window matches BAZtrue. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | The azimuth of the three selected time windows (noise, P, and S) is analysed in more detail in Figure 6(c). The top and bottom rows show histograms of the semi-major azimuth on the angular axis against the inclination on the radial axis. They show the data for a lower (0.3–0.65 Hz, top row) and higher (0.65–1.0 Hz, bottom row) frequency band. To mimic a stereographic projection, the inclination is 90◦ in the middle of the plot and goes to 0◦ at the outer edge. We see that the noise window (left) has a low inclination and a wide range of azimuths, consistent with wind-induced noise (Stutzmann et al., 2021). To help analyse the seismic phases, we mark the P-wave vector in the P and S window plots. It is described by an azimuth and an inclination, which are taken from the maxima of the respective KDE in Figure 6(b). The vector’s intersection with the half-sphere reduces it to a dot. We expect the S-wave to be perpendicular in 3-D space to this vector, i.e. located on a plane with the P-vector as the normal vector. Thus, we calculate the plane perpendicular to the vector and its intersection with the half-sphere, which results in a curve. These two features are plotted in blue in Figure 6(c). For the middle column (corresponding to the P window), we expect the data to cluster around the blue dot. For the right column (corresponding to the S window) the data should cluster somewhere on the blue curve. The data is clustered around the P vector dot in the middle column as expected. The right column shows three clusters, south, north, and northwest, with very low inclination visible in both frequency bands. They lie close to the blue curve and are thus perpendicular to the estimated P-vector. They agree with our predicted S-wave and we consequently interpret these as the S-wave. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Figure 6(d) shows the predicted P-wave based solely on the polarization contained in the S window. The background shading is calculated using Eg. 7, where dark areas have a higher ’likelihood’ for the P-vector, i.e. larger parts of the S window polarization are perpendicular to these areas. The maximum is marked by a red dot, giving the P-vector as predicted by the S window polarization alone. For comparison, the P window KDE-derived P-vector is given by the blue dot. We see in Figure 6(d,top) that these two dots agree well with each other, and the main difference is the inferred inclination of the P-vector. Figure 6(d,bottom) combines the information from the P and S windows by multiplying the azimuth-inclination histogram of P between 0.3–1.0 Hz (Fig. 6(c) combined middle column) with the background from Figure 6(d,top) as described in Eq. 8. Marked in cyan is the resulting P-vector. Since these two results are so similar, combining them to obtain a third set of P-vector offers little improvement for this synthetic case. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | To test the method on events with known back azimuth, we take a dataset recorded at the Warramunga array, station WRAB, located in Tennant Creek, Australia, which has similar topographic conditions as InSight on Mars (i.e. flat surroundings and no close-by mountains) (Scripps Institution Of Oceanography, 1986). Similar to the results in the previous section, we look at the polarization of an earthquake which occurred in Indonesia on April 5th, 2011. The source was at 3.03◦ N 126.94◦ E, in 20 km depth with Mw = 5.9 (Survey, 2011) at an epicentral distance of 2666 km. Figure 7 shows the polarization analysis with the same format as shown in Figure 6. The event is clearly visible on the vertical waveforms (Fig. 7(a)). There is excess energy up to 3 Hz for the P-wave visible in the spectrogram (Fig. 7 (b,top), first column). The S-wave is visible at lower frequencies and has strong excess energy around 0.1 Hz. The amplitude KDE shows strong excess energy for both phases, with an SNR for the P-wave of 48.5. When considering the polarization-filtered azimuth in the second row, the event is clearly visible in the time-frequency plot on the left. The P-wave in particular has a strong, consistent polarization which differs from the noise. The S-wave polarization appears slightly weaker, and shows no dominant azimuth above 1 Hz. The back azimuth estimated from the KDE (right column) of the P-wave is 346◦ with an uncertainty of 343 − 349◦. The estimated back azimuth is very close to the true back azimuth of 342◦. The S window has several peaks, roughly along and opposite to the true back azimuth. The noise window is differently polarized from the earthquake signal. The P window has an (primarily moderate to high) inclination of 40 to 60◦. The S window has a low inclination between 0–20◦. In contrast, the noise window has a wider distribution with less prominent singular peaks. Figure 7(c) provides a deeper analysis of the azimuth of the time windows, where as before, the plot shows azimuth against inclination for two frequency bands. The true back azimuth is marked with a gray line. The noise window has a distinctly different polarization from the P and S windows. The P window matches the true back azimuth (gray line) well in both frequency bands, and consequently, the P vector (blue dot) estimated from the polarization analysis lies very close to the true back azimuth. Several clusters in the S window lie close the blue curve, meaning they are orthogonal to the P-wave in 3-D space. It appears that the KDE-derived P-vector is too low (around 50°) for an optimal fit with the S window polarization. A more vertical inclination for the P would shift the blue curve towards lower inclination, improving agreement with the S window. The fit with the expected orthogonal constraint is good in both frequency bands. In the higher frequency band, there is some additional horizontal energy along the back azimuth direction. Figure 7(d,top) shows the estimated back azimuth from the S window. The P-vector inclination expected from the S window is much steeper than estimated from the P window itself, but the back azimuth is consistent between both. Combining the P and S window (Fig. 7(d,bottom)) gives a result very similar to that of the P window, and thus close to the true back azimuth. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | While the analysis of single events is useful for comparison with marsquakes, in particular regarding the combination of inclination and azimuth, a more systematic approach is needed to assess how the method performs on a larger scale. We analyse a set of 264 earthquakes recorded on the same seismic station (WRAB) to assess the error in estimated back azimuth. Systematic errors could be related to a local site effect. The set of earthquakes has distances between 20◦ and 40◦, source depths of 10 to 30 km, Mw of 5.5 to 6.0, and occurred between beginning of 2010 and end of 2017. These parameters should make the results comparable to Mars (i.e. adjusted magnitudes to account for Earth’s higher noise level, relatively shallow sources) and the epicentral distance covers several seismically active regions from different directions. We use the maximum of the P window KDE to estimate the back azimuth, and we estimate the uncertainty using the peak-half-width. All earthquakes are analysed between 0.3–1.0 Hz. Figure 8 (a) shows the true back azimuth BAZtrue versus the polarization-derived back azimuth BAZpol. The error is calculated as ∆BAZ = BAZtrue−BAZpol, so positive (negative) ∆BAZ values correspond to too small (large) estimated BAZ values. The black bars show the estimated uncertainty. Earthquake magnitude is given by the marker color. A histogram of the back azimuth error is shown in Figure 8 (b). The polarization analysis matches BAZtrue well, with an error of less than 10◦ in 213 out of 264 earthquakes. Only 13 events have an error larger than 30◦. There is no clear trend in the error depending on the back azimuth of the earthquakes, so no obvious site effect is visible in this analysis. The distribution of earthquake back azimuths is, as is expected, not uniform but clustered, related to active fault boundaries near Australia. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | After testing the method on synthetic marsquakes and a number of earthquakes, we apply the polarization analysis to a set of LF and BB marsquakes. The polarization analysis for marsquake S0235b is seen in Figure 9. Same as before, the vertical velocity waveforms are shown bandpass filtered between 0.3–1.0 Hz (Fig. 9(a)). In (b), we show the time-frequency polarization analysis. The amplitude spectrogram in the top left is in velocity, in contrast to the acceleration spectrogram shown in Figure 2. In addition to the P and S-waves, the 2.4 Hz resonance is visible with roughly southward azimuth and near-vertical inclination. Unlike many other marsquakes, S0235b has no glitches close to either the P or S pick. The top row in (b) shows the large signal to noise ratio of the event. Therefore, the polarization for the P and S window should be solely due to the event and not be dependent on background noise. The KDEs for the azimuth ((b), right column, second row) show a distinct peak for both P and S windows, and a flatter distribution for the noise window. The three time windows differ strongly in inclination: the S window has a very low inclination while the P as a steeper inclination. The noise has a broader distribution. The estimated back azimuth from the polarization matches the one from the MQS catalog (gray dashed line, InSight Marsquake Service, 2022). We examine the three time windows more closely in Figure 9(c). The MQS back azimuth is marked with a gray line. The noise shows no preferred azimuth between 0.3–1.0 Hz (Fig. 9(c), left column). The P window shows a clear preferred azimuth over the whole analyzed frequency band. The P window signal has a steep inclination, best visible between 0.65 and 1 Hz. At lower frequencies, the inclination also contains some more horizontal polarization. The S window has two discernible azimuths, roughly southwards and, less dominant, to the north-east. The latter is weakly visible in the higher frequency band, while the former is visible over the whole frequency range. The S window azimuth-inclination signal lies clearly on the blue curve, perpendicular to the estimated P-vector (blue dot). Figure 9(d) shows the expected P-vector as determined solely from the S window, and compares it with the P window-derived P-vector. While inclinations differ slightly, the back azimuth is consistent between both approaches. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | We apply the same methodology to a number of other marsquakes, all of the LF or BB event category (Clinton et al., 2021), which are selected for clear phase arrivals. We estimate the back azimuth from the maximum of the P-wave, and corroborate this with the S-wave, whenever possible. The back azimuth uncertainty is determined by the half-width of the P-wave KDE. Results are summarised in Table 1 and Fig. 10. A detailed polarization analysis of each event in the table can be found in the Supplementary Material together with a short description. The upper part of Table 1 shows results for the best marsquakes with clear phase arrivals and strong polarization. The lower part of the table shows results for events with less clear polarization. They often require a combination of P and S-wave polarization analysis, since the P-wave polarization alone is not sufficient, it is contaminated by significant glitches, or the phase arrival timing is uncertain. Results are therefore less certain. We find that while events with large SNR typically mean reliable back azimuths can be estimated, events with smaller SNRs can still provide back azimuths if the polarization of the event is sufficient. The SNR can also be affected by a glitch close the P-wave onset, or by phase picks with large uncertainty. We see that results obtained from the polarization analysis are close to all those provided by MQS, where available, but that we do find new back azimuths for 16 more quakes. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Figure 6: Polarization analysis of a synthetic marsquake with Mw = 3.5 and a distance of 40◦. (a) Vertical velocity waveforms with P and S arrivals (blue dashed lines). (b) Polarization analysis based on the velocity timeseries, (b,top) amplitude in (m/s)2/Hz [dB], (b,middle) azimuth, (b,bottom) and inclination. (b,left) time-frequency plots for each parameters, with three marked windows: pre-event noise (gray), P-wave arrival (blue), S-wave arrival (red). Phase arrival signal windows are -5 s to 10 s around the corresponding pick, and the noise window has 120 s duration. (b,columns 2–4) Histogram representations of each of the three windows marked in the left column, with the amplitudes on the x-axis corresponding to the scale shown in the time-frequency colorbars from the left column. Vertical dashed gray lines mark the true back azimuth. Shaded areas indicate the preferred frequency band (0.3–1.0 Hz). (b,right) KDEs calculated within the shaded frequency band for each time window. The back azimuth estimated from this analysis is defined as the peak of the P-wave azimuth. Dashed gray and solid red lines in the second row mark the true and calculated back azimuth, respectively. (c) Azimuth-inclination histograms for the time windows separated into two frequency bands (0.3–0.65 Hz, 0.65–1.0 Hz). KDE-derived P-wave vector and its perpendicular plane are marked by a blue dot and curve, respectively. (d) P-vector estimation from the S window (top), and combination of P and S window results (bottom). The S-derived P-vector is marked with a red dot. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Figure 7: Polarization analysis for an earthquake recorded at station WRAB, Australia on April 5th, 2011. The plot follows the same structure as Fig. 6. The earthquake has Mw = 5.9, and the back azimuth is 342◦. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Figure 8: Error of polarization-derived back azimuth for earthquakes recorded at station WRAB, Australia between 2010 and 2017. Shown are (a) the error in back azimuth versus the true back azimuth, and (b) a histogram of the error distribution. Also depicted are the earthquake magnitude Mw (color dots) and estimated uncertainty (black lines). Mw is between 5.5 and 6.0 for all events, resulting in a total of 264 events. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Figure 9: Polarization analysis for marsquake S0235b (Mw3.7 BB QA event at 27.9◦ distance, July 26, 2019). The plot follows the same structure as Fig. 6. The signal windows are -5 s to 10 s around the corresponding wave pick. The noise window is used from MQS, with 261 s length. Pick uncertainties are taken from MQS and are indicated by horizontal bars (b, top row, left column). The MQS back azimuth is marked with a gray dashed line in the second row in (b), and with a gray line in (c). The KDE maximum of the P window in (b, right column) is marked with a red line and diamond indicator. The 1 Hz tick noise has been reduced to avoid contamination of the polarization analysis. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Table 1: Table summarising marsquake back azimuth estimates. The marsquake events, type, quality, and MQS back azimuth are taken from the MQS catalog (InSight Marsquake Service, 2022). MQS uncertainties are described in Clinton et al. (2021). Results from Drilleau et al. (2022) are given in ’Drilleau (2022)’. SNR is calculated from the P-wave and indicated noise window in the same frequency band where the back azimuth is estimated. Uncertainty for the polarization back azimuth is calculated from the width of the KDE at half-maximum. Events in italics have less clear results and are more speculative. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | S0105a LF S0154a BB S0167b LF S0205a BB LF S0290b LF S0325a LF S0407a LF S0409d S0474a LF S0484b BB S0784a BB S0850c BB S0918a LF S1000a BB † Uncertainty set manually. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | We also show results from Drilleau et al. (2022). The authors present their own phase arrival times and do not use the MQS Catalog V8 phase picks, so the results are independent. We find that our results agree well, with the largest deviation for S0474a (21◦ (Drilleau et al., 2022) versus 97◦ (this paper)). For 4 events (S0154a, S0474a, S0484b, S0918a) our method delivers a significantly reduced uncertainty, increasing the value of the result for tectonic interpretation. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Marsquakes have generally low SNRs compared with earthquakes. To better compare the robustness of the method between different quake sets, we analyse the SNR of each quake and compare it to the error in azimuth estimation. For comparison, we test the terrestrial data set and a synthetic Martian one. We use the earthquake set from station WRAB described previously. Synthetic quakes are generated with magnitudes between Mw = 2.8 − 4.2 and distances between 20◦ and 40◦. This range is representative of intermediate to high quality quakes recorded on Mars. Both sets estimate the back azimuth solely from the P-wave KDE maximum between 0.3–1.0 Hz. Figure 11 shows the error in back azimuth estimation (taken as the absolute difference between true back azimuth and estimated back azimuth) versus SNR for earthquakes and synthetics, and the estimated errors (from the width of the KDE peak) and SNR of marsquakes. Further, we show the estimated error for earthquakes and synthetics for each event. The top plot in Figure 11 shows the synthetic marsquakes. Most events have errors less than 15◦, with SNR values ranging from around 2 to 80’000 (Figure axis is limited to SNR = 300). All synthetic quakes with error larger than 20◦ correspond to the lowest magnitude in the data set. Higher SNR values reduce the error in back azimuth estimation. The estimated error decreases with increasing SNR. The middle plot shows the earthquake set used previously, with 264 events in total. SNR ranges from around 1 to 1000. The largest deviations from the true back azimuth occur when SNR values are low, and decrease when SNR is increasing. Events with low SNR values can still have accurately estimated back azimuths however, with the majority of events having errors below 10◦. The estimated error is comparable to what is seen for the synthetics, though with a larger spread. However, there is no obvious trend visible which would link SNR and estimated error, and the very odd ’blunder’ - an event with a large true error with much smaller estimated errors - do occur even for high SNR events. The bottom plot shows marsquakes, which have generally lower SNR values. Back azimuth and SNR are calculated for an individual frequency range for each event depending on frequency content and glitch presence, making direct comparisons not straightforward in terms of SNR. The estimated errors are similar to what is estimated for the earthquake set, barring some outliers there. The marsquake set in general is comparable to low-SNR earthquakes in terms of estimated error. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | In contrast to the previously discussed LF and BB marsquakes, the HF and VF events have their main energy content above 1 Hz, sometimes reaching 30 Hz and higher. Since these high frequencies are dominated by scattering, and several prominent lander modes are present between 1 and 9 Hz, it is much more challenging to find a stable polarization. We try the same methodology but with an adapted frequency band of 0.25– 1.25 Hz for the KDE. This broader band, which extends to higher frequencies, should better capture the characteristics of these events. While their main energy is at high frequencies, for some larger events there is a tail of energy below 1 Hz. The amplitude of this differs between events. At frequencies much above our selected band, scattering is dominant and a stable polarization is not expected. Unfortunately, we do not find a stable back azimuth for any high frequency family event recorded so far. This suggests that the arrival of a ballistic body wave is not observed above noise level and that the first arrival is already strongly scattered (van Driel et al., 2021; Menina et al., 2021). |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | limiting tectonic interpretation from the MQS catalog. More complex techniques are needed to obtain back azimuth estimates for previously unlocated events, and to improve back azimuth estimates of already located events. The polarization analysis presented in this paper is capable of reliably estimating the back azimuth of seismic events with sufficient quality from the P and S-wave arrivals. Analysis of synthetic marsquakes shows that the back azimuth error is generally below 10◦ for quakes with distances and magnitudes similar as those recorded on Mars. For small SNRs (< 2), the error can reach around 40◦. While the background noise recorded on Mars contains several features which interfere with the analysis of seismic data, we find that the removal of the 1 Hz tick noise is sufficient to retrieve the true back azimuth in the desired frequency range. The earthquake set recorded at station WRAB in Tennant Creek, Australia, shows a similar pattern. Over 260 earthquakes are analysed, with SNRs ranging from 1 to 1000. Similar to the synthetic quakes, we see an improved estimated back azimuth for events with higher SNR. The majority of events have an error of less than 10◦, with maximum errors just over 40◦. The estimated uncertainty from the polarization analysis is larger for the earthquake data set than for the synthetic set. While the uncertainty covers the true back azimuth in the majority of cases, it is sometimes estimated too narrow. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | tick noise, glitches, resonances and lander modes, amongst others) and thus requires careful handling to extract the desired signal. We applied the same method to a set of high quality marsquakes to estimate their back azimuths. The polarization back azimuths are close to the back azimuths estimated by MQS and Drilleau et al. (2022) for almost all events. We also provide back azimuth estimates for a number of events where no MQS back azimuth is available. Most of these are more speculative however, owing to poorer event quality and/or low SNR. Some events are contaminated by glitches close to the phase arrivals. However, careful analysis of both P and S-wave polarization and considering the perpendicularity constraint, a back azimuth can still be estimated, albeit with less certainty. We place the majority of these previously unlocated events in a similar, eastern, direction as most of the better constrained events. We combine our results with distances provided by MQS (InSight Marsquake Service, 2022) to obtain locations for the events (Fig. 1). Within uncertainties, we therefore confirm that most seismic activity near InSight happens in the Cerberus Fossae graben system (Banerdt et al., 2020). However, we also identify new active regions. S0899d heralds a source region due North, in Utopia Planitia; S0290b and S0850c, lie north-west, near Isidis basin. S0205a and S0918a are the only events that seem to originate from southern hemisphere, though S0205a in particular is a challenging event to locate. Neither S0205a nor S0899d has a second phase pick, making their distances uncertain. As both events are similar to S0183a, we place them at similar distances (46°) (Ceylan et al., 2022). The two most distant events recorded by InSight, S0976a and S1000a, are located in Valles Marineris and the northern Tharsis region (albeit with a large uncertainty), respectively. S0167b could be located close to the boundary between Utopia Planitia and the polar Vastitas Borealis. While the parameters used in this paper seem well suited to estimate event back azimuths, future marsquakes may need adjustments based on event characteristics. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | We showed that back azimuth determination can be improved by a combination of eigenvector analysis on both P and S-arrivals. This study increases the amount of information on source regions that is possible with the InSight data set. The method allows for an automated determination and quality control of back azimuths from single stations that could also be applied to remote and sparse networks on Earth. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | We found that the majority of marsquakes located with this method are originating from the general Cerberus Fossae region. A few events seem to originate from different directions. The two most distant marsquakes are located in the Tharsis region on Mars, with one quake possibly in Valles Marineris. The new locations will enrich tectonic interpretation of the single-plate planet Mars. InSight continues to record the Martian seismicity and expand the Martian seismic catalog. Each additional event with a constrained location will help build a more comprehensive picture of Mars’ tectonic activity and its interior. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | CNRS, ETHZ, ICL, MPS-MPG), and the operators of JPL, SISMOC, MSDS, IRIS-DMC and PDS for providing SEED SEIS data. This is InSight contribution 221. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | The InSight seismic event catalog Version 9 (InSight Marsquake Service, 2022), waveform data and station metadata are available from IRIS-DMC and the IPGP datacenter (InSight Mars SEIS Data Service, 2019). Earthquake data were obtained from seismic network II (Scripps Institution Of Oceanography, 1986) and accessed with ObsPyDMT (Hosseini and Sigloch, 2017) using the IRIS catalog. Synthetic seismograms were calculated using the Instaseis (van Driel et al., 2015) Martian model InSight KKS21GP (St¨ahler et al., 2021) which can be accessed at http://instaseis.ethz.ch/marssynthetics/InSight_KKS21_GP/. Seismic data was handled with ObsPy (Beyreuther et al., 2010; Krischer et al., 2015). Calculations in Python were done with Numpy (Harris et al., 2020) and Scipy (Virtanen et al., 2020) and the results were visualised with Matplotlib (Hunter, 2007) and seaborn (Waskom, 2021). The Supplementary Material contains a detailed polarization analysis for each marsquake in Table 1 not presented in the main text. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Anderson, P. N., Duennebier, F. K., and Cessaro, R. K. (1987). Ocean borehole horizontal seismic sensor orientation determined from explosive charges. Journal of Geophysical Research: Solid Earth, 92(B5):3573– 3579. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Banerdt, W. B., Smrekar, S. E., Banfield, D., Giardini, D., Golombek, M., Johnson, C. L., Lognonn´e, P., Spiga, A., Spohn, T., Perrin, C., St¨ahler, S. C., Antonangeli, D., Asmar, S., Beghein, C., Bowles, N., Bozdag, E., Chi, P., Christensen, U., Clinton, J., Collins, G. S., Daubar, I., Dehant, V., Drilleau, M., Fillingim, M., Folkner, W., Garcia, R. F., Garvin, J., Grant, J., Grott, M., Grygorczuk, J., Hudson, T., Irving, J. C. E., Kargl, G., Kawamura, T., Kedar, S., King, S., Knapmeyer-Endrun, B., Knapmeyer, M., Lemmon, M., Lorenz, R., Maki, J. N., Margerin, L., McLennan, S. M., Michaut, C., Mimoun, D., Mittelholz, A., Mocquet, A., Morgan, P., Mueller, N. T., Murdoch, N., Nagihara, S., Newman, C., Nimmo, F., Panning, M., Pike, W. T., Plesa, A.-C., Rodriguez, S., Rodriguez-Manfredi, J. A., Russell, C. T., Schmerr, N., Siegler, M., Stanley, S., Stutzmann, E., Teanby, N., Tromp, J., van Driel, M., Warner, N., Weber, R., and Wieczorek, M. (2020). Initial results from the InSight mission on Mars. Nature Geoscience, 13(3):183–189. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Brinkman, N., St¨ahler, S. C., Giardini, D., Schmelzbach, C., Khan, A., Jacob, A., Fuji, N., Perrin, C., Lognonn´e, P., Beucler, E., B¨ose, M., Ceylan, S., Charalambous, C., Clinton, J. F., van Driel, M., Euchner, F., Horleston, A., Kawamura, T., Knapmeyer-Endrun, B., Mainsant, G., Panning, M. P., Pike, W. T., Scholz, J.-R., Robertsson, J. O. A., and Banerdt, W. B. (2021). First Focal Mechanisms of Marsquakes. Journal of Geophysical Research: Planets, 126(4):e2020JE006546. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | B¨ose, M., Clinton, J. F., Ceylan, S., Euchner, F., van Driel, M., Khan, A., Giardini, D., Lognonn´e, P., and Banerdt, W. B. (2017). A probabilistic framework for single-station location of seismicity on Earth and Mars. Physics of the Earth and Planetary Interiors, 262:48–65. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Ceylan, S., Clinton, J. F., Giardini, D., St¨ahler, S. C., B¨ose, M., Charalambous, C., Dahmen, N. L., van Driel, M., Dur´an, C., Horleston, A., Kawamura, T., Plasman, M., Zenh¨ausern, G., Euchner, F., Banerdt, W. B., Lognonn´e, P., Banfield, D., Beucler, E., Garcia, R. F., Kedar, S., Panning, M. P., Perrin, C., Pike, W. T., Smrekar, S. E., Spiga, A., and Stott, A. E. (2022). The marsquake catalogue from InSight, sols 0–1011. Manuscript in preparation. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Clinton, J., Ceylan, S., van Driel, M., Giardini, D., St¨ahler, S. C., B¨ose, M., Charalambous, C., Dahmen, N. L., Horleston, A., Kawamura, T., Khan, A., Orhand-Mainsant, G., Scholz, J.-R., Euchner, F., Banerdt, W. B., Lognonne, P., Banfield, D., Beucler, E., Kedar, S., Panning, M., Perrin, C., Pike, W. T., Smrekar, S. E., Spiga, A., and Stott, A. E. (2021). The Marsquake Catalogue from InSight, Sols 0-478. Physics of the Earth and Planetary Interiors, 310:106595. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Clinton, J., Giardini, D., B¨ose, M., Ceylan, S., van Driel, M., Euchner, F., Garcia, R. F., Kedar, S., Khan, A., St¨ahler, S. C., Banerdt, B., Lognonne, P., Beucler, E., Daubar, I., Drilleau, M., Golombek, M., Kawamura, T., Knapmeyer, M., Knapmeyer-Endrun, B., Mimoun, D., Mocquet, A., Panning, M., Perrin, C., and Teanby, N. A. (2018). The marsquake service: Securing daily analysis of seis data and building the martian seismicity catalogue for insight. Space Science Reviews, 214(8):133. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Dahmen, N. L., Zenh¨ausern, G., Clinton, J. F., Giardini, D., St¨ahler, S. C., Ceylan, S., Charalambous, C., van Driel, M., Hurst, K. J., Kedar, S., Lognonn´e, P., Murdoch, N., Myhill, R., Panning, M. P., Pike, W. T., Schimmel, M., Schmelzbach, C., Scholz, J.-R., Stott, A. E., Stutzmann, E., and Banerdt, W. B. (2021). Resonances and Lander Modes observed by InSight on Mars (1-9 Hz). Bulletin of the Seismological Society of America. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Drilleau, M., Samuel, H., Garcia, R. F., Rivoldini, A., Perrin, C., Michaut, C., Wieczorek, M., Tauzin, B., Connolly, J. A. D., Meyer, P., Lognonn´e, P., and B., B. W. (2022). Marsquake locations and 1-D seismic models for Mars from InSight data. Earth and Space Science Open Archive, page 77. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Fontaine, F. R., Barruol, G., Kennett, B. L., Bokelmann, G. H. R., and Reymond, D. (2009). Upper mantle anisotropy beneath Australia and Tahiti from P wave polarization: Implications for real-time earthquake location. Journal of Geophysical Research, 114(B3):B03306–B03306. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Giardini, D., Lognonn´e, P., Banerdt, W. B., Pike, W. T., Christensen, U., Ceylan, S., Clinton, J. F., van Driel, M., St¨ahler, S. C., B¨ose, M., Garcia, R. F., Khan, A., Panning, M., Perrin, C., Banfield, D., Beucler, E., Charalambous, C., Euchner, F., Horleston, A., Jacob, A., Kawamura, T., Kedar, S., Mainsant, G., Scholz, J.-R., Smrekar, S. E., Spiga, A., Agard, C., Antonangeli, D., Barkaoui, S., Barrett, E., Combes, P., Conejero, V., Daubar, I., Drilleau, M., Ferrier, C., Gabsi, T., Gudkova, T., Hurst, K., Karakostas, F., King, S., Knapmeyer, M., Knapmeyer-Endrun, B., Llorca-Cejudo, R., Lucas, A., Luno, L., Margerin, L., McClean, J. B., Mimoun, D., Murdoch, N., Nimmo, F., Nonon, M., Pardo, C., Rivoldini, A., Manfredi, J. A. R., Samuel, H., Schimmel, M., Stott, A. E., Stutzmann, E., Teanby, N., Warren, T., Weber, R. C., Wieczorek, M., and Yana, C. (2020). The seismicity of Mars. Nature Geoscience, 13(3):205–212. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Golombek, M., Warner, N. H., Grant, J. A., Hauber, E., Ansan, V., Weitz, C. M., Williams, N., Charalam- bous, C., Wilson, S. A., DeMott, A., Kopp, M., Lethcoe-Wilson, H., Berger, L., Hausmann, R., Marteau, E., Vrettos, C., Trussell, A., Folkner, W., Le Maistre, S., Mueller, N., Grott, M., Spohn, T., Piqueux, S., Millour, E., Forget, F., Daubar, I., Murdoch, N., Lognonn´e, P., Perrin, C., Rodriguez, S., Pike, W. T., Parker, T., Maki, J., Abarca, H., Deen, R., Hall, J., Andres, P., Ruoff, N., Calef, F., Smrekar, S., Baker, M. M., Banks, M., Spiga, A., Banfield, D., Garvin, J., Newman, C. E., and Banerdt, W. B. (2020). Geology of the InSight landing site on Mars. Nature Communications, 11(1):1–11. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Harris, C. R., Millman, K. J., van der Walt, S. J., Gommers, R., Virtanen, P., Cournapeau, D., Wieser, E., Taylor, J., Berg, S., Smith, N. J., Kern, R., Picus, M., Hoyer, S., van Kerkwijk, M. H., Brett, M., Haldane, A., del R´ıo, J. F., Wiebe, M., Peterson, P., G´erard-Marchant, P., Sheppard, K., Reddy, T., Weckesser, W., Abbasi, H., Gohlke, C., and Oliphant, T. E. (2020). Array programming with NumPy. Nature, 585(7825):357–362. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Hobiger, M., Halo, M., Schmelzbach, C., St¨ahler, S., F¨ah, D., Giardini, D., Golombek, M., Clinton, J., Dahmen, N., Zenh¨ausern, G., Knapmeyer-Endrun, B., Carrasco, S., Constantinos, C., Hurst, K., Kedar, S., and Banerdt, B. (2021). The shallow structure of Mars from inversion of high-frequency ambient noise Rayleigh wave ellipticity at the InSight landing site. Nature Geoscience. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Khan, A., Ceylan, S., van Driel, M., Giardini, D., Lognonn´e, P., Samuel, H., Schmerr, N. C., St¨ahler, S. C., Duran, A. C., Huang, Q., et al. (2021). Upper mantle structure of mars from insight seismic data. Science, 373(6553):434–438. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Krischer, L., Megies, T., Barsch, R., Beyreuther, M., Lecocq, T., Caudron, C., and Wassermann, J. (2015). ObsPy: a bridge for seismology into the scientific python ecosystem. Computational Science & Discovery, 8(1):014003. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Lammlein, D. R. (1977). Lunar Seismicity, Structure, and Tectonics. Philosophical Transactions of the Royal Society of London. Series A, Mathematical and Physical Sciences, 285(1327):451–461. Publisher: The Royal Society. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Lognonn´e, P., Banerdt, W. B., Giardini, D., Pike, W. T., Christensen, U., Laudet, P., de Raucourt, S., Zweifel, P., Calcutt, S., Bierwirth, M., Hurst, K. J., Ijpelaan, F., Umland, J. W., Llorca-Cejudo, R., Larson, S. A., Garcia, R. F., Kedar, S., Knapmeyer-Endrun, B., Mimoun, D., Mocquet, A., Panning, M. P., Weber, R. C., Sylvestre-Baron, A., Pont, G., Verdier, N., Kerjean, L., Facto, L. J., Gharakanian, V., Feldman, J. E., Hoffman, T. L., Klein, D. B., Klein, K., Onufer, N. P., Paredes-Garcia, J., Petkov, M. P., Willis, J. R., Smrekar, S. E., Drilleau, M., Gabsi, T., Nebut, T., Robert, O., Tillier, S., Moreau, C., Parise, M., Aveni, G., Ben Charef, S., Bennour, Y., Camus, T., Dandonneau, P. A., Desfoux, C., Lecomte, B., Pot, O., Revuz, P., Mance, D., tenPierick, J., Bowles, N. E., Charalambous, C., Delahunty, A. K., Hurley, J., Irshad, R., Liu, H., Mukherjee, A. G., Standley, I. M., Stott, A. E., Temple, J., Warren, T., Eberhardt, M., Kramer, A., K¨uhne, W., Miettinen, E.-P., Monecke, M., Aicardi, C., Andr´e, M., Baroukh, J., Borrien, A., Bouisset, A., Boutte, P., Brethom´e, K., Brysbaert, C., Carlier, T., Deleuze, M., Desmarres, J. M., Dilhan, D., Doucet, C., Faye, D., Faye-Refalo, N., Gonzalez, R., Imbert, C., Larigauderie, C., Locatelli, E., Luno, L., Meyer, J.-R., Mialhe, F., Mouret, J. M., Nonon, M., Pahn, Y., Paillet, A., Pasquier, P., Perez, G., Perez, R., Perrin, L., Pouilloux, B., Rosak, A., Savin de Larclause, I., Sicre, J., Sodki, M., Toulemont, N., Vella, B., Yana, C., Alibay, F., Avalos, O. M., Balzer, M. A., Bhandari, P., Blanco, E., Bone, B. D., Bousman, J. C., Bruneau, P., Calef, F. J., Calvet, R. J., D’Agostino, S. A., de los Santos, G., Deen, R. G., Denise, R. W., Ervin, J., Ferraro, N. W., Gengl, H. E., Grinblat, F., Hernandez, D., Hetzel, M., Johnson, M. E., Khachikyan, L., Lin, J. Y., Madzunkov, S. M., Marshall, S. L., Mikellides, I. G., Miller, E. A., Raff, W., Singer, J. E., Sunday, C. M., Villalvazo, J. F., Wallace, M. C., Banfield, D., Rodriguez-Manfredi, J. A., Russell, C. T., Trebi-Ollennu, A., Maki, J. N., Beucler, E., B¨ose, M., Bonjour, C., Berenguer, J. L., Ceylan, S., Clinton, J., Conejero, V., Daubar, I., Dehant, V., Delage, P., Euchner, F., Est`eve, I., Fayon, L., Ferraioli, L., Johnson, C. L., Gagnepain-Beyneix, J., Golombek, M., Khan, A., Kawamura, T., Kenda, B., Labrot, P., Murdoch, N., Pardo, C., Perrin, C., Pou, L., Sauron, A., Savoie, D., St¨ahler, S., Stutzmann, E., Teanby, N. A., Tromp, J., van Driel, M., Wieczorek, M., Widmer-Schnidrig, R., and Wookey, J. (2019). SEIS: Insight’s seismic experiment for internal structure of Mars. Space Science Reviews, 215(1):12. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Lognonn´e, P., Banerdt, W. B., Pike, W. T., Giardini, D., Christensen, U., Garcia, R. F., Kawamura, T., Kedar, S., Knapmeyer-Endrun, B., Margerin, L., Nimmo, F., Panning, M., Tauzin, B., Scholz, J.-R., Antonangeli, D., Barkaoui, S., Beucler, E., Bissig, F., Brinkman, N., Calvet, M., Ceylan, S., Charalambous, C., Davis, P., van Driel, M., Drilleau, M., Fayon, L., Joshi, R., Kenda, B., Khan, A., Knapmeyer, M., Lekic, V., McClean, J., Mimoun, D., Murdoch, N., Pan, L., Perrin, C., Pinot, B., Pou, L., Menina, S., Rodriguez, S., Schmelzbach, C., Schmerr, N., Sollberger, D., Spiga, A., St¨ahler, S., Stott, A., Stutzmann, E., Tharimena, S., Widmer-Schnidrig, R., Andersson, F., Ansan, V., Beghein, C., B¨ose, M., Bozdag, E., Clinton, J., Daubar, I., Delage, P., Fuji, N., Golombek, M., Grott, M., Horleston, A., Hurst, K., Irving, J., Jacob, A., Knollenberg, J., Krasner, S., Krause, C., Lorenz, R., Michaut, C., Myhill, R., Nissen-Meyer, T., ten Pierick, J., Plesa, A.-C., Quantin-Nataf, C., Robertsson, J., Rochas, L., Schimmel, M., Smrekar, S., Spohn, T., Teanby, N., Tromp, J., Vallade, J., Verdier, N., Vrettos, C., Weber, R., Banfield, D., Barrett, E., Bierwirth, M., Calcutt, S., Compaire, N., Johnson, C., Mance, D., Euchner, F., Kerjean, L., Mainsant, G., Mocquet, A., Rodriguez Manfredi, J. A., Pont, G., Laudet, P., Nebut, T., de Raucourt, S., Robert, O., Russell, C. T., Sylvestre-Baron, A., Tillier, S., Warren, T., Wieczorek, M., Yana, C., and Zweifel, P. (2020). Constraints on the shallow elastic and anelastic structure of Mars from InSight seismic data. Nature Geoscience, 13(3):213–220. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Menina, S., Margerin, L., Kawamura, T., Lognonn´e, P., Marti, J., Drilleau, M., Calvet, M., Compaire, N., Garcia, R., Karakostas, F., Schmerr, N., van Driel, M., St¨ahler, S. C., Plasman, M., Giardini, D., Carrasco, S., Knapmeyer-Endrun, B., Sainton, G., and Banerdt, W. B. (2021). Energy Envelope and Attenuation Characteristics of High-Frequency (HF) and Very-High-Frequency (VF) Martian Events. Bulletin of the Seismological Society of America. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Savoie, D., Richard, A., Goutaudier, M., Lognonn´e, P., Hurst, K. J., Maki, J. N., Golombek, M. P., van Driel, M., Clinton, J., Stutzmann, E., Mimoun, D., Banerdt, W. B., and Williams, N. R. (2021). Find- ing SEIS North on Mars: Comparisons Between SEIS Sundial, Inertial and Imaging Measurements and Consequences for Seismic Analysis. Earth and Space Science, 8(3):e2020EA001286. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Scholz, J.-R., Barruol, G., Fontaine, F. R., Sigloch, K., Crawford, W. C., and Deen, M. (2016). Orienting ocean-bottom seismometers from P-wave and Rayleigh wave polarizations. Geophysical Journal Interna- tional, 208(3):1277–1289. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Scholz, J.-R., Widmer-Schnidrig, R., Davis, P., Lognonn´e, P., Pinot, B., Garcia, R. F., Hurst, K., Pou, L., Nimmo, F., Barkaoui, S., et al. (2020). Detection, analysis, and removal of glitches from InSight’s seismic data from Mars. Earth and Space Science, 7(11):e2020EA001317. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Smith, D. E., Zuber, M. T., Frey, H. V., Garvin, J. B., Head, J. W., Muhleman, D. O., Pettengill, G. H., Phillips, R. J., Solomon, S. C., Zwally, H. J., and others (2001). Mars Orbiter Laser Altimeter: Experiment summary after the first year of global mapping of Mars. Journal of Geophysical Research: Planets, 106(E10):23689–23722. Publisher: Wiley Online Library. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Sollberger, D., Igel, H., Schmelzbach, C., Edme, P., van Manen, D.-J., Bernauer, F., Yuan, S., Wassermann, J., Schreiber, U., and Robertsson, J. O. A. (2020). Seismological Processing of Six Degree-of-Freedom Ground-Motion Data. Sensors, 20(23). |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Stachnik, J. C., Sheehan, A. F., Zietlow, D. W., Yang, Z., Collins, J., and Ferris, A. (2012). Determination of New Zealand Ocean Bottom Seismometer Orientation via Rayleigh-Wave Polarization. Seismological Research Letters, 83(4):704–713. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | St¨ahler, S. C., Khan, A., Banerdt, W. B., Lognonn´e, P., Giardini, D., Ceylan, S., Drilleau, M., Duran, A. C., Garcia, R. F., Huang, Q., et al. (2021). Seismic detection of the martian core. Science, 373(6553):443–448. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Stutzmann, E., Schimmel, M., Lognonn´e, P., Horleston, A., Ceylan, S., van Driel, M., Stahler, S., Banerdt, B., Calvet, M., Charalambous, C., et al. (2021). The polarization of ambient noise on Mars. Journal of Geophysical Research: Planets, 126(1):e2020JE006545. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | St¨ahler, S., Khan, A., Banerdt, W. B., Lognonn´e, P., Giardini, D., Savas Ceylan, Drilleau, M., A. Cecilia Duran, Garcia, R. F., Quancheng Huang, Doyeon Kim, Lekic, V., Samuel, H., Schimmel, M., Schmerr, N., Sollberger, D., El´eonore Stutzmann, Zongbo Xu, Antonangeli, D., Constantinos Charalambous, Davis, P., Irving, J. C., Kawamura, T., Knapmeyer, M., Maguire, R., Marusiak, A. G., Panning, M. P., Perrin, C., Ana-Catalina Plesa, Rivoldini, A., Schmelzbach, C., Zenh¨ausern, G., ´Eric Beucler, Clinton, J., Dahmen, N., Van Driel, M., Gudkova, T., Horleston, A., W. Thomas Pike, Plasman, M., and Smrekar, S. E. (2021). Interior Models of Mars from inversion of seismic body waves. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Survey, U. G. (2011). M 5.9 - 186 km NW of Tobelo, Indonesia. Accessed: 2022-01-20. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | van Driel, M., Ceylan, S., Clinton, J. F., Giardini, D., Horleston, A., Margerin, L., St¨ahler, S. C., B¨ose, M., Charalambous, C., Kawamura, T., Khan, A., Orhand-Mainsant, G., Scholz, J.