The active region source of a type III radio storm observed by Parker Solar Probe during Encounter 2
L. Harra, D. H. Brooks, S. D. Bale, C. H. Mandrini, K. Barczynski, R. Sharma, S. T. Badman, S. Vargas Dominguez, M. Pulupa
aa r X i v : . [ a s t r o - ph . S R ] F e b Astronomy & Astrophysicsmanuscript no. ISSI_E2_Hinode-final © ESO 2021February 10, 2021
The active region source of a type III radio storm observed byParker Solar Probe during Encounter 2
L. Harra , , D. H. Brooks , S. D. Bale , C. H. Mandrini , , K. Barczynski , , R. Sharma , S. T. Badman , S. VargasDomínguez , and M. Pulupa PMOD / WRC, Dorfstrasse 33 CH-7260 Davos Dorf, Switzerlande-mail: [email protected] e-mail: [email protected] ETH-Zurich, Hönggerberg campus, HIT building, Zürich, Switzerland College of Science, George Mason University, 4400 University Drive, Fairfax, VA 22030 USAe-mail: [email protected] Physics Department and Space Sciences Laboratory, University of California, Berkeley, USA. 94720-7450e-mail: [email protected] Instituto de Astronomía y Física del Espacio (IAFE), CONICET-UBA, Buenos Aires, Argentina Facultad de Ciencias Exactas y Naturales (FCEN), UBA, Buenos Aires, Argentinae-mail: [email protected] Fachhochschule Nordwestschweiz (FHNW), Bahnhofstrasse 6, 5210 Windisch, Switzerlande-mail: [email protected] Universidad Nacional de Colombia, Observatorio Astronómico Nacional, Bogotá, Colombiae-mail: [email protected]
Received September 2020
ABSTRACT
Context.
To investigate the source of a type III radio burst storm during encounter 2 of NASA’s Parker Solar Probe (PSP) mission.
Aims.
It was observed that in encounter 2 of NASA’s Parker Solar Probe mission there was a large amount of radio activity, and inparticular a noise storm of frequent, small type III bursts from 31st March to 6th April 2019. Our aim is to investigate the source ofthese small and frequent bursts.
Methods.
In order to do this, we analysed data from the Hinode EUV Imaging Spectrometer (EIS), PSP FIELDS, and the SolarDynamics Observatory (SDO) Atmospheric Imaging Assembly (AIA). We studied the behaviour of active region 12737, whoseemergence and evolution coincides with the timing of the radio noise storm and determined the possible origins of the electron beamswithin the active region. To do this, we probe the dynamics, Doppler velocity, non-thermal velocity, FIP bias, densities, and carry outmagnetic modelling.
Results.
We demonstrate that although the active region on the disk produces no significant flares, its evolution indicates it is a sourceof the electron beams causing the radio storm. They most likely originate from the area at the edge of the active region that showsstrong blue-shifted plasma. We demonstrate that as the active region grows and expands, the area of the blue-shifted region at the edgeincreases, which is also consistent with the increasing area where large-scale or expanding magnetic field lines from our modellingare anchored. This expansion is most significant between 1 and 4 April 2019, coinciding with the onset of the type III storm and thedecrease of the individual burst’s peak frequency, indicating the height at which the peak radiation is emitted increases as the activeregion evolves.
Conclusions.
Key words. solar physics–
1. Introduction
Type III radio bursts are observed regularly on the Sun, and theirsources are beams of energetic electrons streaming outwardsalong open magnetic field lines. The largest Solar Energetic Par-ticle (SEP) events are due to shocks associated with coronal massejections (CME) and solar flares e.g. Reames (2017), thoughothers can be associated with jets. For example, Krucker et al.(2011) made a study of jets associated with flares as sources ofsupra-thermal electron beams. The flares in this paper all hadenergies higher than C2 GOES classification. However, elec-trons can be accelerated in smaller energy release events in ac-tive regions, when magnetic field in a coronal hole interacts with nearby closed magnetic field, and even in bright points that pro-duce jets (see the review by Reid & Ratcli ff e 2014).NASA’s Parker Solar Probe (PSP) mission has opened up anew way of probing type III bursts due to the close vicinity tothe Sun. Pulupa et al. (2020) have explored the statistics of typeIII bursts during the first two encounters of PSP. They found thatonly a few bursts occurred during encounter 1 (E01, October-November 2018, perihelion November 6th 2018), during whichthere was minimal solar activity. In encounter 2 (E02, March-April 2019, perihelion April 4th 2019), however, there were anumber of dynamic active regions, and a large number of typeIII radio bursts observed by PSP. In particular, two active re-gions (AR 12737 and AR 12738) were prominent during this Article number, page 1 of 13 & Aproofs: manuscript no. ISSI_E2_Hinode-final time interval and at solar locations such that radio waves wouldbe likely to reach PSP. AR 12738 was the larger active region,receiving its NOAA designation on April 6 as it rotated onto thesolar disk as viewed from Earth, but had existed at least 2 weeksprior and was visible in STEREO A / SECCHI data. Krupar et al.