1 of 10

How precisely we can determine the thermal energy of weak flares based on STIX images? �

Litwicka Michalina, Mrozek Tomasz and Arun Kumar Awasthi

Space Research Center, Polish Academy of Science in Wrocław, Poland.

​

The Solar Physics High Energy Research (SPHERE) workshop 

​

2 of 10

A „weak” flare

3 of 10

MARLIN imaging algorithm

  • Lucy – Richardson algorithm (similar to EM, different approach to transmission matrix calculation)
  • Uses counts detected in pixel
  • Gives us possibility to subtract or include background counts
  • Number of iteration should be declared before the calculation – it has direct impact on source sizes
  • Overresolution after many iteration steps (typical beaviour for L-R algorithm)
  • Volume stabilizing condition
  • ΔV = 2|V20-V10|/(V20+V10) <= 1%

​

​

​

​

​

​

​

​

Algorithm stops at 360 step

​

4 of 10

��FFLARE = FOBS  - FBKG

FCAL =

FLARE

IMAGE

MTRANS * 

   FCAL                FOBS - FBKG

(n)

Wn

FLARE

IMAGE 

(n+1) =

FLARE 

IMAGE 

(n)

Wn*

N iteration

FOBS - vector of counts measured by each pixel stripe

FCAL – vector counts values calculated

FCAL =

MTRANS * 

FLARE

IMAGE

(n)

+FBKG

   FCAL                 FOBS

Wn

FLARE

IMAGE 

(n+1) =

Wn*

FLARE 

IMAGE 

(n)

N iteration

1) Background subtracted

from pixel data counts

2)  Background included  to 

image calculation

Different approaches to background in MARLIN

5 of 10

Background subtracting or including �

  • Smaller sources with higher signal in center
  • Sidelobes visibles in image,
  • Negative values

​

  • Overresolution
  • Worse counts statistics
  • Not physical shapes (sharp edges, verry narrow structures etc.)

​

6 of 10

Chosen microflares (B1 – C1 classes) 20-21 Sep 2021 volume estimation

Volume estimation depend on:

  • Real geometry and size of x-ray source
  • Number of MARLIN iteration step
  • Background approach (subtracted or included)
  • Size definition: 50% of total signal contour

50% of maximum value contour

​

 

In order to compare results volume was also estimated for mem-ge reconstructed sources.

7 of 10

Weak flare example �20-Sep-2021 23:04

1)Bkg Includ 50 % sig

2) Bkg subtract 50 % sig

3) Bkg includ max val

4) Bkg subtract max val

1) Iteration number: 300

Semi axis: 7.32 10.55

50% sig. = 28% max. value

2) Iteration number: 250

Semi axis: 7.29 10.45

50% sig. = 29% max. Value

3) Iteration number: 490

Semi axis: 4.77 8.02

4) Iteration number: 440

Semi axis: 4.65 7.93

​

8 of 10

Fitting all spectra with OSPEX

Temp : 11.81 mln K

Temp : 12.42 mln K

Temperatures determine for all flares are in ranges : 9.44 mln K – 14.80 mln K

9 of 10

Thermal energy range all�weak flares

1] 1.38 +/- 0.17

2] 1.16 +/- 0.17

3] 1.07 +/- 0.14

4] 1.54 +/- 0.17

5] 2.09 +/- 0.42

6] 1.14 +/- 0.14

7] 1.40 +/- 0.36

8] 2.85 +/- 0.36

9] 2.73 +/-0.73

10] 4.01 +/- 0.48

11] 4.55 +/- 0 .61

12] 0.86 +/- 0.10

13] 1.94 +/- 0.17

28

Different approach to volume estimation from image gives us

Errors ranges from 9% to 27%

10 of 10

Thank you