The velocity and magnetic fields at two heights of the Sun's atmosphere: Flare forecasting contribution

Transcription

1 The velocity and magnetic fields at two heights of the Sun's atmosphere: Flare forecasting contribution R. Forte 1, S. Jefferies 2,3, F. Berrilli 1, E. Pietropaolo 4, S. Scardigli 1, D. Del Moro 1 ; F. Pucci 1, N. Murphy 5 ; M. Oliviero 6 Thanks to K. Loumou 1 University of Rome Tor Vergata, ITALY 2 Institute for Astronomy - University of Hawaii, Maui, USA; 3 Georgia State University, Atlanta, USA 4 University of L Aquila, ITALY 5 NASA Jet Propulsion Laboratory, USA; 6 INAF/Osservatorio Capodimonte, ITALY ISSI Flare Meeting, October 2016

2 UNITOV Solar and Space Team info at:

3 + 3 staff for SEP Space Instruments (PAMELA and ALTEA experiments) + 3 PhD Students (F. Pucci, M. Lovric, G. Viavattene) + 1 PhD-AQ (G. Napolitano) + 4 PostDocs (S. Scardigli, M. Mergè, M. Martucci, A. Rizzo)

4 UNITOV main funded projects: European Solar Telescope: FP7-EST, FP7-SOLARNET, H2020- GREST, H2020-INFRADV-PREEST (members from 15 European countries) WP on large etalons and Heat-Rejector. IPS Ionospheric Predicion Service project has been financed by the European Commission through a Tender PRIME Contractor: TELESPAZIO PARTNERS: INGV + UNOTT + UTOV SWERTO: Space-Weather at the University of Rome Tor Vergata has been financed by the Regione Lazio FILAS-RU for the period April 2015 March Design and realize a data-base with the particle fluxes recorded by the space missions: PAMELA, ALTEA, Alteino, SilEye, NINA, and with the spectro-polarimetric measurement of the solar photosphere-chromosphere from the IBIS (HiRes) and MOTH (Full Disk) multi-lines observations.

5 MOTH II - Magneto-Optical filters at Two-Heights Instrument specifications: Full disk images CMOS 2048x2048 pixels Aperture: 20 cm Pixel scale: 1 arcsec/pixel Sensitivity: 7 m/s for v; 5 Gauss for B. line λ (nm) Formation height (km) K I Na D Ca I He I Fe I (HMI) Ni I (MDI)

6 HMI B LOS (h 125km) K B LOS (h 330km) NaB LOS ( 700km) MOTH instrument is able to evaluate both horizontal x,y B LOS and vertical magnetic z B LOS gradients, Dopplergrams (i.e., V LOS images) and intensity images. Line-of-Sight approximation of vector magnetogram LOS WL SG = ( B LOS )dl integration is along strong-field intervals of the AR neutral lines.

7 Preliminary analysis of B Na /B K normalized images in a selected ROI shows interesting features inbetween and near bipolar regions. A new flare forecasting algorithm will be based on parameters which characterize the properties of the solar active region. free magnetic energy, dynamics proxies and vertical magnetic features.

8 MOTH Synoptic Telescopes: a second instrument, necessary to realize a simple network for solar h24 multiline observations, can be realized in Europe and installed at Canary islands (Tenerife or La Palma observatories). Mees Solar Observatory Maui, Hawaii Izana Observatory Tenerife, Canary Islands

9 MOTH Calibration Pipeline

10 Calibration Pipeline (Level 0) Data Acquisition: Flat field frames FF Leak frames L Dark frames D Observations I Preliminary Data check Image statistics (mean values, max, min, rms) Temperature check (MOF, CMOS, prefilter, box,...) Tracking Dark current Flat Filed & Leakage Images

11 Calibration Pipeline (Level 1) <D> FFFF CC = FFFF/FFFF bb LL CC =Lf Mask(j) I (j) - <D> I c Flat field correction I C (j)= I Lc (j) / FFc Leakage correction I Lc (j) = I Dcorr (j) - L c Atmosp. transp. & Doppler tr. corrections I atcorr (j)= I dc (j) C(j) I Dcorr (j) = I atcorr (j) - K(j) Registration Pipeline (Level 2) Geometrical corrections σ+,σregistration (Same camera) B,R registration (Two cameras) K,Na registration (Two channels ) I c

12 Magnetograms and Dopplergrams computation (Level 3) BB LLLLLL vv LLLLLL Calibration parameters Magnetograms M = RR+ BB + RR BB RR + +BB + RR +BB Dopplergrams D = RR+ BB + + RR BB RR + +BB + RR +BB I c Results: Magnetograms at two heights of the solar atmosphere, in Na D2 (589 nm) h= km K I (770 nm) h= km which allow to evaluate the magnetic ratio along the Line- Of-Sight: LOS B Na LOS B K Example of a LOS magnetogram in K channel. July 10, :27:00 UT

13 EVENT FORECASTING data data acquisition data acquisition data acquisition acquisition features events detection indexes knowledge DB forecasting models heterogeneous data prediction - data harvesting - standardization procedures - post-elaboration algorithms - indexes definition databases - storages - mirroring - dissemination -...

