A substorm-associated enhancement in the XUV radiation measuring channel observed by ESP/EVE/SDO

Transcription

1 Research in Astron. Astrophys. Vol.0 (200x) No.0, Research in Astronomy and Astrophysics A substorm-associated enhancement in the XUV radiation measuring channel observed by ESP/EVE/SDO Yan Yan 1,2,4, Hua-Ning Wang 1,2, Chao Shen 3 and Zhan-Le Du 1,2 1 National Astronomical Observatories, Chinese Academy of Sciences, Beijing , China; yyan@bao.ac.cn 2 Key Laboratory of Solar Activity, Chinese Academy of Sciences, Beijing , China; 3 School of Natural Sciences and Humanity, Harbin Institute of Technology Shenzhen Graduate School, Shenzhen , China; 4 University of Chinese Academy of Sciences, Beijing , China Received 2014 April 7; accepted 2015 December 1 Abstract Comparing the ESP/EVE/SDO flux data of Feb , with the counterparts of XRS/GOES and SEM/SOHO, we find that there is an enhancement, not demonstrated in the two latter datasets. The enhancement, possibly regarded as a flare at a glimpse, nevertheless, is not involving an energy-releasing occurrence from the sun. Based on the enhancement, we combine data from SXI/GOES 15 into a synthesize analysis, becoming aware of that it attributes to a particle-associated enhancement in the XUV radiation measuring channel. Paradoxically, it seems to be somewhat of a particle-avalanching process. Prior to the event, a moderate geomagnetic storm has taken place. Subsequently, while the event is proceeding, a geomagnetic substorm is observed simultaneously. Therefore, the particles, though unidentified, are probably energetic electrons induced by substorm injection. Key words: Sun: X-rays, gamma rays Sun: flares Sun: solar-terrestrial relations 1 INTRODUCTION The solar electromagnetic irradiance had been studied for a very long time, nearly throughout the entire astronomical history, from the era of ancient time in optical, to the modern times, extending longward to radio, shortward to ultraviolet, X-ray, even in γ ray. The ultraviolet irradiance, a considerable parameter for space weather, however, was less studied until human beings marched into the space age. In the very beginning, the observational data were less reliable, and often discontinuous, confined to balloons or rockets(lean, 1987). After that, several spacecrafts launched into space, solar astronomers began to, systematically say, pay more and more attention to the irradiance range of Extreme UltraViolet (whereafter EUV) and Soft X-ray (whereafter XUV). What we focus primely on here is the so-called XUV, the electromagnetic wave with the wavelength ranging from 0.1 to 10nm, the most sensitive one applying to track solar eruptions. It is generally recognized that XUV irradiance can be the most agile tracer of solar flares. In order to surveille them in realtime, the Geostationary Operational Environment Satellite (whereafter GOES) family were successionally sent into the geostationary orbit, from 1974(Grubb, 1975) to today. So far, GOES 15 is in operation, and has been the primary satellite, equipped with the solar X-Ray Sensor (whereafter XRS), the Soft X-ray Imager (whereafter SXI), and other instruments. More recently,