(2020) used radio triangulation of a single strong radio burst onApril 3 and showed its location to be consistent with a Parkerspiral emerging from AR 12738, indicating it was responsiblefor at least some of the radio activity observed at this time. Inaddition, Cattell et al. (this issue), find some evidence of corre-lating periodicities in AR 12738 and in radio at PSP, althoughthe interval they investigated is well after the time period studiedin this paper, once AR 12738 had rotated on disk and AR 12737had rotated o ff disk.AR 12737, meanwhile, was observed to develop out ofa coronal bright point near the East limb on March 31 andevolve and grow before reaching a more steady state on approx-imately April 6th. At this time, in addition to strong impulsivebursts which PSP measured throughout this encounter and whichKrupar et al. (2020) were able to associate with the larger activeregion, a significant type III radio noise storm occurred consist-ing of a huge number of smaller and more quickly repeating ra-dio bursts. In addition, this noise storm showed the interestingfeature of a systematic decrease in peak frequency with time atthe exact time that AR 12737 was emerging and developing. Thiswill be discussed further in relation to Figure 2, but here we sim-ply state that for the above reasons, it is well motivated to studythis active region in relation to the noise storm as this time. Inaddition, in contrast to the larger active region, AR 12737 is vis-ible on disk during PSP’s closest approach to the Sun (April 4th),and thus more likely to be measuring weaker radio events whichare harder to study at 1AU. Hence AR 12737 is observed at thistime with all the available Earth-based instrumentation includ-ing imaging spectroscopy, radio interferometry, magnetographsand more. Since the active region is not flare active during ourperiod of interest, we consider other possible sources of the typeIII bursts. Other possibilities are jets, micro-flares and active re-gion outflows.Jets are an important aspect of electron acceleration,and indeed have been put forward as a means to explainthe magnetic switchbacks that have been seen so clearlyby PSP (Bale et al. 2019; Kasper et al. 2019; Horbury et al.2020). Sterling & Moore (2020) have proposed that reconnectedminifilament eruptions can manifest as outward propagatingAlfvénic fluctuations that steepen into a disturbance as theymove through the solar wind. Mulay et al. (2019) showed a spa-tial association of active region jet and the interplanetary type-III burst source co-spatial with extrapolated open magnetic fieldlines. The bunches of type-III bursts occurred before and duringjet eruption suggesting particle acceleration begins before theEUV jet eruption.Another potential source of type III bursts are the blue-shifted regions found at the edges of active regions (e.g.,Harra et al. 2008). The blue-shifted regions have been put for-ward as one of the potential sources of the slow solar wind, andthey have been found to often have elemental abundances con-sistent with the slow solar wind (e.g., Brooks & Warren 2011).Del Zanna et al. (2011) have also studied these outflows and sug-gest that the continuous growth of active regions maintains asteady reconnection across the separatrices at the null point inthe corona. The acceleration of electrons in the interchange re-connection region between the closed loops in the AR core andthe outflow on open field lines at the boundary, produces a radio noise storm in the closed loop areas, as well as weak type IIIemission along the open field lines.In this paper, we explore the data from the second solar en-counter of PSP (E02) around perihelion on 4 April 2019, andinvestigate the possible source(s) associated with AR 12737 ofthe type III radio storm that occurs in the lead up to this time.
2. Instrumentation and data analysis
The Hinode EUV Imaging Spectrometer (EIS) instrument(Culhane et al. 2007) is an imaging spectrometer that has twonarrow slits that raster to build up images, and two slots. In thiswork we use studies with the 2 ′′ slit, where spectral images havebeen built up that cover the whole active region starting from 1April less than a day after the region first emerged. In addition,there is a fast 3-step raster with a cadence of 42 seconds on 1April starting at 17:00 UT lasting an hour. The field of view inthis case covers a strip towards the eastern edge of the active re-gion. We processed the EIS data using the standard calibrationprocedure eis_prep. The emission lines were fitted with single ormultiple Gaussian profiles (depending on the presence of knownblends) in order to extract plasma parameters such as Dopplervelocity, non-thermal velocity, and FIP (first ionization poten-tial) bias.The spectroscopic data were combined with AtmosphericImaging Assembly (AIA) data from the Solar Dynamics Obser-vatory (SDO; Pesnell et al. 2012) to provide context imaging andan understanding of the dynamical behaviour of the action regionas it emerged on 31st March and developed through to 7th April2019.Measurements of interplanetary type III radio bursts weremade by the Radio Frequency Spectrometer (RFS) subsystem(Pulupa et al. 2017) of the FIELDS instrument suite (Bale et al.2016) on the NASA Parker Solar Probe (PSP) mission (Fox et al.2016). The FIELDS / RFS system is a base-band radio receiverthat produces full Stokes parameters in the range 10.5 kHz -19.17 MHz, corresponding to plasma frequencies at radial dis-tances of ∼ R S to 1 AU (Leblanc et al. 1998). Voltage mea-surements are made on two ∼
