Evidence for precursors of the coronal hole jets in solar bright points

A set of 23 observations of coronal jet events that occurred in coronal bright points has been analyzed. The focus was on the temporal evolution of the mean brightness before and during coronal jet events. In the absolute majority of the cases either single or recurrent coronal jets were preceded by slight precursor disturbances observed in the mean intensity curves. The key conclusion is that we were able to detect quasi-periodical oscillations with characteristic periods from sub-minute up to 3-4 min values in the bright point brightness which precede the jets. Our basic claim is that along with the conventionally accepted scenario of bright point evolution through new magnetic flux emergence and its reconnection with the initial structure of the bright point and the coronal hole, certain MHD oscillatory and wave-like motions can be excited and these can take an important place in the observed dynamics. These quasi-oscillatory phenomena might play the role of links between different epochs of the coronal jet ignition and evolution. They can be an indication of the MHD wave excitation processes due to the system entropy variations, density variations or shear flows. It is very likely a sharp outflow velocity transverse gradients at the edges between the open and closed field line regions. We suppose that magnetic reconnections can be the source of MHD waves due to impulsive generation or rapid temperature variations, and shear flow driven nonmodel MHD wave evolution (self-heating and/or overreflection mechanisms).

Sample BP before and during CJ ejection process. The top three row represent overlap of SDO/AIA 193Å, 304Å and 171Å intensity images and SDO/HMI photospheric magnetograms. The bottom row represents HMI magnetogram separately. Blue and yellow contours indicate positive and negative magnetic field polarities, respectively. These regions are also indicated with arrows in the bottom row of panel (a). Panel (a) -the beginning of precursor; panel (b) -the peak of the precursor brightening; panel (c) -the time after the precursor when still there is no signature of main jet outflow; panel (d)the moment when the structure is destabilized and the jet-type instability starts; panel (e) -the fully developed transient jet outflow. In panel (e) we take a wider observational window as shown on the corresponding axes.

INTRODUCTION
Coronal Jets (CJs) are collimated plasma flows observed in extreme ultraviolet (EUV), X-Rays and white light images across the entire solar atmosphere: in quite corona and active regions Chandra et al. 2017;Guo et al. 2014;Raouafi et al. 2016) and in coronal holes (CHs) (Sterling et al. 2015;Young & Muglach 2014a,b;Young 2015). The contemporary space based observational missions (such as Hinode, Solar Dynamics Observatory, and others) with state-of-the-art spatiotemporal resolution revealed that jet-like transient processes occur very often in the solar atmosphere (Savcheva et al. 2007;Paraschiv et al. 2010).
Despite some observational and theoretical studies (Shibata et al. 1992;Yokoyama & Shibata 1995;Shimojo et al. 2001;Ryutova et al. 2008;Pariat et al. 2009;Moore et al. 2010;Magara 2010;Pariat et al. 2010;Archontis et al. 2010) the mechanism for CJ formation and evolution is not entirely understood. There is a general consensus that all of the jet-like features should originate from the common basic formation mechanism, viz. the release of free magnetic energy due to magnetic reconnection processes (Shibata et al. 2007;Pariat et al. 2009). The majority of the investigated CJs eject from newly emerging or developed bright points (BPs) (Nisticò et al. 2009) during the peak of their brightness (Pucci et al. 2012). In the past BPs visible in H-alpha and in EUV have been frequently observed before cool jets called surges and collimated with EUV jets (Gu et al. 1994;Canfield et al. 1996;Schmieder et al. 1996). Sterling et al. (2015Sterling et al. ( , 2016 presented new results that suggest that mini-filament eruptions trigger CJs, and there might be no difference in the scenario of standard and blowout jet formation. Furthermore, Panesar et al. (2016) presented an observational  Figure 2. Coronal Jet intensity curves for dataset 1 (without threshold -bottom curves) and dataset 2 with threshold -top curves) in top panels (CJ1) and (CJ2/CJ3/CJ4). Red (dataset 1) and blue (dataset 2) solid-line parts of the curves indicate precursors events. Accordingly, we plot zooms of these parts in the detrended form in bottom panels (CJ1), (CJ2), (CJ3), (CJ4), respectively. Panels (CJ1) (top and bottom) correspond to the case of the single CJ that started at 2015-12-09 17:28 UT with a precursor start approximately 9.8 minutes before the CJ release. While, panels (CJ2/CJ3/CJ4), (CJ2), (CJ3) and (CJ4) demonstrate the case of the recurrent CJs that started at 2015-12-30 23:16 UT, which includes three subsequent plasma ejections. Each of them has precursors starting 19.6, 7.8 and 5.6 minutes before the respective CJs. Vertical dashed lines represent the end of precursor ignition.
study that reports that mini-filament eruptions, which lead to CJ acceleration, are driven by flux cancellation in the photosphere. The matter of our focus, in this letter is a particular aspect of the observed dynamics, namely a systematic observational study of the CJs associated with on-disk CHs. Our aim was to deduce from the extreme EUV data analysis the character of the temporal behavior of the mean intensity and the connection to the jet outflow processes themselves.