2 2 Y. Yan, et al. Table 1 Parameters of the unusual enhancement Start time Peak time End time Duration Peak flux Relative increment 08:48 09:18 10:20 90min Wattsm 2 40% launched into space on Feb 11th 2010, the Solar Dynamical Observatory (whereafter SDO)(Pesnell et al., 2012), the first instrument of NASA s Living With a Star Program, helps us unveil both the sun s interior characteristics and its exterior influence on the Earth and its nearby space. The Extreme ultraviolet Variability Experiment (whereafter EVE)(Woods et al., 2012) on board SDO measures the EUV irradiance of the sun from 0.1 to 105nm, including the Multiple EUV Grating Spectrographs and EUV SpectroPhotometer (whereafter ESP)(Didkovsky et al., 2012a), a broadband spectrograph and almost the heritage of the Solar EUV Monitor (whereafter SEM) on board the Solar and Heliospheric Observatory (whereafter SOHO)(Judge et al., 1998), holding the highest temporal-resolution of 0.25 second, the champion of peers. ESP has two orders: the first order covers four channels of 18.2, 25.7, 30.4, 36.6nm in measuring the line intensity of four spectral lines, whereas, the zeroth one covers the broadband between 0.1 and 7nm, primarily for space weather. Fruitful results were seen in the journals relevant to ESP/EVE. Didkovsky et al. (2011) discovered the 5-min oscillations in the corona and Didkovsky et al. (2013) further found its variability. Both papers talked about the EUV/XUV radiation directly relevant to the sun. Space weather community nearly considered the fluctuations of XUV radiation to be solar flares as well, and Du & Wang (2012) gave a tentative relationship of solar flares with both sunspot and geomagnetic activity. This letter, in another way, covering an unusual enhancement observed by ESP/EVE/SDO, and analysing dozens of imaging data from SXI/GOES 15 together, will enumerate a counterexample which is not a solar flare, but substorm-associate enhancement indeed. Section 2 describes observations in detail and makes data comparison and analysis. At last, summary and discussion will be talked in section 3. 2 OBSERVATION AND DATA Three soft X-ray light curves on flux data on February 6th 2011 are introduced, respectively, see Figure 1. Figure 1(a) comes from XRS/GOES 15 in 24 hours, with the nm wavelength band; Figure 1(b) is based upon ESP/EVE/SDO, with the broader wavelength band of 0.1-7nm; Figure 1(c) is originally taken by SEM/SOHO. The comparison work will be talked later. Except for the flux data, we also take up dozens of images obtained by SXI(Hill et al., 2005; Pizzo et al., 2005) onboard GOES-15 to see what has happened in the meantime. The images are all level-1 fits data with the filter of BEA covering the spectral band pass of nm. The exposure time of them is approximately 1s. February 6th 2011 was indeed a calm day on solar activity. The solar disk appeared merely two active regions, NOAA and 11153, both of which kept silence throughout the 24 hours, simple in magnetic configuration and small in sunspot area, originally from Space Weather Prediction Center solar region summary. GOES soft x-ray flux diagram is plotted as Figure 1(a). From the plot, one can clearly learn that no flare happened, even in B class. Analogously, ESP is another version to measure soft x-ray flux in higher cadence. We survey ESP zeroth data in the same interval as XRS has, and find something unlike which is not displayed in the plot of XRS. As shown in Figure 1(b), a remarkable increase is burst into this plot. This soaring resembles a solar flare so much, both in shape and duration. The main phase of the event, beginning at about 08:48UT, afterwards, mounts up rapidly. It peaks at 09:18UT, with the flux of Watts m 2. The relative increment almost reaches 40% comparing with the background flux. The phase lasts almost 100 min, behaving a solar-flare like process, whose main features are listed in table 1. Interestingly, it is contradictory. Whether it is a true flare or not perplexed us at that time. We pick the SEM/SOHO data up at the corresponding time to reconsider which is authentic. The result plotted in Figure 1(c), compatible with our hypotheses, does not exhibit any notable increase yet. SDO travels geosynchronously whereas SOHO shuttles its odyssey around the L1 orbit where is far from the Earth.

3 A substorm-associated enhancement of XUV radiation 3 Fig. 1 Time-intensity profiles of three monitors on February 6th (a) nm XUV variation from XRS/GOES 15 (b) 0.1-7nm XUV variation from ESP/EVE/SDO (c) nm XUV/EUV variation from SEM/SOHO Therefore, the enhancement observed by ESP/EVE is possibly not a sun-associated, but a process taking place in earth s nearby space likely (Didkovsky, 2012b). Didkovsky et al. (2012a) has clearly stated that it is possible to use the Cdark measurements (available at any time) as a proxy for dark counts in the science bands, and as a proxy for particle-background signal. In view of the statement, we have checked the simultaneous data in the dark band, which is completely closed at all times, as seen in Figure 2. From the figure, we can see that the Cdark channel also demonstrates a time-coincident peak. It suggests us that the Cdark peak has a certain connection with the counterpart appearing in the XUV measuring channel. The absolute scale of the ESP zeroth order and Cdark channel is different because of their different sensitivity to particle contamination. For clarifying the truth of the event, we examine further the data of SXI. Co-equipped onboard GOES as XRS, SXI is responsible for imaging the Sun in soft x-ray wavelength, tallying with the wavelength ESP covers. We select the data which comprise the period of the event and check them up in image processing with care. Twenty-four images in total from 08:00UT to 11:04UT with eight-minutecadence are investigated, all of which are in pixels. We cut out them into pixel images, and co-align them to the first image taken at 08:00UT, that is Figure 3(a). Since 08:48UT, unexpected dazzling specks mushroom onto the images, besides, the amount of the specks on each image is beginning to raise rapidly. Figure 3(c), near the peak takes the largest amount of the specks. For interpreting this more clearly, we set to our work on separating the images into various parts. The agglomerate, brilliant and congregated portions stand for the XUV-intense regions (whereafter XIR) belonging to the sun, whereas the scattering bright specks are unrelated. We mark the XIR alphabetically,