7m crossed dipoles and digitizedinto 2 virtual receivers (the High Frequency Receiver - HFR andthe Low Frequency Receiver - LFR) in 128 pseudo-log spaced( ∆ f / f ≈ th April 2019 04:00:16 to 04:05:16 UT taken during theG0002 solar observation program. The solar observations weremade in ‘picket-fence’ mode, i.e. the total available bandwidth of30.72 MHz can be distributed within 80-300 MHz in 12 roughly
Article number, page 2 of 13. Harra et al.: The active region source of a type III radio storm observed by Parker Solar Probe during Encounter 2 log-spaced chunks, each of 2.56 MHz wide. Here, we analyse 4frequency bands at 107.5, 163.8, 192.0, 240.6 MHz.Since our aim is to find the origin of the type III noise stormmeasured at PSP from 1 to 6 April 2019, we have modelledthe coronal magnetic field of AR 12737 during that period insearch of large-scale or expanding field lines (i.e. ’open’ withinthe limitations of a local magnetic field model approach) thatcould be associated to the areas where EIS observes blue-shiftedregions. Our model is carried out by taking as boundary condi-tion the vertical component of the photospheric magnetic fieldderived from SDO / HMI observations. The magnetic field verti-cal component is computed from the line-of-sight magnetogramsdownloaded from the Joint Science Operation Center (JSOC,http: // jsoc.stanford.edu / HMI / Magnetograms.html). We selecteddata with a 720 ms resolution from that database.
3. A type III storm and the active region
During PSP encounter 2, frequent type III radio bursts were ob-served as described by Pulupa et al. (2020). We concentrate onthe period between 31 March 2019 - 6 April 2019.Figure 1 shows PSP / RFS data during a ∼ I in a 40 second window normalized to the mode (most proba-ble) value in that window, at frequencies 18.28-19.17 MHz (thehighest two frequency bins). The 2nd panel is the full RFS / HFRStokes I spectogram, at full spectral and temporal resolution. Inthese panels, one can see both larger intensity type III bursts,extending over the full frequency band and exhibiting the classi-cal frequency drift, and the more impulsive, frequency-localizedfeatures associated with type III radio storms. The third panelshows, as intensity, the number of 7-second intensity measure-ments that exceed 2 × the mode value (as per the top panel)in a 40 second window (hence the maximum value is 5), asa function of measurement frequency. This shows the locationin time-frequency of ‘bursts’ rather than background (galactic)noise. The bottom panel is the frequency of the maximum nor-malized intensity (in the panel above). While not shown in Fig-ure 1, Stokes linear polarization U / I and Q / I are measured forthis type III storm to be less than %2. While strong linear, orcircular, polarization can be an indication of mode conversionprocesses (e.g. Zlotnik 1981) and, therefore, a clue to emissionmode (fundamental or harmonic), the relatively weak polariza-tion signature observed here probably indicates strong scatteringin density fluctuations. We take it as an indication of fundamen-tal emission.Furthermore, as seen in Figure 2, Panel 6, the frequency ofmaximum radio intensity is decreasing with time over this inter-val. As shown in Panel 7, this implies a source moving to higheraltitudes (and hence lower heliospheric plasma density) or an ex-panding source region whose density decreases with lateral ex-pansion. The lower panel in Figure 2 shows the inferred heightfrom a heliospheric density profile n e ( r ) (Leblanc et al. 1998),assuming radio emission at the fundamental f pe as is commonlyassumed for type III radio storms (Morioka et al. 2015).During this time period, AR monitored the solar radio flux andrecorded an increase towards the end of the week. The top panelof Figure 2 plots the daily-averaged solar radio flux for 245MHz. A distinct increase can be seen from 4 th April, which isdue dominantly to AR 12737 as it is the only region on the diskat that time. We note that AR 12738 is also visible from April 6,and may contribute to the increasing RSTN flux after that date(Krupar et al. 2020).During E02, PSP does not observe the type III electronbeams in situ causing the radio emission. This is not surpris-ing since simple ballistic modeling and a PFSS model (seeBadman et al. 2020, for details on method) shows that PSP didnot connect magnetically to either active region during this time.Such a connectivity would make the origin of the type III burstsless ambiguous but unfortunately is not possible in this case.Type III radio emission is widely beamed, and bursts can of-ten be seen by widely separated spacecraft, and thus proximityto one active region or another is not a strong constraint on thesource location. As such, it is plausible for PSP to be receivingradio emission from electron beams injected by AR 12737 andso we proceed with a full analysis of its dynamics and propertiesto determine its nature as a radio source.Our goal is to understand where the source of the SEPs couldbe within this active region. The first check is to see if thereare any flares, but the active region produces no flares greaterthan GOES ‘A’ level (see second panel, Figure 2). There are nosignificant energy release events in this time period.The active region emerges as a simple