-R., Euchner, F., Knap- meyer, M., Schmerr, N., Pike, W. T., Lognonn´e, P., and Banerdt, W. B. (2021). High-frequency seismic events on mars observed by insight. Journal of Geophysical Research: Planets, 126(2):e2020JE006670. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Virtanen, P., Gommers, R., Oliphant, T. E., Haberland, M., Reddy, T., Cournapeau, D., Burovski, E., Peterson, P., Weckesser, W., Bright, J., van der Walt, S. J., Brett, M., Wilson, J., Millman, K. J., Mayorov, N., Nelson, A. R. J., Jones, E., Kern, R., Larson, E., Carey, C. J., Polat, ˙I., Feng, Y., Moore, E. W., VanderPlas, J., Laxalde, D., Perktold, J., Cimrman, R., Henriksen, I., Quintero, E. A., Harris, C. R., Archibald, A. M., Ribeiro, A. H., Pedregosa, F., van Mulbregt, P., and SciPy 1.0 Contributors (2020). SciPy 1.0: Fundamental Algorithms for Scientific Computing in Python. Nature Methods, 17:261–272. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Zweifel, P., Mance, D., ten Pierick, J., Giardini, D., Schmelzbach, C., Haag, T., Nicollier, T., Ceylan, S., St¨ahler, S., van Driel, M., Sollberger, D., Euchner, F., Clinton, J. F., Bierwirth, M., Eberhardt, M., Lognonn´e, P., Pike, W. T., and Banerdt, W. B. (2021). Seismic High-Resolution Acquisition Electronics for the NASA InSight Mission on Mars. Bulletin of the Seismological Society of America. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Figure 10: Overview of results as seen in Table 1. Probability density functions for this study (purple), Drilleau et al. (2022) (blue), and MQS (green). For this study, we show the KDE for events without marker. Events marked with † have the back azimuth and uncertainty set manually and use a Gaussian curve - for a detailed discussion of each event, see Supplementary Material. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Figure 11: Robustness analysis for different types of quakes. (a) True error (difference between estimated back azimuth and true back azimuth) against SNR of the event for synthetic marsquakes. (b) True error against SNR for earthquakes recorded at station WRAB. (c) Estimated error from the width of the KDE peak against SNR for marsquakes. The colors of the dots in the first two rows indicate the estimated error. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Figure S1: Polarization analysis for marsquake S0105a (Mw3.0 LF QB event at 32.5◦ distance, March 14, 2019). The plot follows the same structure as Fig. 6. The frequency band of 0.3–0.65 Hz shows moderate to high inclination values for the P window, as well as low inclination values for the S window. At frequencies below 0.3 Hz, P is dominated by a glitch 30 seconds after the picked arrival. The S-arrival has a strong polarization at 0°and 190°, which is consistent with the P-polarization, but suggests that the value could be 10-20°too high. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Figure S2: Polarization analysis for marsquake S0154a (Mw2.9 BB QC event at 33.7◦ distance, May 4, 2019). The plot follows the same structure as Fig. 6. The two arrivals are marked as x1 and x2 by MQS since they are not unequivocally P and S using standard tools. In our analysis, we find that they can be identified well as P and S in a relatively narrow frequency band from 0.4–0.7 Hz. x2 has a high-inclination signal at 0.3–0.4 Hz, which seems inconsistent with S, but could be explained by the low SNR. Its horizontal signal is consistent with a back azimuth from the east (d). A weaker x1 and some contamination of x2 make this event challenging, though results are consistent for both phases. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Figure S3: Polarization analysis for marsquake S0167b (Mw unknown LF QC event at unknown distance, May 17, 2019). The plot follows the same structure as Fig. 6. This event is at a larger distance than most others discussed in this article. The identification of x1 as P is difficult based on polarization only, since the direction does not differ significantly from the pre-event noise window, which is contaminated by wind. x2 shows a low-inclination signal at azimuth 40 or 230°, distinct from the wind-induced noise before. Its signal is also significantly stronger than the noise. We assign a manual back azimuth of 315°, consistent with the polarization of 300-340°from x2. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Figure S4: Polarization analysis for marsquake S0173a (Mw3.7 LF QA event at 30◦ distance, May 23, 2019). The plot follows the same structure as Fig. 6. One of the best events recorded by InSight during the mission so far. The P is contaminated by a glitch arriving 20 seconds after, visible by a strongly horizontal signal (blue in the azimuth plot). Using frequencies between 0.3–1.0 Hz avoids the glitch. Both P and S are strongly polarized and show consistent results of 90°. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Figure S5: Polarization analysis for marsquake S0183a (Mw3.1 LF QB event at 46.4◦ distance, June 3, 2019). The plot follows the same structure as Fig. 6. The P-wave has a consistent polarization between 0.2–1.0 Hz, with a strongly vertical signal. The S-wave has a horizontal inclination in the same frequency band, though it is possibly glitch-contaminated (see (a) and Clinton et al. (2021) for a detailed discussion). Results from P and the presumptive S are consistent with each other. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Figure S6: Polarization analysis for marsquake S0205a (Mw3.0 BB QC event at 42.4◦ distance, June 25, 2019). The plot follows the same structure as Fig. 6. The event lacks a second pick, so only x1 is available. x1 shows a high inclination between 0.3–1.0 Hz, and a southward polarization. Since no second pick is available, the distance is obtained from alignment (InSight Marsquake Service, 2022). |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Figure S7: Polarization analysis for marsquake S0290b (Mw3.5 LF QB event at 30.4◦ distance, September 21, 2019). The plot follows the same structure as Fig. 6. There is a strong glitch shortly before the P with horizontal inclination, but the P window is not affected (cf. inclination row). The S window contains a clear horizontal signal between 0.4 and 0.7 Hz, clearly different from the largely vertical energy present otherwise. The frequency band 0.3–0.8 Hz catches this horizontal energy while containing the stable azimuth signal in the P window. There is general agreement between the P and S results as seen in (d), resulting in an estimated back azimuth of 257-341 °. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Figure S8: Polarization analysis for marsquake S0325a (Mw3.7 LF QB event at 39.7◦ distance, October 26, 2019). The plot follows the same structure as Fig. 6. The frequency band of 0.15–0.5 Hz contains both the high inclination part of the P window and the low inclination part of the S window. As seen in (c), the P-wave window has a polarization that is significantly different from the noise window below 0.3 Hz. While the noise above 0.3 Hz has a similar azimuth as the P window, agreement with the S window and higher inclination suggest that the P-wave polarization is robust and shows the true event back azimuth of 43-73°. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Figure S9: Polarization analysis for marsquake S0407a (Mw2.9 LF QB event at 29.3◦ distance, January 19, 2020). The plot follows the same structure as Fig. 6. There is clear vertical and horizontal energy visible at the P and S arrival, respectively. The P-wave has a narrow-peaked azimuth distribution above 0.6 Hz, but is more broad below. The S-wave polarization shows the opposite picture: Peaks at 0 and 180°below 0.6 Hz (consistent with an event BAZ of 90°), but 30°and 210°above 0.6 Hz (consistent with BAZ of 100-120°. In summary, the event is likely located due east, but with a relatively high uncertainty, if one takes all frequencies into account. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Figure S10: Polarization analysis for marsquake S0409d (Mw3.2 LF QB event at 31.1◦ distance, January 21, 2020). The plot follows the same structure as Fig. 6. The P arrival is dominated by several large glitches, and the event signal is difficult to retrieve. The S arrival shows clear horizontal energy and it suggests a back azimuth around 80◦. Analysing the P window more closely with this knowledge shows vertical energy in a similar direction. The back azimuth is manually set to 70◦, and (c) and (d) show that there is good agreement with the S window for this. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Figure S11: Polarization analysis for marsquake S0474a (Mw2.9 LF QC event at 29.1◦ distance, March 28, 2020). The plot follows the same structure as Fig. 6. The P-wave pick is speculative with ±60 s uncertainty. The amplitude KDE shows that the P window has no strong amplitude, resulting in a very weak SNR based on the P. However, the S window shows both a stronger amplitude and clear horizontal energy, suggesting a back azimuth of around 90◦. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Figure S12: Polarization analysis for marsquake S0484b (Mw2.9 BB QB event at 31.8◦ distance, April 7, 2020). The plot follows the same structure as Fig. 6. The P-arrival is contaminated by a strong glitch (visible in (a) and (b)). There is some vertical energy visible in the P window at higher frequencies. The S window has a clear horizontal arrival with strong amplitude. It suggests a back azimuth around 110◦. The back azimuth is set to 100◦ after careful analysis of the P window. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Figure S13: Polarization analysis for marsquake S0784a (Mw3.3 BB QB event at 34.5◦ distance, February 9, 2021). The plot follows the same structure as Fig. 6. Both P and S arrival show clear vertical and horizontal energy respectively. The S window shows the horizontal energy as more dominant at lower frequencies, while the P window shows dominantly vertical energy over the whole frequency band. Results from both windows are consistent around 115◦. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Figure S15: Polarization analysis for marsquake S0809a (Mw3.3 LF QA event at 29.8◦ distance, March 7, 2021). The plot follows the same structure as Fig. 6. Both P and S arrivals are clear with no glitch contamination. The P-wave polarization is similar to the pre-event noise, but the S-wave polarization is distinct. They both show a consistent polarization which agree with each other. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Figure S16: Polarization analysis for marsquake S0820a (Mw3.3 LF QA event at 30.2◦ distance, March 18, 2021). The plot follows the same structure as Fig. 6. Both P and S arrivals are clear with no glitch contamination. The P-wave polarization is similar to the pre-event noise, but the S-wave polarization is distinct. They both show a consistent polarization which agree with each other. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Figure S17: Polarization analysis for marsquake S0850c (Mw2.6 BB QC event at 31.6◦ distance, April 18, 2021). The plot follows the same structure as Fig. 6. There is a strong glitch after the S-wave arrival, though its polarization appears clearly separated from the event signal. The P window shows two dominant azimuths, either at 310°below 0.6 Hz, or at 60°above 0.6 Hz. The S window excludes the 60°direction and instead confirms the range between 277 and 325°. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Figure S18: Polarization analysis for marsquake S0864a (Mw3.1 BB QA event at 28.7◦ distance, May 2, 2021). The plot follows the same structure as Fig. 6. Both P and S arrivals are clear with no glitch contamination. Their amplitude is significantly above the noise and they both show a consistent polarization. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Figure S19: Polarization analysis for marsquake S0899d (Mw unknown LF QB event at unknown distance, May 2, 2021). The plot follows the same structure as Fig. 6. Since there is no second phase pick, the event has no assigned distance. x1 shows a strongly polarized signal, with a vertical inclination and a back azimuth at 22◦. Though it cannot be corroborated with a second pick, the strong signal amplitude suggests this is the event polarization and not simply noise. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Figure S20: Polarization analysis for marsquake S0916d (Mw2.9 BB QB event at 29.3◦ distance, June 25, 2021). The plot follows the same structure as Fig. 6. The P arrival is contaminated by a strong glitch. Its effect can be minimized by taking a frequency band between 0.4-1.5 Hz. The P-wave azimuth is similar in distribution to the pre-event noies. The S window however shows a clear and strong polarization, with dominant horizontal energy and clear north-south azimuth. Results from both P and S windows agree with each other. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Figure S21: Polarization analysis for marsquake S0918a (Mw3.0 LF QB event at 27.9◦ distance, June 27, 2021). The plot follows the same structure as Fig. 6. Both P and S signals show amplitudes above the noise. Stable P and S polarization is seen between 0.4–0.8 Hz, where P has high inclination and S has low inclination. The P-wave polarization is bimodal, with peaks at 50°and 135°, an ambiguity that is not entirely resolved by the S-wave. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Figure S22: Polarization analysis for marsquake S0976a (Mw4.2 LF QA event at 146◦ distance, August 25, 2021). The plot follows the same structure as Fig. 6. This is the most distant event recorded by InSight so far. The P window polarization shows a consistent back azimuth around 94◦. The S window suggests a back azimuth in opposite direction around 270◦. |
Processed_Low_Frequency_Marsquakes_and_Where_to_Find_Them:_B.txt | Figure S23: Polarization analysis for marsquake S1000a (Mw4.1 BB QB event at 116◦ distance, September 18, 2021). The plot follows the same structure as Fig. 6. The P window shows a dominant azimuth over a wide frequency band, in northeast direction. Taking a frequency band of 0.1–0.35 Hz puts the back azimuth at 57◦. In this band, the S window is the most horizontal The S window suggests a more north-northwest direction. However, some of that is due to glitch contamination. |
Processed_Are_There_Radical_Cyanogen_Abundance_Differences_B.txt | Abstract. On reading an old paper about galactic globular cluster abundance observations (of NGC 6752) we came across an intriguing result. Norris et al. (1981) found that there was a dis- tinct lack of cyanogen-strong (CN-strong) stars in their sample of AGB stars, as compared to their sample of RGB stars (which had roughly equal numbers of CN-normal and CN- strong stars). Further reading revealed that similar features have been discovered in the AGB populations of other clusters. Recently, Sneden et al. (2000) followed up on this pos- sibility (and considered other proton-capture products) by compiling the existing data at the time and came to a similar conclusion for two more clusters. Unfortunately all of these studies suffer from low AGB star counts so the conclusions are not necessarily robust — larger, statistically significant, sample sizes are needed. In this conference paper, presented at the Eighth Torino Workshop on Nucleosynthesis in AGB Stars (Universidad de Granada, Spain, 2006), we outline the results of a literature search for relevant CN observations and describe our observing proposal to test the sugges- tion that there are substantial abundance differences between the AGB and RGB in galactic globular clusters. The literature search revealed that the AGB star counts for all studies (which are not, in general, studies about AGB stars in particular) are low, usually being ≤ 10. The search also revealed that the picture may not be consistent between clusters. Although most clusters appear to have CN-weak AGBs, at least two seem to have CN- strong AGBs (M5 & 47 Tuc). To further complicate the picture, clusters often appear to have a combination of both CN-strong and CN-weak stars on their AGBs – although one population tends to dominate. Again, all these assertions are however based on small sam- ple sizes. We aim to increase the sample sizes by an order of magnitude using existing high quality photometry in which the AGB and RGB can be reliably separated. For the observa- tions we will use a wide-field, low- to mid-resolution multi-object spectroscope to obtain data not only on the AGB but also on the horizontal branches and first giant branches of a sample of clusters. With the new information we hope to ascertain whether significant abundance differences really exist. |
Processed_Are_There_Radical_Cyanogen_Abundance_Differences_B.txt | We are attempting to perform a conclusive test of the suggestion put forward by Norris et al. (1981), which has been touched upon by many authors since and recently explored by Sneden et al. (2000), that there are differences in cyanogen abundance distributions between the first and second giant branches in galactic globular clusters. |