4 4 Y. Yan, et al. Fig. 2 Time profile of the dark band counts from UTC 8:00 to 11:00 on Feb 6, The profile is shown with a 10-sec smoothing window. see Figure 3. Region A, B, C and D are all XIR, however, Region E is a void where any XUV blocks are seen, but numerous specks. Afterwards, for the sake of comparing the images to the ESP s flux plot quantitatively, we cut each square region into independent integration. The results are itemized in Table 2. The Table gives the result of integration of the full disk as well. In the column of full disk, we can know that the value in the row of 09:20UT is the maximum around the time interval, according with the peak time in Figure 1(b). Figure 3, plots of the relative increment of integrated DN of the selected regions in Figure 2, according to the formula DN r (%)= DNi DN1 DN 1 100%(i = 1 24), where DN r means relative increment, and DN i represents the DN integral of each region in each image, reflects the trends demonstrating the relative increment of corresponding region in Figure 3 in the time interval. It is worthy of note that all XIR are descending at just time the peak happens. Hence, there should be an external cause driving the total flux climbing up. That is, as a consequence, the rapid increment of the specks. Region E, a void of XUV radiation, disagrees with the XIR. And, the result of Region E, shown in Figure 4, as we expect, coincides with the shape of the light curve of Figure 1(b), confirming that the two plots have a fine time-correspondence relationship, although the two flux value of which may not equivalent. The reason for the different amplitude of the fluxes was perhaps that the two instruments underwent different astronomical position, angle of view, clear aperture, and the event itself has its own angular distribution. By quantitative image analysis, we confirm that the flare-like enhancement is not attributed to the sun, but the cumulative effect of the speckles spreading all over the images. What the speckles are should be a very interesting and opening question. Naturally, the geosynchronous relativistic electron enhancement (whereafter GREE) is a feasible consideration relating to

5 A substorm-associated enhancement of XUV radiation 5 Fig. 3 Selected regions in the pixel images taken by SXI/GOES-15 at 08:00UT(a), 08:48UT(b), 09:20UT(c) and 10:24UT(d), February 6th Each image is marked in five different regions alphabetically. Region A, B, C and D are all XIR, whereas Region E is a void of XUV radiation. the event. GREEs have, mostly, strong connection with geomagnetic activity. A moderate geomagnetic storm has taken place with a minimum Dst index of -63nT on UTC 22:00 Feb 4, and lasts six days at least, from the beginning phase to the recovery phase. The happening time of the event is on the recovery phase of the magnetic storm, while the magnetic field is still active, as Figure 5 shows. Apart from the magnetic storm, meanwhile, the speed of the solar wind stream goes up to large value, with a maximum of 670km/s, and preserves rather long time (almost 50 hours or more) with the speed higher than 500km/s. Rathore et al. (2015) highlighted that the importance of solar wind parameters on space weather, two of which were the solar wind speed and southward component of the Interplanetary Magnetic Field(whereafter IMF). Figure 6, not only demonstrates the situation of the solar wind stream, also, more attracted, conveys the message to us that within the days during the magnetic storm, IMF is always pointing negative, indicating that the southward component of which has arisen. The deep depression of the plot shown in Figure 6, acts as the culprit who is responsible for the geomagnetic storm, however, other drops with duration more than three hours though less strength will, with high probability, develop geomagnetic substorms. Substorm, the other fundamental magnetospheric process, unlike geomagnetic storm, is a frequent occurrence, almost daily, in the magnetosphere, and takes shorter time-scale, often 2-3 hours. The processes always manifest colorful observational phenomena, making them hard to be classified. In addition, the relationship of geomagnetic storm and substorm is as well as a troublesome debate, however, some evidence shows that substorms become more frequent during a geomagnetic storm, especially on its recovery phase, while the southward IMF has still continued. In this case, we have examined the