bipole on 31 March2019. Figure 3 shows AIA images from its emergence until 7April 2019. In the first two time-frames the active region is com-pact. By 3 April, there are many more bright loops, covering alarger area, and the edges of the active region are showing ex-tended structures. In the same time period, as shown in Figure 2,the type III storm of interest commences and shows a transientand monotonic change in its peak frequency (implying sourceevolution), at the same time as the active region evolves.To determine if the AR is a radio source, we produced 2.56MHz bandwidth and 10s time-averaged MWA solar maps. Thesolar emission in the image was defined by choosing a 5- σ threshold, where σ is the RMS noise computed over the regionfar from the Sun in the image. The resultant MWA radio imageswere converted into brightness temperature (T B ) maps followingthe procedure described in Mohan & Oberoi (2017). The contourplots of the T B maps are shown in Figure 4 for the 4 frequencybands. The location of the contours suggests radio sources at162 MHz, 192 MHz and 240 MHz are associated with the activeregion. No other radio source was observed in the MWA solarmaps.Figure 5 shows the global coronal structure derived from ourmodeling for 1 and 4 April for which we have used the clos-est in time magnetograms to each EIS observation. The figureis constructed superposing field lines computed using di ff erentlinear-force-free field (LFFF) models, i.e. ∇ × B = α B , where B is the magnetic field and α is a constant (see Mandrini et al.2015, and references therein for a description of the model andits limitations). Because we are interested in the large scale coro-nal configuration and the potential presence of ‘open’ field lines,we selected a region four times larger than that covered by AR12737, and centered on it, for each model. Each of these mag-netic maps was embedded within a region twice as large padded Article number, page 3 of 13 & Aproofs: manuscript no. ISSI_E2_Hinode-final with a null vertical field component to, on one hand, decreasethe modification of the magnetic field values since the methodto model the coronal field imposes flux balance on the full pho-tospheric boundary (i.e. flux unbalance is uniformly spread on alarger area and the removed uniform field is weaker as the areais larger), and, on the other, to decrease aliasing e ff ects resultingfrom the periodic boundary conditions used on the lateral bound-aries of the coronal volume. Doing so, we are able to discrimi-nate field lines that connect to the surrounding quiet-Sun regionsfrom those that are potentially ’open’ magnetic field lines as theyleave the extrapolation box. In particular, field lines ending in anopen circle in Figure 5 are those that leave the computationalbox shown in each panel. In all our models the height above thephotospheric boundary was 400 Mm.The free parameter of each of our LFFF models, α , has beenselected to better match the shape of the observed loops in AIA193 Å (see Figure 3). To do this comparison, the model is firsttransformed from the local frame to the observed frame as dis-cussed in Mandrini et al. (2015) (see the transformation equa-tions in the Appendix of Démoulin et al. 1997). This allows adirect comparison of our computed coronal field configuration toAIA EUV images obtained at almost the same time and shownas background in Figure 5. Furthermore, in order to determinethe best matching α values we have followed the procedure dis-cussed by Green et al. (2002). AR 12737 is a mainly bipolar ARthat emerges close to the eastern solar limb around 31 March; itexpands and decays as it evolves on the solar disc, until it dis-appears on the western limb on 7 April. During this period onlya few minor flux emergence episodes are seen. The two bottompanels in Figure 5 illustrate very well that the blue-shifted regionis mostly consistent with large-scale or expanding magnetic fieldlines whereas the red-shifted region is mostly consistent withclosed loops.Figure 6 shows EIS Fe xii ff ects (Kamio et al. 2010). This isan important step in producing accurate velocities since EIS doesnot have an absolute wavelength calibration. The maps in Fig-ure 6 therefore show Doppler velocities relative to a chosen ref-erence wavelength. In this case we used the mean centroid inthe top part of the raster FOV. The neural network model alsoutilizes instrument housekeeping information from early in theEIS mission, so the correction is less applicable to more recentdata and often leaves a residual orbital variation across the raster.We therefore corrected the velocities in a final step by modelingthis residual e ff ect (see Appendix of Brooks et al. 2020). Forthe non-thermal velocity calculation we followed the procedureoutlined by (Brooks & Warren 2016).The AR grows and expands during this period. Althoughthere are no rasters between 1 April and 4 April, we can clearlysee that between these two dates the area of the blue-shifted re-gion to the eastern side of the active region is larger. The non-thermal velocities in the same region also expand in area andincrease in magnitude. This is consistent with the images in Fig-ure 3 which show clear expansion of the active region. The blue-shifted region also coincides with large-scale or expanding mag-netic field lines from our modelling (Figure 5), so plasma canfind a way to escape into the solar wind. We examined the area of the blue-shifted region more quantitatively in Figure 7 (top leftpanel). It shows the number of pixels in the blue-shifted region.This area of blue-shifted plasma increases as the AR evolves.We also measured the FIP bias in the outflow region (high-lighted by white boxes) as it crossed the disk (Figure 7; top rightpanel). The FIP bias is enhanced by about a factor of 2 abovephotospheric abundances throughout the period of the observa-tions, but appears to increase on April 4th, when the blue-shiftedarea also reaches a maximum. The FIP bias then slowly de-creases over the subsequent 5 days. The increase on the 4th isnot dramatic, but it is about 50% larger than measured on the1st. This is potentially indicative of SEPs, or is at least consis-tent with SEP activity, since they show a higher FIP bias in-situthan is measured in the slow solar wind (Reames 2018).To measure the FIP bias we used the Si x x ff ects. We measure the electron density using the Fe xiii / <
10 eV) FIP elements observed by EIS. In this casewe used lines of Fe covering a broad range of temperatures(0.52 – 2.75 MK). These are then used to compute the di ff er-ential emission measure (DEM) distribution from the observedintensities. With the DEM established, we can model the inten-sity of the high ( >
10 eV) FIP S line. Since we used low FIPelements to derive the DEM, the calculated S x ff ect is operating. The ra-tio of computed to observed S x – The AR is a radio source and its dynamical evolution coin-cides with the evolution of the peak emission frequency ofthe dominant type III radio storm observed by PSP at thistime. – The active region has just emerged and as it evolves the mag-netic field lines expand. In particular, at the edges of the ac-tive region. – The edges of the active region both show increased Dopplervelocities, increasing areas of upflows and increasing mag-nitude of upflows and non-thermal velocity between the 1stand 4th April. – The active region does not flare or have jets.We conclude that the active region is the most likely source ofthe energetic electron beams causing the type III radio storm,and more precisely, that the extended blue-shifted region couldbe a source. We also conclude that the changing nature of typeIII bursts (peak emission at a higher altitude or lower density
Article number, page 4 of 13. Harra et al.: The active region source of a type III radio storm observed by Parker Solar Probe during Encounter 2 region) must be related to the evolution and expansion of theactive region. We investigate this further in the next section withhigh cadence observations.
4. High cadence observations from Hinode EIS andcomparison to PSP data
From the foregoing discussion, it is clear that the blue-shifted re-gion is the most likely source of the SEPs within the studied ac-tive region. In this section, we analyse a high cadence EIS datasetto search for dynamical behaviour, similar to that which is seenin the PSP data, during a one hour period starting from 17:00UTon April 1. The EIS data consisted of a series of fast rasters overa small field-of-view, each of which took 41 seconds to com-plete. As already noted, the type III storm consists of continuoussmall and frequency localized bursts, and these are repeating ontimescales of minutes. This is another property that may helpdiscriminate the source region.In the right panel of Figure 8, the EIS Doppler velocity inthe blue-shifted region is shown above, and the PSP data isshown below for the one hour time period. The Doppler veloc-ities in this region are showing small but continuous variations,also on timescales of minutes. This is consistent with the na-ture of the type III bursts, and also with analysis carried outby Ugarte-Urra & Warren (2011). Ugarte-Urra & Warren (2011)also found that the blue-shifted outflow regions showed tran-sient blue wing enhancements within the 5 minute cadence of thetheir observations. The errors on the Hinode EIS Doppler veloc-ity measurements are on the order of a few km / s, which meansthese fluctuations are on the edge of detectability for these mea-surements.We investigated the properties of the AR core to see if itcould potentially be the source through small-scale brightenings.We analysed the variability of the AR core in AIA data duringthe same one hour time period as the EIS high cadence scan.We already ruled out significant flaring during this time, but it isimportant to check for any small scale micro-flaring, so we pro-duced a running di ff erence movie to highlight any brighteningsthat occur in the core. We then extracted the number of brighten-ings which were defined as having an area >
15 AIA pixels andhave an intensity enhancement compared to the previous imageof at least 100 DN. Figure 9 shows the light curve of the wholeactive region. The longer timescale changes that are seen over ≈
30 minutes are due to the intensity increase of ‘new’ loops.The plot below shows when and how many brightenings occurin the core, and we find that they do not happen consistently dur-ing the whole hour period. We also searched in the XRT data forjets, and found none. Hence we infer that the fluctuations seen inthe AR core are not viable as sources of the continuous type IIIbursts seen by PSP.