In general, recent studies report on the BP brightness fluctuations and a variety of their oscillation periods (Sheeley & Golub 1979;Nolte et al. 1979;Strong et al. 1992;Mandrini et al. 1996;Aulanier et al. 2007;Tian et al. 2008;Pontin et al. 2011;Kumar et al. 2011;Guo et al. 2013;Schmieder et al. 2013). Besides, Pucci et al. (2012) followed the BPs brightness evolution for several CJs and found that most of the jets occurred in close temporal association with the brightness maxima. This observational evidence leads us to the assumption that in the dynamics of BPs, different MHD wavelike or oscillatory disturbances may take place. In this regard, Pariat et al. (2012) assumes that the observed structure of CJs could be explained using the generation of a helical kink-like wave. Lee et al. (2015) reports that jets manifest oscillatory motions during their ejection linking them to wave processes. Moreover, such disturbances may provide a link for the coupling of the dynamic processes occurring in BPs. It is natural to assume (in the current study only at an intuitive level) that these wavelike processes can be excited in the substantially non-equilibrium environment such as shear flows (Shergelashvili et al. 2006) or flows with thermal variations (Shergelashvili et al. 2007).
This kind of disturbances can be excited in the vicinity of and triggered by the above-mentioned, conventionally accepted coronal jet energy sources viz. magnetic reconnection and flux emergence/cancelation processes. For an analytical modeling under these frameworks, one should keep in mind that such wave-like dynamical processes follow more or less similar scenarios: ignition of a precursor disturbance, its evolution in time, and triggering of a main larger event. Besides, the overall contribution of CJ incidents in the energy budget of the solar atmosphere and wind can be estimated using the statistical properties of CHs reported by Bagashvili et al. (2017), with which the BPs of interest can be linked. However, before achieving such far reaching consequences regarding analytical modeling, a thorough observational analysis and confirmation of the regular presence of such quasi-oscillatory type of precursors are needed. These are the ultimate goals of the observational studies reported in the current letter.

THE METHODOLOGY OF OBSERVATIONS AND DATA ANALYSIS
We used data from the Atmospheric Imaging Assembly (AIA) on board of the Solar Dynamic Observatory (SDO) , which monitors the Sun with 0.6" spatial and 12 sec temporal resolution (Boerner et al. 2012). The SDO/AIA data are retrieved, processed and analyzed using standard procedures with the SolarSoft (SSW) package. Our particular interest goes to BPs and CJs situated within and at the edges of the CHs. The coronal holes have been chosen from different areas of both hemispheres, excluding regions very close to limbs, during the period from 1 December 2015 till 1 May 2016. In total, we investigated 23 (spatiotemporarily independent) CJs using SDO/AIA 193 A channel images. Among them were recurrent jets which, nonetheless, we treated each jet phenomenon as a separate event.
For despiking, bad pixel/cosmic-ray influence correction and flat-fielding, level 1 data has been processed into level 1.5 using the aia prep.pro code. From the obtained files, we cut out rectangular boxes for each event and create their time series. We applied the rot xy procedure to exclude the solar rotation effect. Figure 1 shows several example snapshots from such an image sequence. The images are created using the AIA 193Å,304Å and 171Å channel data both in combination with HMI photospheric magnetograms. Blue and yellow contours indicate positive and negative magnetic field polarities, respectively. In the middle (small yellow and blue closed contours) of the BP there is a small bipolar closed magnetic structure anchored in the photosphere and extended in the form of a magnetic loop through the base of the solar corona. Besides, it is seen that the footpoints of the central loop are surrounded by a system of magnetic field concentration areas (edge of supergranules) in the chromospheric network. These magnetic flux concentration regions are closely situated to each other and also embedded in the global CH magnetic field of positive polarity (Fig. 1). Panel (a) shows the beginning of the precursor for each wavelength and for the magnetogram as well; panel (b) corresponds to the peak of the precursor brightening; panel (c) corresponds to the time after the precursor when there is still no signature of the main jet outflow; panel (d) demonstrates the moment when the structure is destabilized and the jet-type instability starts and filamentary darkening aside the BP loop appears, just like in many similar observations; panel (e) corresponds to the fully developed transient jet outflow.