6 6 Y. Yan, et al. Table 2 Integrated DN of the selected regions in each image Time Region A Region B Region C Region D Full Disk Region E 08: : : : : : : : : : : : : : : : : : : : : : : : one-minute data of AE, AU, AL and AO indices at the right time when the event is ongoing, as seen in Figure 7, noticing that a classical substorm is coming through the timescale. AE index, the indicator of the substorm climbs up at 904nT and the peak time happens along the peak of the enhancement. Still, the time-profile of the substorm coincides with the enhancement. On account of the geophysical condition, GREEs in the outer radiation belt, where geosynchronous orbit, about 6.6 Re live in, are expected. This evidence supports our initial speculation that the enhancement arises not from the sun, but from a geophysical event, probably a substorm accelerating electrons into energetic state which containments the XUV signal. Furthermore, we examine the electron fluxes observed by the MAGnetospheric Electron Detector (hereafter MAGED) onboard GOES 15, so as to have a view on what has happened at that time. MAGED is an instrument suite, consisting nine telescopes covering nine different directions, each with five electron energy channels from 30KeV to 600KeV. We pick up the data on the consistent timescale, UTC 08:00 to 11:00, and the plotted result is shown in Figure 8. The violet, green, red, blue and grey curves respectively indicate the omni-direction electron flux with the energy channels of 40KeV, 75KeV, 150KeV, 275KeV and 475KeV, as the caption notes. From the figure, we can read that there are still peaks on the time scale, and the peak amplitude appears somewhat middle-focusing effect. That means that the electron fluxes with the energy channels from 75KeV to 275KeV, especially in 150KeV, may constitute the major contribution of the enhancement observed by ESP. Substorms often introduce the so-called substorm-injection, rapidly increasing in tens to hundreds of KeV electrons in nearby space, and the injection in this case presents the nature of dispersion due to out of synchronization of the electron flux raising in different energy channels. That is to say, the more energetic electrons arrive at the synchronous orbit earlier than the less ones. Thanks to the enhancement observed by ESP, it is accessible to initiate the multi-instrument observation, and cross-match of the space physics event can be therefore carried out. From figure 8, it is clearly seen that the earliest arriving time of the flux-raising is after 9:00UT, around 9:06UT, whereas the enhancement observed by ESP has begun at 8:48UT, about a quarter before the former. SDO and GOES 15 are both synchronous satellites, but they are placed in two different longitudes, 102 W and 89.5 W, living almost one magnetic local time apart. The injection may burst at

7 A substorm-associated enhancement of XUV radiation 7 Relative increment of integrated DN of the selected regions (%) Region A Region B Region C Region D Region E :00 08:20 08:40 09:00 09:20 09:40 10:00 10:20 10:40 11:00 Universal Time Fig. 4 Plots of the relative increment of integrated DN of the selected regions. We take the values on 08:00UT as the initial values, and other values are calculated from DN r(%)= DN i DN 1 DN 1 100%(i = 1 24). Three reference lines (the dot line, dash line and dot-dash line) are ticked to represent the start time, peak time and end time, respectively. midnight, on the west of the two instrument, diffusing eastward from SDO to GOES 15, and the angular velocity of the eastward propagation is estimated by12 /18min, or 40 /min, approximately. 3 SUMMARY AND DISCUSSION ESP/EVE/SDO records an enhancement in the XUV radiation channel on the band of 0.1-7nm which the zeroth data cover. In the meantime, there is not a remarkable upheaval in the corresponding time in the light curves of both XRS/GOES and SEM/SOHO. We then examine some relevant images taken from SXI/GOES 15, and are conscious that the contribution of the speckles on the images is responsible for the enhancement. When we anatomize the images into parts and integrate them individually, a strong time-correspondence relationship between the enhancement and the mount of the speckles is unveiled. The instance we encounter is scarce or accidental, because some requirements must be satisfied at the same time. When the sun turns to be inactive, the signal enhancement in the XUV observing channel might be prominent from the background noise, and should be detected in time. Except for this case, other few cases alike would be further reconfirmed in the future. If they were true, we would hope to study the relationship of the flux intensity and the geomagnetic index. Observed energetic particles at geosynchronous orbit are considered to be a variety of ingredients. They are protons, electrons, α particles, heavy ions, and some peculiar particles not ever-present. More complicatedly, they have different origins. It is not effortless to say from where they come. But we can

8 8 Y. Yan, et al. Fig. 5 Hourly Dst index from Feb to Feb , adapted from WDC for Geomagnetism, Kyoto. A moderate geomagnetic storm has taken place, with the minimum Dst index of -63nT. Fig. 6 Time profiles of one-minute proton speed and z component of magnetic field in GSM from Feb 4 to Feb , adapted from WDC for Geomagnetism, Kyoto. The proton speed(red line) reflects the solar wind speed, whereas the z component of magnetic field in GSM (black line) means the z component of IMF. The left vertical dot-dash line refers to the onset of the magnetic storm, whereas the right one means the peak of the enhancement. view the flux diagrams of energetic particles observed at geosynchronous orbit to determine the particles whether them belong to solar-origin or not. The particles are unlikely to be protons because neither major solar eruption takes place in the meantime, nor solar energetic proton event has been recorded. Otherwise, they are rather not heavier ions due to the great amount of the speckles. The reasonable and comprehensible explanation may be that they are energetic electrons. Still, they are not an event of solar impulsive energetic electron, because the two active regions are not seemed to uphold the coronal magnetic topology that Li et al. (2013) has proposed. One day before the event, or February , the solar wind speeded up to 670km/s, injecting much more kinetic energy into the magnetosphere. As a result, the supersaturated magnetosphere would trans-

