A Search for Fast X-ray Variability from Active Galactic Nuclei using Swift 5 Matthew Pryal, Abe Falcone, Michael Stroh 1 0 Department of Astronomy & Astrophysics 2 The Pennsylvania State University, University Park, PA 16802 n a J 8 2 Abstract Blazars are a class of active galactic nuclei (AGNs) known for their very rapid variabilty in ] E the high energy regions of the electromagnetic spectrum. Despite this known fast variability, H X-ray observations have generally not revealed variability in blazars with rate doubling or . halving timescales less than approximately 15 min. Since its launch, the Swift X-ray Tele- h scope has obtained 0.2-10 keV X-ray data on 143 AGNs, including blazars, through intense p target of opportunity observations that can be analyzed in a multiwavelength context and - o used to model jet parameters, particularly during flare states. We have analyzed this broad r Swift data set in a search for short timescale variability in blazars that could limit the size t s of the emission region in the blazar jet. While we do find several low-significance possible a [ flares with potential indications of rapid variability, we find no strong evidence for rapid (<15 minutes)doublingorhalvingtimesinflaresinthesoftX-rayenergybandfortheAGNsanalyzed. 1 v 8 Subject headings: acceleration of particles - galaxies: active - galaxies: jets - gamma rays: general - 8 X-rays: general 2 7 1. Introduction R < (cTδ)/(1 + z) where T is the variability 0 . timescale of the flux, δ is the Doppler factor, and 1 Analysisofthefluxvariabilityofactivegalactic z is cosmological redshift. The Doppler factor δ 0 5 nuclei(AGNs) allowsconstraintsto be put onthe is defined as δ−1 = γ(1−βcosθ), where γ is the associated emission mechanisms and jet parame- 1 Lorentz factor in the jet, β is the ratio of the jet : ters of the AGN jet. Flux variability in AGNs velocity to the speed of light in vacuum, and θ is v is seenacrossthe entireelectromagneticspectrum i the angle of the jet with respect to the observer. X with the fastest variability seen in the higher en- Thevariabilitytimescale,T,istypicallydescribed r ergyregions,inparticularthekeVtoTeVregions. by the flux doubling time, which will be referred a Of all AGNs, the strongest and fastest variability to as T or the flux halving time, which will double is often seenin a class of AGNs knownas blazars, be referred to as T1/2. whichareAGNsthathaveajetpointedinadirec- There has been repeated evidence of fast flux tionthatisveryclosetoourlineofsight. Analysis variability in the TeV γ-ray region with doubling ofthefluxvariabilityofblazarsinthehigh-energy or halving times as short as 224±60 s from PKS regionof the electromagnetic spectrum is a useful 2155-304 (Aharonian et al., 2007), 15 minutes methodforlimitingthesizeoftheemissionregion from Mrk 421 (Gaidos et al., 1996) and ∼ 9 min in an AGN, thereby limiting the emission mecha- from BL Lac (Arlen et al., 2012). Results from nisms. simultaneousTeVandX-raystudiesprobingvari- Due to light propagation time and geometry ability on the order of hours have shown evidence effects, fast variability can limit the size of the that X-rays and TeV γ-rays may originate from emission region R of a blazar by the inequality the same region (e.g. Maraschi et al. 1999). This 1 correlationat largertimescales, in addition to ob- The data are comprised of 7.43×106 seconds served fast variability of TeV γ-rays, has moti- of AGN observations from 12544 continuous ob- vated the search for fast variability in blazars at servations of the 143 AGN observed at the time X-ray wavelengths. of analysis. These data were obtained between In general, X-ray observations have not re- 2004 December 17 and 2014 May 28. The data vealed variability with doubling timescales less were processed using the most up to date version than approximately 15 min. There has been iso- of Swift tools at the time of analysis: Swift Soft- latedevidenceformuchfastervariabilityinX-rays wareversion4.1andFTOOLSversion6.14(Black- from a single blazar,H0323+022(Feigelsonet al., burn,1995)withobservationsprocessedusingxrt- 1986), with a halving time less than 30 seconds pipeline version 0.12.8. For