In order to perform a rigorous analysis the observed dynamical processes we created two types of 193Å intensity curves in Fig. 2. The first type of data (dataset 1) comprises the calculation of the mean intensity values of the entire cutout BP-boxes. The second type of data (dataset 2) is created using the average over all pixels of modified intensity values cutout BP-boxes obtained through noise deduction (using the methodology similar to one outlined in Chandrashekhar et al. (2013)). The final values in dataset 2 exclude low-intensity noise containing regions of BPboxes. Consequently, in dataset 2, the effect of background noise is removed and all the transient disturbances are more sharply observable. All 23 BPs and corresponding jets we investigated have more or less similar morphological and dynamical properties. Examples of brightness evolution curves are shown in top panels (CJ1) and (CJ2/CJ3/CJ4) of Fig. 2.
Despite the fact that the examples of the BPs we studied show the structure and dynamics conventionally standardized in the related literature (Raouafi et al. 2016;Sterling et al. 2015). Our study uncovers the systematic presence of relatively low amplitude (compared to the jets) quasi-oscillatory dynamic processes before each main jet events (the solid-line, black parts of the curves in top panels (CJ1) and (CJ2/CJ3/CJ4) of Fig. 2). We argue that these processes can represent precursors of jets (the solid-line, red and blue colored parts of the curves in top panels (CJ1) and (CJ2/CJ3/CJ4) of Fig. 2). Further, we plot these precursor parts zoomed in and detrended for an isolated single (top panel (CJ1)) and for a recurrent jet event (panels (CJ2), (CJ3), (CJ4)). It is apparent that regardless of the type of the event (single or recurrent) the quasi-oscillatory variation of the mean intensity of precursors is systematically observed.

ANALYSIS OF THE RESULTS
Our first finding is that in the absolute majority of instances (20 out of 23) we detect a characteristic brightening of the observed BP preceding, by a few minutes, the main jet outflows. We consider these processes as CJ precursors (as shown in Fig. 2). In this paper we show only illustrative examples of single and recurrent jet events and the full analysis all 23 jets will be given in the form of http://solspanet.eu/solspanet/static/Supplementary_Data_Bagashvili_1001.pdf on our project web site. Figure 2 shows only illustrative examples of single (top panel (CJ1)) and recurrent (panel (CJ2/CJ3/CJ4)) jet events (out of all 23 investigated cases). The enhancement of the BP intensity is visible several minutes before the CJ ejection in the case of both dataset 1 and dataset 2 mean intensity curves. In both cases of single and recurrent jets the precursors are systematically observed before each plasma ejection. In top panel (CJ1) Fig. 2  Note-Complete set of 23 selected CJ and corresponding precursor parameters. Location represents BP (x, y) coordinates in arcsecs; in the column Type we have two possible values S -single and R -recurrent; τP I represents the precursor ignition duration; τP T is the precursor total evolution time span; (τCJ ) -CJ durations, ∆τ P eaks -The time intervals between precursor and CJ peaks; Osc. Periods shows the minimum and maximum values of precursor oscillation periods. The parameters having time dimension are measured in minutes.   Table  1.
We investigated the statistical distributions of the precursor and the CJ individual parameters for both dataset 1 and dataset 2. We organized the parameters as follows (Table 2): (i) The precursor ignition duration (τ P I ), representing the half-width of the Gaussian fit to the first disturbance in the precursor. The τ P I contains the information about the temporal properties of the precursor sources which enables us to judge the particular periods of the initially ignited disturbances. It should be noticed that some precursors consist of a single peak disturbance and in such cases the ignition and total duration of the precursors are co-measurable. In other cases, however, the precursors include an Note-The CJ and corresponding precursor parameters with related variance and error estimations. τP I represents the precursor ignition duration; τP T is the precursor total evolution time span; (τCJ ) -CJ durations, ∆τ P eaks -The time intervals between precursor and CJ peaks. All the quantities in the top four rows are measured in minutes and in bottom four rows all are dimensionless. All values are given in accordance with the order data set 1/data set 2. In two bottom row values are calculated only for dataset 2. oscillatory-like behavior, and in this case, the total length of the process is significantly larger than the ignition (see Table 1). (ii) The precursor total evolution time span (τ P T ) showing how long the precursor process lasts before the jet outflow starts. There is another parameter also enabling to judge on time spans between the precursors and the CJs, viz. (iii) The time intervals between precursor and CJ peaks (∆τ P eaks ). (iv) CJ durations (τ CJ ) calculated in a similar way as the precursor ignition, as described above. We also evaluated the interrelation between the precursors and coronal jets by introducing the following dimensionless parameters: (i) the ratio of the coronal jet and precursor ignition peak intensities (I CJ /I P I ) (the parameter manifests the rate of the coupling and the energy pumping from the precursor, and maybe some external source, to the main jet) and (ii) their durations (τ CJ /τ P I ). Finally, the ratio of the temporal gap between the precursor ignition and the jet peaks over (iii) the precursor ignition (∆τ P eaks /τ P I ) and (iv) jet (∆τ P eaks /τ CJ ) durations.