9 A substorm-associated enhancement of XUV radiation 9 Fig. 7 Time profiles of one-minute AE, AU, AL and AO indices from UTC 8:00 to 11:00 on Feb 6, 2011, adapted from WDC for Geomagnetism, Kyoto. Three reference lines(dash line) are ticked to represent the start time, peak time and end time, respectively, the same as Figure 4. fer the redundant energy to the electrons nearby the magnetotail by reconnection or any other unclear accelerating processes. SDO and GOES 15 were both geosynchronous satellites, with the longitudes of 102 W and89.5 W on February While the event was going on from 08:48UT to 10:20UT, the two satellites were both, coincidentally, turning around at the moment of midnight in local time, working at the nightside facing on the magnetotail within the magnetosphere, where was vulnerable to the danger of energetic electrons accelerated from the location of magnetic reconnection. Simultaneously, a stormtime substorm is ongoing, making substorm-injections possible to be identified at geosynchronous orbit. The substorm-injection induced electrons will, in bulk, pass by some instruments working at geosynchronous orbit, playing a role of the black sheep in the aliasing of signal and noise. Energetic electron storms occur frequently, especially in the years which are not around the maximum of a solar cycle, for the reason that coronal holes, which are frequently appearing in the corona in the quiet years, often drive the solar wind up to very high speed. In consequence, nearby space becomes energetic, creating a great volume of accelerated electrons which can damage the satellites traveling in scheduled orbits. In result, for space weather, it is equally essential to be considered in the years while solar activity is lower. In the end, we will reiterate that the event is a rare case, whereas it is still worthwhile even in a low probability. It reminds us that if we examine the datasets in the timescale of weaker solar activity and stronger geomagnetic activity, especially in the recovery of a geomagnetic storm, we will take some care. Furthermore, if it were not a real flare, one would think that it might be related to a geophysical event, probably a substorm.

10 10 Y. Yan, et al. Fig. 8 Time profile of the electron fluxes measured by MAGED onboard GOES 15 from UTC 8:00 to 11:00 on Feb 6, The violet, green, red, blue and grey curves represent the omni-direction electron flux with the energy channels of 40KeV, 75KeV, 150KeV, 275KeV and 475KeV, alternately. Acknowledgements The authors thank the anonymous referee for his/her valuable suggestions on greatly improving the manuscript. This work is jointly supported by the National Basic Research Program of China (973 Program) through grant 2011CB and the National Natural Science Foundation of China (NSFC) through grants , and Y.Y. owes a great deal to the enlightenment of Prof. Leonid Didkovsky s opinions on the accomplishment of the work. Dr. Don Woodraska at the LASP/CU team provided the L0D data on our request. Prof. Ronglan Xu gave us some valuable advices on space physics. We acknowledge the XRS/GOES, SXI/GOES, MAGED/GOES, ESP/EVE/SDO, SWEPAM/ACE, MAG/ACE and SEM/SOHO consortia for usage of the data. The WDC for Geomagnetism, Kyoto is also strongly acknowledged. References Didkovsky L., Judge D., Kosovichev A. G., Wieman S., & Woods T., 2011, ApJ, 738, L7 Didkovsky L., Judge D., Wieman S., Woods T., & Jones A., 2012a, Sol. Phys., 275, 179 Didkovsky L., 2012b, private communication Didkovsky L., Kosovichev A., Judge D., Wieman S., & Woods T., 2013, Sol. Phys., 287, 171 Du Z. L., & Wang H. N., 2012, Research in Astron. Astrophys. (RAA), 4, 400 Grubb R. N., 1975, The SMS/Goes Space Enviroment Monitor Subsystem, NOAA Tech. Memo., ERL SEL-42, Boulder, CO Hill S. M., Pizzo V. J., Balch C. C., et al., 2005, Sol. Phys., 226, 255 Judge D. L., McMullin D. R., Ogawa H. S., et al., 1998, Sol. Phys., 177, 161 Lean J., 1987, J. Geophys. Res., 92, 839 Li C., Sun L. P., Wang X. Y., & Dai Y., 2013, A&A, 156, L2 Pesnell W. D., Thompson B. J., & Chamberlin P. C., 2012, Sol. Phys., 275,3 Pizzo V. J., Hill S. M., Balch C. C., et al., 2005, Sol. Phys., 226, 283 Rathore B. S., Gupta D. C., & Kaushik S. C., 2015, Research in Astron. Astrophys. (RAA), 1, 85 Woods T. N., Eparvier F. G., & Hock R., et al., 2012, Sol. Phys., 275, 115