WT mode data, we anda quotedstatisticalsignificanceofP <10−13. additionally filtered out the first 150 seconds of However, this fast variability in X-rays was a sin- data after a Swift telescope slew since the tele- gle, isolated event that has not been seen again. scopepointing wasnotalwayssettledwellenough The authors did consider systematic and/or in- to allow accurate standard WT mode data anal- strumental effects, but none could be identified. ysis. Greater detail on how these data were ob- tained and processed is available in Stroh & Fal- With the launch of the Swift X-Ray Telescope cone (2013). (Burrows et al., 2005), X-ray data in the 0.2-10 keV region has been obtained on many AGNs on The number of observations for each object timescales ranging from seconds to over 8 years. range from as many as 1192 observations in the caseofMrk 421to aslittle as1 observationin the We report on our search for fast variability in case of PKS 0244-470. The average continuous blazar flares. We have searched our large Swift observation time is about 592 s, with a maximum database for blazar flares with doubling or halv- observation time of 2565 s and a minimum obser- ing times less than 15 minutes. The data taken vation time of 21 s. from the Swift database for analysis are outlined in Section 2. The analysis techniques, including 3. Analysis the necessary constraints set on the data in the searchforfastblazarvariability,isoutlinedinSec- The goal of the analysis is to determine if fast tion 3. We provide the results of the analysis in blazar variability with a rate doubling or halv- Section 4 with discussion of the results in Section ing time less than 15 minutes exists within these 5. data. The data were analyzed on an observation byobservationbasistotestforthisfastvariability. 2. Data Variousconstraintswere neededto: (1)define the data set of observations that would have enough ThesedatawereobtainedwiththeX-RayTele- counts and a long enough duration to provide the scope (XRT) on the Swift observatory (Burrows potential to find significant flaring, and (2) define et al. 2005), and the reduced light curves are all the observations within the aforementioned data available on a public web site and database de- set that would have variability, and in particular, scribed by Stroh & Falcone (2013). The Swift- the potential for flux doubling or halving. The XRTisoneofthreeinstrumentsonboardtheSwift necessary constraints to define our data set in- GammaRayBurstExplorer(Gehrelsetal.,2004). clude: amaximumcountrateofatleast0.1cts/s, In particular, the Swift-XRT is an X-ray imag- atleastthreedatapointsineachobservation,and ing spectrometer with the ability to obtain data a length of observationof at least 5 minutes. The for lightcurves with high timing resolution. The constraintsthatdefineapossibleflareinclude: re- Swift-XRT data in this study are from two differ- strictingthemaximumcountratetobeatleast1.8 ent modes, window timing mode (WT) and pho- timesgreaterthantheminimumcountrateduring ton counting mode (PC). All data are screenedto the observation, 3 consecutive data points either ensurethat no individualobservations,andthere- increasing or decreasing in count rate, and a re- fore no potential flares, include a switch from one ducedχ2 valuefromastraightlinefitgreaterthan modetotheother. SeeBurrowsetal.,2005foran 3.0. In addition to these pre-defined constraints, in depth discussion of the XRT modes. 2 0954+658, Obs ID: 00035381001 1ES 2344+514, Obs ID: 00035031050 3C 279, Obs ID: 00035019131 -1Rate (0.3-10.0 keV) (cts s)000000......112112050050 -1Rate (0.3-10.0 keV) (cts s)000000......345345 -1Rate (0.3-10.0 keV) (cts s)0111101111..........8024680246 Count 00..0055 Count 00..22 Count 0000....4646 5533992200..444400 5533992200..444455 5533992200..445500 5533992200..445555 5555119999..441155 5555119999..442200 5555119999..442255 5555119999..443300 5566665577..447755 5566665577..448800 5566665577..448855 5566665577..449900 Time (MJD) Time (MJD) Time (MJD) 3C 279, Obs ID: 00053516001 3C 345, Obs ID: 00036390005 3C 454.3, Obs ID: 00031493027 -1Rate (0.3-10.0 keV) (cts s)00000000........45674567 -1Rate (0.3-10.0 keV) (cts s)0000....12125050 -1Rate (0.3-10.0 keV) (cts s)00000000........45674567 Count 00..33 Count 00..1100 Count 0000....2323 5533992266..220000 5533992266..220055 5533992266..221100 5533992266..221155 5555006666..880055 5555006666..881100 5555006666..881155 5555006666..882200 5566556677..777755 5566556677..778800 5566556677..778855 5566556677..779900 Time (MJD) Time (MJD) Time (MJD) PKS 0235+164, Obs ID: 00030880008 PKS 0537-441, Obs ID: 00030138004 PKS 0537-441, Obs ID: 00050150003 -1Rate (0.3-10.0 keV) (cts s)000000......234234 -1Rate (0.3-10.0 keV) (cts s)0000000000..........12233122335050550505 -1Rate (0.3-10.0 keV) (cts s)00000000........34563456 Count 00..11 Count 00..1100 Count 00..22 5544113377..228855 5544113377..229900 5544113377..229955 5544113377..330000 5544774477..886600 5544774477..886655 5544774477..887700 5544774477..887755 5533339966..660000 5533339966..660055 5533339966..661100 5533339966..661155 Time (MJD) Time (MJD) Time (MJD) PKS 1424+240, Obs ID: 00041539012 PKS 1510-089, Obs ID: 00031173069 RX J0324.6+3410, Obs ID: 00036533033 -1Rate (0.3-10.0 keV) (cts s)0000000000..........12233122335050550505 -1Rate (0.3-10.0 keV) (cts s)00000000........23452345 -1Rate (0.3-10.0 keV) (cts s)0000000000..........4567845678 Count 00..1100 Count 00..11 Count 0000....2323 5566339955..331100 5566339955..331155 5566339955..332200 5566339955..332255 5555667799..885555 5555667799..886600 5555667799..886655 5555667799..887700 5566330077..885555 5566330077..886600 5566330077..886655 5566330077..887700 Time (MJD) Time (MJD) Time (MJD) S5 0716+714, Obs ID: 00035009047 S5 0716+714, Obs ID: 00035009048 W Com, Obs ID: 00035018040 -1Count Rate (0.3-10.0 keV) (cts s)00010001........46804680 -1Count Rate (0.3-10.0 keV) (cts s)00000000000000..............34567893456789 -1Count Rate (0.3-10.0 keV) (cts s)000000000000............000111000111468024468024 5555117799..770000 5555117799..770055 5555117799..771100 5555117799..771155 5555118800..229955 5555118800..330000 5555118800..330055 5555118800..331100 5555335566..773355 5555335566..774400 5555335566..774455 5555335566..775500 Time (MJD) Time (MJD) Time (MJD) Fig. 1.— The light curves of the 15 observations flagged for a potential quick time X-ray flare are shown. The observation ID specified in the title of each plot is the observation ID assigned by Swift. 3 manual inspection was required for the observa- flux doubling or halving since the purpose of the tions that passed the constraints to ensure that analysisis to searchfor AGN flareswith fast dou- flux variations from systematic effects (e.g. mode bling or halving timescales. switching)werenotflaggedaspossibleflares. The necessity of each of these constraints is discussed 3.2.2. Consecutive Points in more detail below. We require at least 3 consecutive data points to be either increasing or decreasing in count rate 3.1. Constraints That Define Data Set to minimize random noise in the count rate and 3.1.1. Max Count Rate guarantee a possible steady increase or decrease in the flare. Amaximumcountrateofatleast0.1cts/sdur- ing an observation is necessary to allow for a sta- 3.2.3. Reduced χ2 Value tisticallysignificantflaretooccurinthetimescales being probed. Taking into account light curve Finally, a reduced χ2 value from a straightline error bars at this count rate and applying χ2 fit of the weighted mean of the count rate greater statistics,astatisticallysignificantflarecannotbe than 3.0 is necessary to ensure sufficient variabil- detected at rates lower than 0.1 cts/s over the ityinthecountrateofthesource. Thisconstraint timescales of concern. Therefore, if the maximum ensures that the selected observation is not con- count rate of an observation is not at least 0.1 sistent with a quiescent state of the AGN. This cts/s then it would be impossible to detect a fast reducedχ2 valuewasalsousedindeterminingthe flarefromthatobservation,thusweeliminatesuch p-valueofthenullhypothesisofa straightline fit. observations from the data sample. 