DISCUSSION AND CONCLUSIONS
We performed a statistical analysis of the brightness evolution of 23 small-scale jet-like events located within on-disk CHs. We studied the corresponding bright point brightness average intensity evolution in time, using high temporal and spatial resolution SDO/AIA 193Å images. We found that the vast majority of coronal jets (CJs) are accompanied by a minor increase in mean intensity occurring several minutes before each intensive jet ejection. The precursor was identified in 20 out of 23 cases. We consider such enhancements of brightness as precursors of CJs. This result allows us to conclude that the presence of the precursor is a systematic property of CJs and the absence of the precursor in 3 of the 23 considered cases might be due to either incompleteness of the observational data. Alternatively, this can also be caused by the overlap between the precursor (τ P I ) and the jet (τ CJ ) duration intervals, which can happen when both are larger than the characteristic time between the peaks (∆τ P eaks ) (zero values in corresponding fields). Despite the fact that there are some studies of the BP brightness evolution, for instance by Pucci et al. (2012), we performed introduction of the notion of CJ precursors.
According to our observation, the average lifetime of all examined τ P I is 2.77/2.85 min, the mean ratio of CJ and precursor ignition peak intensities (I CJ /I P ) is 2.18/2.59, the average time between CJ and its precursor (∆τ P eaks ) is 10.25/10.13 min, while the duration of the CJs (τ CJ ) is 8.52/8.62 min (see Table 2) which is in good agreement with other observational indications (see e.g. Schmieder et al. 2013, and references therein). These parameters allow drawing preliminary conclusions on the observed processes. (i) We made an estimation of the mean maximum oscillation period by introducing the parameter τ P I and also trough the FFT analysis. As we see from the table both estimations coincide very well. (ii) The relation between the total (τ P T ) and ignition (τ P I ) durations of the precursor shows that the longest period oscillation makes on average 2-3 oscillations before the jet starts which is also justified by the mean values of ratio ∆τ P eaks /τ P I and τ CJ /τ P I . (iii) The value of ∆τ P eaks /τ CJ demonstrates the fact that the mean period of the mode that is involved in the jet outflow must be larger than that of the one involved in the precursor. (iv) The ratio of I CJ /I P proves that the amplitude of the jet mode is larger than that of the precursor which indicates that perhaps the presence of the wave mode nonequilibrium driving (entropy variations or shear flows). This kind of parameter analysis can be continued. However, we stop at this moment as further analysis ultimately requires mathematically rigorous modeling which is planned in the near future (see corresponding remarks below).
The key conclusion that can be drawn from our analysis is that we were able to detect quasi-periodical oscillations with characteristic periods from sub-minute up to 3-4 min values in the BP brightness which precede the jets. The AIA cadence favors such detection. The basic claim what can be made at this stage of pure observational analysis is that along with the conventionally accepted scenario of BP evolution through new magnetic flux emergence and its reconnection with the initial structure of the BP and the CH, certain MHD oscillatory and wave-like motions can be excited and these can take an important place in the observed dynamics. One can even imagine that these quasi-oscillatory phenomena might play the role of links between different epoches of the CJ ignition and evolution. However, we do not have at this moment rigorous information neither on the nature of this quasi-oscillatory behavior nor on the mentioned possible links with the standard evolutionary scenario of the BPs and CJs. We just give a rough estimation of characteristic periods by applying a standard FFT routine to the data. A complete understanding of this issues requires further analytical and perhaps even numerical modeling of the processes discovered here and such investigations will become a matter of future more extensive studies. These are not compatible with the format of the present short letter, which only aims at the announcement of the novel observational evidence. In other words, the goal is to establish a notion of the CJ and their precursor patterns. However, we can make some general qualitative indications on the observed oscillatory processes and their link with some theoretical background. The quasi-oscillatory variations of intensity can be an indication of the MHD wave excitation processes due to the system entropy variations (Shergelashvili et al. 2007) or density variations (Shergelashvili et al. 2005;Zaqarashvili & Roberts 2002). The observed mutual positioning of open and closed magnetic field structures, indicates that there is very likely a sharp outflow velocity gradients at the edges between the open and closed field line regions. All these conditions suggest a sequence of local magnetic reconnection events that could be the source of MHD waves due to impulsive generation or rapid temperature variations (Shergelashvili et al. 2007) on the one hand, and shear flow driven MHD wave excitation, coupling and dissipation (self-heating mechanism) processes (Shergelashvili et al. 2006) and explosive type strongly non-adiabatic overreflections in the shear flows (Gogoberidze et al. 2004), on the other hand. The results obtained in this letter create a solid ground for further studies in this direction.