5. Conclusions
We have analysed the behaviour of AR 12737 during the timeperiod of 31 March - 6th April 2019 around perihelion (April 4)of the second encounter of PSP. During this time, on a backdropof larger, more impulsive type III bursts, PSP / FIELDS detectednumerous small type III bursts constituting a radio noise storm.These type III bursts: – were rapid and persistent during the time interval, – exhibited a decreasing peak frequency indicating a sourcealtitude which is climbing in time, or a source region whichis becoming more rarefied with time, – both such possibilities are consistent with an expandingsource region becoming more open to the solar wind.AR 12737 is the most probable candidate source region forthis type III noise storm. It is seen to emerge near the east limbat the same time as the radio noise storm develops. Between 1and 4 April, as the noise storm evolved as described above, thisactive region also showed significant changes: – the area of the blue-shifted outflow region increased by anorder of magnitude, – the FIP bias increased in the blue-shifted region by a signif-icant amount (consistent with an increase in SEPs; Reames2018), – the whole active region expanded and, consequently, large-scale or expanding magnetic field lines anchored at the ARedge are more evident (including in the outflow region). – the magnitude of the Doppler velocity and the non-thermalvelocity increases significantly as the active region expandsin its first few days of formation.The behaviour and changes in the AR during this time periodare consistent with the source of the type III bursts being theblue-shifted outflow region. We also explored the dynamics ofthe outflow region which do show fluctuations on the time scaleof the cadence of the observations. However these fluctuationsare close to the measurement limit of the instrument, and providea tantalising hint of dynamics. The high cadence aspect of thesemeasurements is key to further understanding, and we encouragethe future observing campaigns with PSP aim to have some fastcadence measurements.In addition, the expansion of the blue-shifted area of the ac-tive region may o ff er an insight into the generation mechanism oftype III storms. At least in this example, the decreasing peak fre-quency of the noise storm with time is suggestive of a scenario inwhich an expanding open field region allows the energetic elec-trons to more readily escape and thus produce their peak emis-sion higher in the corona. Alternatively, the increasing open fieldregion may be allowing more plasma to evacuate into interplane-tary space, causing a rareification of the source region and hencea decreasing plasma frequency. The EIS composition measure-ments could be consistent with either scenario. Both an increasein escaping energetic particles and / or source plasma would leadto a greater fraction of the total emission in the outflow areas wemeasured being contributed by higher FIP bias plasma; thereforeincreasing the mean value. The decreasing trend in FIP bias afterthe peak on the 4th is consistent with a decrease in the magnitudeof the type III bursts. Acknowledgements.
We acknowledge support from ISSI for the team 463, enti-tled ’Exploring The Solar Wind In Regions Closer Than Ever Observed Before’.The work of DHB was performed under contract to the Naval Research Labo-ratory and was funded by the NASA Hinode program. LH is grateful for SNSFsupport. CHM acknowledges financial support from the Argentine grants PICT2016-0221 (ANPCyT) and UBACyT 20020170100611BA (UBA). CHM is amember of the Carrera del Investigador Científico of the Consejo Nacional de In-vestigaciones Científicas y Técnicas (CONICET). Hinode is a Japanese missiondeveloped and launched by ISAS / JAXA, with NAOJ as domestic partner andNASA and STFC (UK) as international partners. It is operated by these agenciesin cooperation with ESA and NSC (Norway). This scientific work makes use ofthe Murchison Radio-astronomy Observatory (MRO), operated by the Common-wealth Scientific and Industrial Research Organisation (CSIRO). We acknowl-edge the Wajarri Yamatji people as the traditional owners of the Observatorysite. Support for the operation of the MWA is provided by the Australian Govern-ment’s National Collaborative Research Infrastructure Strategy (NCRIS), undera contract to Curtin University administered by Astronomy Australia Limited.We acknowledge the Pawsey Supercomputing Centre, which is supported by theWestern Australian and Australian Governments. The SDO data are courtesy ofNASA / SDO and the AIA, EVE, and HMI science teams. CHIANTI is a collab-orative project involving George Mason University, the University of Michigan
Article number, page 5 of 13 & Aproofs: manuscript no. ISSI_E2_Hinode-final (USA), University of Cambridge (UK) and NASA Goddard Space Flight Cen-ter (USA). The FIELDS instrument on the Parker Solar Probe spacecraft wasdesigned and developed under NASA contract NNN06AA01C. Contributionsfrom S.T.B. were supported by NASA Headquarters under the NASA Earth andSpace Science Fellowship Program Grant 80NSSC18K1201. RS acknowledgessupport from the Swiss National Science foundation (grant no. 200021_175832).Contribution from K.B. were supported by Swiss National Science Foundation.