3.3. Manual Analysis 3.1.2. Data Point Requirement Manual inspection was required to ensure that all observations flagged as possible flares did not The requirement of at least three data points haveanyissuesthatcouldnotbeflaggedsystemat- within an observation as a means of defining our ically. One of these issues includes the Swift-XRT data setis closelyrelatedto the discussioninSec- mode switching from PC to WT or vice versa, tion 3.2.2. We cannot include observations that which often caused an abrupt break in the light have less than three data points as part of our curve and spurious points to occur in the data as data set because they do not have the ability to shown in light curve on the top of Figure 2. This have three increasing or decreasing data points issue would occasionally cause observations to be and therefore would not possibly be able to be flagged as a possible flare and would have to be flagged as a possible flare. discardedmanually. Otherobservationsthatwere 3.1.3. Orbit Length processed out through manual inspection had a singlespuriousdatapointresponsiblefortheiden- We also require observations to be at least 5 tification of the observation as a potential flare, minutes in length. This is to allow enough time with the single point occurring either at the be- for a significantflare to occur within a typical ob- ginning or end of an observation with no flaring servation. Our data set also has max observation evidence throughout the rest of the light curve. time of 2565seconds, which implicitly sets an up- These observations (e.g. Figure 2, bottom) also per limit to the length of the doubling or halving had to be discarded from the data set. time observed of a potential flare, though no ex- plicit cut was set. 4. Results 3.2. Constraints That Define Flaring Of the initial 12544 AGN observations in the large data set, 8606 AGNs observations had the 3.2.1. Doubling/Halving Factor possibility of having a detectable quick time flare We require the maximum count rate to be 1.8 basedon the count-rateconstraints andthe a pri- times greaterthanthe minimumcountratetoen- ori assumptions that define our data set as dis- sure that any potential flares at least approach a cussed in Section 3.1. Of the remaining 8606 4 Mrk 421, Obs ID: 00030352205 saic of Figure 1. 7700 -1eV) (cts s) 6600 imcaeWnthceoeduosfceodtmhtepwaqoreusmictkehthetiomldigsehttsoocfutersvtXiem-rdaaatyetafltahtoreesas;ingonanisfe-- 0.0 k 5500 sumption that the X-ray rate was constant with 1 3- a value equal to the mean rate during the obser- 0. ate ( 4400 vation, and a second method compares the light R nt curve data to an assumption that the X-ray rate u Co 3300 was constant with a value equal to a best guess 5555224422..001100 5555224422..001155 5555224422..002200 5555224422..002255 at the quiescent source rate near the time of the Time (MJD) potential flare. In the first method, the chance probability p-value was determined based on the 1156+295, Obs ID: 00036381002 null hypothesis probability of a straight line fit of -1V) (cts s) 0000....11114646 tthhaetwweiagshtfeodunmdeaansadnisdcuusssinedgtihneSreecdtuiocned3χ.22.3v.aluIne e 0 k 00..1122 the secondmethod the chance probabilityp-value 3-10. 00..1100 wasdeterminedbasedonthenullhypothesisprob- e (0. 00..0088 ability of the light curve data being described by unt Rat 00..0066 asucroronusntadnint,gsetaracihghpto-ltiennetifiatl fltoarteh,ewmhiicnhimwuams ursaetde Co 00..0044 as a best-guess estimate of the ”quiescent” state 5544442288..6655 5544442288..6655 5544442288..6666 5544442288..6677 Time (MJD) of emission surrounding the potential flare. We computed the reduced χ2 value of the data dur- ing the potential flare relative to this ”quiescent” Fig. 2.— The light curve from Mrk 421 (top) constantratelightcurveandthen determinedthe showsasharpbreakinthelightcurvethatresults p-value. These values are shown in Table 1 un- in spurious data points as the result of the Swift- der ”Weighted Mean p-value” and ”Quiescent p- XRTswitchingobservingmodesfromPCmodeto value” respectively. WT mode. Thelightcurveof1156+295(bottom) Oncethep-valuewasdeterminedusingthetwo shows a single spurious data point that was po- different methods, we had to take into account a tentially a result of fluctuations at the beginning trials factor based on the number of observations of an observation that could not be considered a