Processed_Are_There_Radical_Cyanogen_Abundance_Differences_B.txt | Although galactic globular clusters (GCs) are chemically homogeneous with respect to Fe and most other heavy elements (see eg. Kraft et al. (1992)), it has long been known that they show inhomogeneities in many lighter el- ements (eg. C, N, O, Mg, Al). These inhomo- geneities are considered anomalous because they are not seen in halo field stars of similar metallicity (see eg. Gratton et al. (2000)). |
Processed_Are_There_Radical_Cyanogen_Abundance_Differences_B.txt | Ivans et al. (1999)). However these studies have hinted at a tantalising characteristic: most (observed) GCs show a lack of CN-strong stars on the AGB. If this is true then it is in stark contrast to the red giant branch (RGB) and ear- lier phases of evolution, where the ratio of CN- Strong to CN-Weak stars is roughly unity in many clusters. |
Processed_Are_There_Radical_Cyanogen_Abundance_Differences_B.txt | This possible discrepancy was noted by Norris et al. (1981) in their paper about abun- dances in giant stars in NGC 6752. They state that “The behaviour of the CN bands in the AGB stars is... quite difficult to understand... not one of the stars studied here has enhanced CN... yet on the [first] giant branch there are more CN strong stars than CN weak ones.” (also see Figure 3 in that paper). More re- cently Sneden et al. (2000) presented a con- ference paper on this exact topic. Compiling the contemporaneous preexisting data in the literature they discussed the relative amounts of CN in AGB and RGB stars in the GCs NGC 6752 (data from Norris et al. (1981), M13 (data from Suntzeff (1981)) and M4 (data from Norris et al. (1981) and Suntzeff & Smith (1991)). They also discuss Na abundance vari- ations in M13 (data from Pilachowski et al. (1996a) and Pilachowski et al. (1996b) ). Their conclusion for the CN variations was that the clusters in question all showed significantly less CN on the AGB as compared to the RGB. However the data compiled only contained about 10 AGB stars per cluster. In their clos- ing remarks they suggest observations with larger sample sizes are needed — which may be done using wide-field multi-object spec- troscopes. This is exactly the conclusion the present authors also came to, inspiring this seminar/conference paper at the Eighth Torino Workshop on Nucleosynthesis AGB Stars held at the Universidad de Granada, Spain, in 2006. |
Processed_Are_There_Radical_Cyanogen_Abundance_Differences_B.txt | search was that the available number of AGB star observations are not statistically signifi- cant enough to come to any real conclusion about the nature of the CN abundance distribu- tions. This has mainly been due to technologi- cal constraints. However, the data that does ex- ist shows that there appears to be a strong trend towards CN-weak asymptotic giant branches. The picture is not so simple though, as two clusters in Table 1 actually have CN-strong AGBs. In addition to this, most clusters have some stars of the opposite class on their AGBs – the classifications given in Table 1 (usually) refer to strong majorities in each cluster, rather than totally homogeneous populations. |
Processed_Are_There_Radical_Cyanogen_Abundance_Differences_B.txt | A vital part in being able to observe signif- icant numbers of AGB stars is having photom- etry good enough to separate the AGB from the RGB. Photometric observations have now reached such high accuracy that it is becom- ing feasible to separate the AGB and RGB populations reliably. In addition to this, wide- field multi-object spectroscopes are now avail- able. During our literature search we came across some very high-quality photometric studies. For example, the study of M5 done by Sandquist & Bolte (2004). Their set of obser- vations is complete out to 8-10 arc min. They also tabulate all their stars according to evolu- tionary status – and find 105 AGB stars! This represents a sample size increase of one order of magnitude. Further to this we found colour magnitude diagrams for two more GC candi- dates that have the required accuracy (and high AGB star counts). Thus our current study in- volves three GCs, one of which appears to have a majority of CN-strong stars on its AGB (M5) which makes it an important outlier that may cause problems for some explanations of the (possible) phenomenon. |
Processed_Are_There_Radical_Cyanogen_Abundance_Differences_B.txt | stars as they are very well studied already – and have similar temperatures and luminosities to the AGB stars. Depending on the quality of the final data we will also attempt to derive abun- dances for aluminium and CH (a proxy for car- bon). |
Processed_Are_There_Radical_Cyanogen_Abundance_Differences_B.txt | Assuming for the purpose of discussion that the lack of CN-normal AGB stars is real, then this is actually the opposite to what we would expect based on observations at the tip of the RGB. These stars, which are the precursors to the AGB stars (via the HB), actually tend to become more N-dominated due to ‘extra mix- ing’ (the results of extra mixing are routinely observed in Halo field RGB stars as well GC RGB stars – see eg. Shetrone (2003)). Thus we would predict an increase in the number of CN-strong AGB stars over the RGB mean – rather than a decrease. |
Processed_Are_There_Radical_Cyanogen_Abundance_Differences_B.txt | Table 1. Results of the literature search for CN abundances in GC AGB stars. Note that ‘weak’ or ‘strong’ means that there is a very significant majority of that class of star in each case. |
Processed_The_Flare-Dominated_Accretion_Mode_of_a_Radio-Brig.txt | Transitional millisecond pulsars (tMSPs) are a new sub-class of neutron star low-mass X-ray binaries (NS- LMXBs) that becomes observationally known in the last decade (see Jaodand et al. 2018 for the time-line of some of the most significant events of the class). Unlike typi- cal accreting millisecond X-Ray pulsars (AMXPs), these systems switch between distinct states of being a pul- sar and an LMXB on time-scales that range from weeks to 10+ years (Bond et al. 2002; Thorstensen & Arm- strong 2005; Wang et al. 2009; Archibald et al. 2009; Patruno et al. 2014; Papitto et al. 2013). As the only- known bridge between the radio MSPs and LMXBs, they are widely linked to the standard recycling scenario of neutron stars (Alpar et al. 1982; Radhakrishnan & Srini- vasan 1982). |
Processed_The_Flare-Dominated_Accretion_Mode_of_a_Radio-Brig.txt | To date, only three tMSPs are known: PSR J1824- 2452I in M28 (a.k.a. M28I; Papitto et al. 2013), PSR J1023+0038 (Archibald et al. 2009; Patruno et al. 2014), and PSR J1227 4853 (Roy et al. 2015). They are all identified as “redback” eclipsing millisecond pulsar bina- ries, in which the MSP is ablating the low-mass compan- ion (median mass of 0.36 M(cid:12); Strader et al. 2019) in a compact orbit (orbital periods of (cid:47) 1 day). |
Processed_The_Flare-Dominated_Accretion_Mode_of_a_Radio-Brig.txt | M28I is currently the only known tMSP that showed a typical X-ray outburst as AMXPs (i.e., Lx (cid:38) 1036 erg s−1). In PSRs J1023+0038 and J1227 4853, the accretion state is about two orders of magnitude lower (Lx (cid:46) 1034 erg s−1). In this so-called sub-luminous disk state, at least three accretion modes, namely the low 1033 erg s−1), and (a few × 1034 erg s−1) are observed (de Martino flare modes ( et al. 2013; Bogdanov et al. 2015). At least one candi- date tMSP, 3FGL J1544.6 1125, has been identified via − its display of similar accretion modes and its other op- tical properties (Bogdanov & Halpern 2015; Britt et al. 2017). Like 3FGL J1544.6 1125, PSR J1023+0038 and − 4853 have been observed to emit GeV γ-rays PSR J1227 − in the sub-luminous disk states (Stappers et al. 2014; Johnson et al. 2015). Interestingly, PSR J1023+0038 also exhibited optical pulsations during the sub-luminous disk state, which makes it the first millisecond pulsar ever detected in optical (Ambrosino et al. 2017). Papitto et al. (2019) argued that the pulsed optical emission orig- inates neither from magnetically channelled accretion nor rotation-powered pulsar magnetosphere, but synchrotron emission from the intrabinary shock between the pulsar wind and the accretion disk. This would imply that the rotation-powered activity of a pulsar persists in the sub- luminous disk state. |
Processed_The_Flare-Dominated_Accretion_Mode_of_a_Radio-Brig.txt | Fig. 1.— (a) The black curve shows the reduced EPIC light curve of 3FGL J0427.9−6704 (pn + MOS 1/2), while the gray one indicates the flaring background obtained by pn. (b & c) Two arbitrary zoomed-in views of (a) with 1-σ uncertainties of the EPIC light curve. No correlation is found between the two curves. |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.