References
Badman, S. T., Bale, S. D., Martínez Oliveros, J. C., et al. 2020, ApJS, 246, 23Bale, S. D., Badman, S. T., Bonnell, J. W., et al. 2019, Nature, 576, 237Bale, S. D., Goetz, K., Harvey, P. R., et al. 2016, Space Sci. Rev., 204, 49Brooks, D. H., Ugarte-Urra, I., & Warren, H. P. 2015, Nature Communications,6, 5947Brooks, D. H. & Warren, H. P. 2011, ApJ, 727, L13Brooks, D. H. & Warren, H. P. 2016, ApJ, 820, 63Brooks, D. H., Winebarger, A. R., Savage, S., et al. 2020, ApJ, 894, 144Culhane, J. L., Harra, L. K., James, A. M., et al. 2007, Sol. Phys., 243, 19Del Zanna, G. 2013, Astron. Astrophys., 555, A47Del Zanna, G., Aulanier, G., Klein, K. L., & Török, T. 2011, A&A, 526, A137Del Zanna, G., Dere, K. P., Young, P. R., Landi, E., & Mason, H. E. 2015, Astron.Astrophys., 582, A56Démoulin, P., Bagalá, L. G., Mandrini, C. H., Hénoux, J. C., & Rovira, M. G.1997, A&A, 325, 305Dere, K. P., Landi, E., Mason, H. E., Monsignori Fossi, B. C., & Young, P. R.1997, Astron. Astrophys. Supp. Ser., 125, 149Fox, N. J., Velli, M. C., Bale, S. D., et al. 2016, Space Sci. Rev., 204, 7Green, L. M., López fuentes, M. C., Mandrini, C. H., et al. 2002, Sol. Phys., 208,43Grevesse, N., Asplund, M., & Sauval, A. J. 2007, Space Sci. Rev., 130, 105Harra, L. K., Sakao, T., Mandrini, C. H., et al. 2008, ApJ, 676, L147Horbury, T. S., Woolley, T., Laker, R., et al. 2020, ApJS, 246, 45Kamio, S., Hara, H., Watanabe, T., Fredvik, T., & Hansteen, V. H. 2010,Sol. Phys., 266, 209Kashyap, V. & Drake, J. J. 1998, Astrophys. J., 503, 450Kashyap, V. & Drake, J. J. 2000, Bulletin of the Astronomical Society of India,28, 475Kasper, J. C., Bale, S. D., Belcher, J. W., et al. 2019, Nature, 576, 228Krucker, S., Kontar, E. P., Christe, S., Glesener, L., & Lin, R. P. 2011, ApJ, 742,82Krupar, V., Szabo, A., Maksimovic, M., et al. 2020, ApJS, 246, 57Leblanc, Y., Dulk, G. A., & Bougeret, J.-L. 1998, Sol. Phys., 183, 165Lonsdale, C. J., Cappallo, R. J., Morales, M. F., et al. 2009, IEEE Proceedings,97, 1497Mandrini, C. H., Baker, D., Démoulin, P., et al. 2015, ApJ, 809, 73Mariska, J. T. 1994, ApJ, 434, 756Mohan, A. & Oberoi, D. 2017, Sol. Phys., 292, 168Morioka, A., Miyoshi, Y., Iwai, K., et al. 2015, ApJ, 808, 191Mulay, S. M., Sharma, R., Valori, G., et al. 2019, A&A, 632, A108Pesnell, W. D., Thompson, B. J., & Chamberlin, P. C. 2012, Sol. Phys., 275, 3Pulupa, M., Bale, S. D., Badman, S. T., et al. 2020, ApJS, 246, 49Pulupa, M., Bale, S. D., Bonnell, J. W., et al. 2017, Journal of Geophysical Re-search (Space Physics), 122, 2836Reames, D. V. 2017, Solar Energetic Particles, Vol. 932Reames, D. V. 2018, Space. Sci. Rev., 214, 61Reid, H. A. S. & Ratcli ff e, H. 2014, Research in Astronomy and Astrophysics,14, 773Sterling, A. C. & Moore, R. L. 2020, ApJ, 896, L18Tingay, S. J., Goeke, R., Bowman, J. D., et al. 2013, PASA, 30, e007Ugarte-Urra, I. & Warren, H. P. 2011, ApJ, 730, 37Wayth, R. B., Tingay, S. J., Trott, C. M., et al. 2018, PASA, 35, 33Zlotnik, E. I. 1981, A&A, 101, 250 Article number, page 6 of 13. Harra et al.: The active region source of a type III radio storm observed by Parker Solar Probe during Encounter 2
Fig. 1.
PSP / RFS radio frequency data showing the type III bursts and storm on April 2, 2019. Top panel is the normalized Stokes intensity above 18MHz, 2nd panel is the full spectrogram, 3rd panel shows significant burst above background, and the bottom panel is the frequency of maximumnormalized intensity. This interval shows both more classical type III bursts, and the weaker more localized features associated with type III radiostorms. These data are described further in the text. Article number, page 7 of 13 & Aproofs: manuscript no. ISSI_E2_Hinode-final
Fig. 2.
PSP / RFS, RSTN, and MWA radio frequency measurements, AIA 193 Å intensity, and GOES X-ray data over the full interval March 30,2019 - April 6, 2019. The top panel shows daily RSTN flux from 30th March to 6th April 2019 at 245 MHz. The red circle marks the flux at240 MHz obtained from MWA flux calibration, with error bars from the temporal RMS each day. The second panel is GOES soft X-ray flux datawhich shows no significant flares. The 3rd panel shows the AIA 193 Å intensity integrated over the FOV shown in Figure 3, showing the growthand flattening o ff of the development of the AR from 31 March through 4th April. The 4th panel is again the normalized Stokes intensity above18 MHz and 5th panel is again the full spectrogram. The 6th panel is the frequency of peak normalized signal (as per Figure 1). The bottom panelis the inferred source height of the radio emission, inverting the peak frequency from a solar wind density profile and fundamental emission, asdescribed in the text. The noise storm commences on March 31 and a clear trend between April 1 to April 4 shows the peak type III storm emissionfrequency decreasing, corresponding to a higher source altitude or a more rarefied source region.Article number, page 8 of 13. Harra et al.: The active region source of a type III radio storm observed by Parker Solar Probe during Encounter 2 y [ a r cs e c ] −800 −750 −700 −650 −600 −550 y [ a r cs e c ] l og ( A I A [ DN / s ] ) Fig. 3.