inthedatasetthathadthe potentialtoprovidea potential quick time flare. detection of a rapid soft X-ray flare. The number for the trials factor was determined by counting AGN observations, 31 observations passed the the number of observations that passed the set constraints that defined a possible flare as dis- constraints as discussed in Section 3.2 and Sec- cussed in Section 3.2. Once the manual analy- tion 3.3, as wellas the visualinspection described sis was done, 21 observations remained that did above. Therefore, there were 16 trials that were not have any mode switching issues or were re- taken into account, and the resultant p-values for liant on a single spurious data point, as discussed each method are shown in Table 1 in the ”Post- in Section 3.3. Of the remaining 21 observations, trials” column next to each respective method. six were flagged as being clearly due to random InTable1wealsoincludethereducedχ2values fluctuations, without any clearly discernible flare, thatwerefoundonthesectionofeachobservation and were therefore not considered to be poten- that was deemed part of the potential quick time tialflares,andoneobservationfromPKS0537-441 flareasdiscussedearlierfortheweightedmeanand showedsignsoftwoflaringepisodeswithinasingle quiescent straight line fits. We also include the observation. Therefore,16potentialflaresfrom15 doublingtimesandhalvingtimeswhenapplicable observationsand 12 different AGN remained. De- to each flaggedobservation. These times were de- tails from these 16 potential flares are shown in termined by fitting a straight line to the sections Table 1, and each light curve is shown in the mo- of each observation deemed to be part of a pos- 5 RX J0324.6+3410, Obs ID: 00036533033 find potential doubling times in 6 different AGNs -1V) (cts s) 0000....7878 bt9i.em1twemseieninn. A120.l2ldomifffitenhreeanpntodtAe5Gn.7tNiamslflibnaertawesseiewnneTl2la.2baslemh1ianclvoaimnndge e 0 k 00..66 from the blazar class (FSRQs or BL Lacs), which 0. 3-1 00..55 aremorelikelytoexhibitrapidflaringduetotheir 0. e ( 00..44 aligned relativistic jets. However, none of these at R potential flares represents a truly significant de- nt 00..33 ou tection after accounting for trials, and therefore, C 00..22 no strongstatements canbe made aboutthe exis- 5566330077..885555 5566330077..886600 5566330077..886655 5566330077..887700 Time (MJD) tence, or nonexistence, of short timescale flares in the data set. 1ES 2344+514, Obs ID: 00035031050 The potential flare from the RX J0324.6+3410 -1V) (cts s) 00..55 o(tbospe)r.vatTiohnisipsostheonwtinalinflamreohreasdeatadioluibnliFngigutirmee3, e 0 k 00..44 calculated as discussed in Section 4, of 5.7 min- 10. utes. If the variability is truly from flaring, the 3- e (0. 00..33 potential emission region size can be calculated Rat by applying the inequality R < (cTδ)/(1 + z), unt 00..22 which was discussed in Section 1. The doubling o C time variabilitiy as well as z=0.061(Linford et al. 5555119999..441155 5555119999..442200 5555119999..442255 5555119999..443300 Time (MJD) 2012)equatetoanupperlimitonthe emissionre- gionsizeof9.7×1012δ cm,whereδ istheunknown doppler factor that was discussed in Section 1. Fig. 3.— These plots show the selected section of If the potential flares outlined in Table 1 were the observations from RX J0324.6+3410 on 2013 indeed quick time soft X-ray flares, these would Jan 15 (top) and 1ES 2344+514on 2012 January be the fastest doubling and halving times for X- 3 (bottom) that were fit with either increasing or ray AGN flares ever observed, with the exception decreasing lines to estimate the doubling or halv- of the single, isolated event observed by Feigelson ing time. et al. (1986). As stated previously, variability in AGNs at timescales approximately this short had sible flare and calculating how much time would previously only been seen repeatedly with statis- be required for half of the maximum count rate tical significance in the TeV region (Aharonian et to double or for the maximumcount rateto halve al., 2007;Gaidos et al., 1996,Albert et al., 2007). when appropriate. An example of this fitting can Simultaneousmultiwavelengthstudiesofblazars be seen