14 - data harvesting - SDO/HMI -> JSOC procedure - GONG-H -> web crawling - GOES-X -> ftp -... features events detection indexes knowledge DB - standardization procedures - resolution (imaging) - disk rotation (imaging) - timescale homogenization post-elaboration algorithms - data quality - lacking data management - event detection - active region definition indexes definition - magnetic indexes (local fluxes, R,...) - eliocentric angles databases - under realization NEXT Company - storages - TBD - mirroring - TBD - dissemination - TBD -...

15 quick recap display background picture: latest Hα from the selected time interval candidate events (Hα) mouse-hovering: SDO/HMI indexes selection form

16 candidate event details GONG Hα coordinates mean magnetic fluxes SDO/HMI (MOTH) magnetograms R indexes target areas source (observatory) event time SDO/HMI release SDO/HMI time

17 candidate event record system reference time source (observatory) GONG Hα SDO/HMI quality control H event magnetic fluxes magnetic indexes

18 Active Region Automatic Detection In order to point out the active regions, two thresholds are set, chosen through a statistical analysis of the image values, e.g. : Values II xx, yy < 150 GG are excluded, to remove the Gaussian noise; Values II xx, yy > 5 kkkk are excluded, to remove the saturated points. The remaining pixels are enhanced with a gaussian convolution. Sample of K channel (10 July 2014 h: 19:27:00) Where: 150 GG < II xx, yy < 4000 GG

19 Active Region Automatic Detection The areas with strongest magnetic field, remaining from the previous computation, undergo a gaussian smoothing, that outlines the active regions contours. Each area is labeled with a growing number, from left to right. This labels can be compared and converted in the official NOAA HARP HMI Active Region Patches. Nevertheless, the selected areas do not necessarily correspond to the NOAA ARs, the main purpose of this work is to provide an algorithm which is able to recognize from magnetograms where there is a strong magnetic activity. 1 3a 2 3b 4 5 Sample of K channel (10 July 2014 h: 19:27:00)

20 Active Region Automatic Detection Here is shown the same magnetogram acquired for HMI (left) and MOTH II K channel (right). The two images have been rebinned to have the sime pixel dimension, and the algorithm has been applied to both of them, with different thresholds (i.e. II xx, yy > 600 GG). HMI magnetogram (10 July 2014 h: 19:27:00) MOTH II magnetogram - K channel (10 July 2014 h: 19:27:00) Above is shown the HMI magnetogram (10 July 2014 h: 19:27:00) with active regions number from NOAA.

21 Schrijver photospheric metric R: the unsigned flux on the inversion polarity line which can be calculated for each active region, in order to provide a probability for the region to flare. R Schrijver computation For each AR, the parameter R is computed as follows: In the ARn, are selected the points (x,y) where II xx, yy < 70 GG If it is not true, then II xx, yy = 0 AR n=7 The image obtained ondergoes a gaussian convolution to enhance the values. Then R is computed RR = II xx, yy xx,yy

22 Active Regions detection Region 1 (llllll 10 RR = 5.22) corresponds to NOAA 2113, which developed one M class flare and 4 C class flares within 11th of July a 3b 4 According to Schrijver s study, regions presenting a value of llllll 10 RR > 5, there is an 80% probability to have an M1 class flare, 35% of an M3 flare, 20% of an X1 flare and 1-2% of a X3 flare. Calibrated for MDI, 2 5

23 The Pro and Cons of Flare Forecasting and Prediction with MOTH Pro: simultaneous Na and K magnetograms and dopplergrams (two heights) in solar atmosphere Possibility to extend the observations to Ca line (422nm) and He triplet (1083nm) high-cadence data enable unprecedented possibilities to define dynamic/multi-line forecasting parameters (see example) 2 MOTH telescopes (Hawaii + Europe) can create a basic network for H24 observations Cons: Currently we have sparse data and low statistics Probably, some years are necessary in order to obtain the appropriate statistics Magnetic field lines, that are close to perpendicular to the surface, have occasional significant inclinations possibly associated to topological reconfigurations of magnetic field. At this lesser angle, magneto-acoustic portals open briefly, allowing sound-wave energy to escape into the upper atmosphere. Acoustic power gives information about magnetic topology. Acoustic power Na D2 KI Photos.