The evolution of the active region from its emergence on the 31st March to 7 April 2019. The active region has a simple structure, withexpansion clearly seen, especially between the 31 March and 3 April.
Fig. 4.
AIA’s 171 Å solar image overlaid with the MWA radio contours at 4 frequency bands. The time corresponding to the AIA image is 04:01:45UT. The contour levels corresponding to 164, 192 and 240 MHz are 50%, 70%, 80% and 90% w.r.t map’s maximum T B respectively. The contourlevels corresponding to the 108 MHz radio source are 70%, 80%, 90%, 99%, w.r.t map’s maximum T B . Article number, page 9 of 13 & Aproofs: manuscript no. ISSI_E2_Hinode-final
Fig. 5.
The top two panels show the magnetic field model of the active region on 1 April and 4 April overlaid on AIA 193 Å images with theintensity reversed. The bottom two panels show the same model but overlaid on the Doppler velocity maps with blue showing blue shifts and redshowing red shifts (colour range is +/ -20km / s). The AR expands as it evolves. At its edges on 1 April, we observe structures in AIA data that areclosed within the AR. On 4 April the magnetic field lines derived from our model look more expanded at the edges of the AR, and larger scalemagnetic field lines are relevant in correspondence with the blue-shifted region. The axes in all panels are labelled in Mm, with the origin set inthe AR centre (located at N12E31 on 1 April and at N13W05 on 4 April). The iso-contours of the line-of-sight field correspond to ± ± ±
500 G in continuous magenta (blue) style for the positive (negative) values. Sets of computed field lines matching the global shape of theobserved coronal loops in the AIA 193 Å images have been added in continuous line and red colour in the two top panels and repeated in the twobottom panels.Article number, page 10 of 13. Harra et al.: The active region source of a type III radio storm observed by Parker Solar Probe during Encounter 2 y [ a r cs e c ] y [ a r cs e c ] y [ a r cs e c ] −671 −610 −546 −482 −421−8114237360482−671 −610 −546 −482 −421−8114237360482−671 −610 −546 −482 −421x [arcsec]−8114237360482 −83 −22 41 104 16543166289412534−83 −22 41 104 16543166289412534−83 −22 41 104 165x [arcsec]43166289412534 139 200 264 328 38937160283406528139 200 264 328 38937160283406528139 200 264 328 389x [arcsec]37160283406528 384 445 509 573 63456179302425547384 445 509 573 63456179302425547384 445 509 573 634x [arcsec]56179302425547 Fig. 6.
Top panel: EIS Fe xii xii & Aproofs: manuscript no. ISSI_E2_Hinode-final
Number of pixels with Doppler shifts < −10 km/s B l ue − s h i ft ed a r ea ( p i x e l s ) FIP bias F I P b i a s Mean Doppler velocity D opp l e r v e l ( k m / s ) Mean non−thermal velocity V n t ( k m / s ) Fig. 7.
Top left panel: number of pixels of the blue-shifted region as the active region crossed the disk from 1 – 6 April. The area of the blue-shiftedregion increases as the AR evolves. Top right panel: FIP bias measurements in the outflow area (white boxes in Figure 6). The FIP bias appears toincrease on April 4th when the blue-shifted area also reaches a maximum. Bottom left panel: Variation of the mean of Doppler velocity in the sameregion. Bottom right panel: Variation of mean of non-thermal velocity in the same region. Both the Doppler velocity and non-thermal velocityshow a significant increase in magnitude in the first few days of the active region’s formation.
Fig. 8.
Left panel: AIA 193 Å image with the field-of-view of the EIS fast scan overlaid. Within the EIS raster we chose a small region of interestfocused in the blue-shifted region. A blue arrow highlights this region of interest. Right panel, Top: variation of the Doppler velocity during theone hour raster. Bottom: variations in the PSP data and the frequency of the type III bursts for the same time interval.Article number, page 12 of 13. Harra et al.: The active region source of a type III radio storm observed by Parker Solar Probe during Encounter 2 A v e r age i n t en s i t y intensity > 100, and more than 15 pixels N u m be r o f b l ob s i n r d i ff i m age Fig. 9.
Top panel: lightcurve of the AIA intensity in the 193 AA filter. Bottom panel: the number of brightenings in the AIA running di ff erencemovie that have a ’blob’ size with an area more than 15 pixels and an intensity enhancement of >>