in Figure 3. We also include the date of in TeV γ-rays and X-rays are vital to determine observation of each potential flare as well as the if very fast X-ray variability is coincident with starttime ofthe observationcontainingthe possi- very fast TeV variaiblity as has been seen previ- ble flareas a reference to comparethe data to the ouslyonlongertimescales(Maraschietal., 1999). light curves available on the Swift-XRT monitor- Further X-ray observations of these sources with ing site (Stroh & Falcone, 2013). more sensitive instruments for longer continuous integrations to determine flare timescales will al- 5. Discussion low us to perform better searches for fast X-ray variability in the future. This will enable more Analysis of our large Swift-XRT database in- complete studies of the connection between X-ray dicates that our data set does not show strongly and TeV emission regions in blazar jets. significant evidence for soft X-ray flaring on the timescales probed, i.e. less than 15 minutes. Of the low-significance potential flares that were found and outlined in Table 1 and Figure 1, we 6 REFERENCES 2005, Sp. Sci Rev, 120, 165 Feigelson, E. D., et al., 1986,ApJ, 302, 337F Aharonian, F., et al., 2007,ApJ, 664L, 71A Gaidos, J. A., et al., 1996,Nature, 383, 319G Albert, J., et al., 2007, ApJ, 669, 862A Gehrels, N., et al., 2004, ApJ, 611, 1005 Arlen, T., et al., 2013,ApJ, 762, 92A Linford, J. D., et al., 2012, ApJ, 744, 177L Blackburn,J.K.,1995,inAstronomicalSocietyof Maraschi, L., et al., 1999, ApJ, 526L, 81M the Pacific Conference Series, Vol. 77, Astro- nomical Data Analysis Software and Systems Stroh, M. & Falcone, A., 2013, ApJS, 207, 28S IV,R.A.Shaw,H.E.Payne,&J.J.E.Hayes, ed., pp. 367 This2-columnpreprintwaspreparedwiththeAASLATEX macrosv5.2. Burrows, D.N., Hill, J.E., Nousek, J.A., et al., 7 Table 1: The 16 potential quick time soft X-ray flares flagged for more analysis. Source Date Obs. Time Weighted Quiescent Tdouble T1/2 Weighted Post-Trials Quiescent Post-Trials Mean χ2 Mean p-value red χ2 p-value red TXS 0954+658 04-Jul-06 53920.442 3.23 4.38 - 9.1 3.96×10−2 6.34×10−1 1.25×10−2 1.99×10−1 1ES 2344+514 03-Jan-10 55199.419 4.26 6.70 - 2.8 5.13×10−3 8.20×10−2 1.61×10−4 2.58×10−3 3C 279 31-Dec-13 56657.480 4.45 11.96 3.4 - 3.95×10−3 6.32×10−2 7.91×10−8 1.27×10−6 3C 279 10-Jul-06 53926.208 3.05 4.69 - 3.9 2.73×10−2 4.36×10−1 2.81×10−3 4.49×10−2 3C 345 23-Aug-09 55066.809 3.19 4.29 - 7.0 4.10×10−2 6.56×10−1 1.37×10−2 2.19×10−1 3C 454.3 02-Oct-13 56567.783 3.63 11.52 4.2 - 2.66×10−2 4.26×10−1 9.90×10−6 1.58×10−4 PKS 0235+164 06-Feb-07 54137.292 5.95 7.79 2.5 - 2.60×10−3 4.16×10−2 4.13×10−4 6.61×10−3 PKS 0537-441,1 08-Oct-08 54747.866 4.37 6.10 - 3.6 1.27×10−2 2.03×10−1 2.25×10−3 3.60×10−2 8 PKS 0537-441,2 08-Oct-08 54747.869 5.05 6.76 3.2 - 6.41×10−3 1.02×10−1 1.16×10−3 1.85×10−2 PKS 0537-441 26-Jan-05 53396.605 5.68 8.08 2.2 - 3.40×10−3 5.44×10−2 3.11×10−4 4.97×10−3 PKS 1424+240 13-Apr-13 56395.315 3.50 5.45 - 4.9 1.47×10−2 2.35×10−1 9.68×10−4 1.55×10−2 PKS 1510-089 28-Apr-11 55679.860 5.94 6.96 - 5.8 2.63×10−3 4.21×10−2 9.46×10−4 1.51×10−2 RX J0324.6+3410 15-Jan-13 56307.859 3.91 9.46 5.7 - 3.57×10−3 5.71×10−2 1.21×10−7 1.93×10−6 S5 0716+714 14-Dec-09 55179.704 4.05 6.62 - 2.2 1.74×10−2 2.78×10−1 1.33×10−3 2.13×10−2 S5 0716+714 15-Dec-09 55180.299 5.51 10.98 - 3.2 4.05×10−3 6.48×10−2 1.71×10−5 2.73×10−4 W Com 09-Jun-10 55356.738 3.78 5.36 - 8.1 2.28×10−2 3.65×10−1 4.68×10−3 7.49×10−2 Notes.—Col. (1): ObjectnamereferredtoonSwift-XRTmonitoringsite. Col. (2): Dateofobservation. Col. (3): Starttimeofobservation in MJD. Col. (4): Reduced χ2 of weighted mean straightline fit. Col. (5): Reduced χ2 of simulated quiescent line fit. Col. (6): Estimated doubling time of potential flare in minutes. Col. (7): Estimated halving time of potential flare in minutes. Col. (8): Estimated p-value significance of weighted mean straight line fit before trials factor. Col. (9): Estimated p-value significance of weighted mean straight line fit after trials factor. Col. (10): Estimated p-value significance of quiescent line fit before trials factor. Col. (11): Estimated p-value significance of quiescent line fit after trials factor.