Observing the Sun with LOFAR

1

Peijin Zhang

​

University of Helsinki, FI

IANAO, BG

This project has received funding from the European Union's Horizon 2020 research and innovation programme under grant agreement No 952439.

(A) Quiet sun study

• Majorly in microwave band
• Quiet sun in > 1GHz well described by thermal bremsstrahlung

​

• Based on model of density and temperature of lower solar atmosphere

2

Imaging of Quiet sun with LOFAR LBA: Configuration

​

• Antenna set: LBA (10MHz-80MHz)
• Frequency resolution: ≈1MHz
• Time integral: 2.45 hrs (Quiet sun)
• 24 core stations and 9 remote stations (528 station baselines)
• Longest baseline : 48km
• Calibrator: Cas-A

3

4

Coronal Holes .vs. Radio image

5

Brightness temperature spectrum

Model Model and observation

6

Center averaged (<0.5 Rs)

Spectrum of coronal holes and dark regions

7

Slice in the equator and meridian

8

Size of the sun in radio

9

Conclusions

Scientific

• The brightness temperature measured as center average is lower than model predicted of thermal bremsstrahlung
• Equator coronal holes are bright in frequency range of 10-80MHz
• We find an extremely low brightness temperature region on the solar disk
• The size of the sun is larger in E-W direction.
• The size shares the same trend with the local plasma radius

​

Technical

• Quiet sun imaging requires long time integral and flagging

​

10

(B) Radio Frequency Interference (RFI)

• Definition:

unwanted electromagnetic radiation

interferes with the intended radio signals.

• Sources:

​

​

11

7/7/23

Farming infrastructure

Natural radio emission

Instrument itself

Wireless communication

Why should we care

• Interferometric imaging:

Significantly influence the image quality

(Some experience: If you lose 10% of your visibility, you will still have an OK image, but if you have 1% super strong RFI, the image will look like a mess)

​

• Spectroscopy:

It covers interesting structures in the dynamic spectrum

​

​

12

7/7/23

Visualization and identify

• First: know your target source

Time-frequency variation

​

​

• Tools:

rfigui (André R. Offringa)

casaplot

​

13

7/7/23

RFI Mitigation (1)�Stay away from RFI

14

7/7/23

Rob Millenaar

MIDPREP/AAMID workshop 2016

Exloo, NL

(LOFAR)

Karoo

(MeerKAT, SKA-mid)

Daocheng, CN

(DSRT)

But RFI travels far and it is everywhere and it is getting worse

RFI Mitigation (2)�Instrument filtering and narrow beam

15

7/7/23

Credit: Cees Bassa

RFI Mitigation (3)�Data processing

• Manually removing (with visualization)

​

• Threshold, semi-auto

​

• Algorithm, e.g. AOFlagger, ConvRFI

​

​

16

7/7/23

Possible methods

• Statistical based:
• SumThreshold
• AOFlagger

[Offringa et al. 2012, A&A, 539, A95]

​

• Machine learning based:
• CNN-based [Sun, H., et al MNRAS 2022]

​

• Feature matching based:
• ConvRFI

17

Features

• Most of the RFI are lines and edges

​

​

​

​

​

​

​

• Solution: to find the pixels where there are lines and edges.

18

[Fourier]→

time

frequency

time

frequency

[Fourier]→

Matching cores

• Mini-array (3*3)

19

-1

-1

-1

2

2

2

-1

-1

-1

1

1

1

1

1

1

1

1

1

Elementary multiply and sum()

~0

=0

>0

Matching cores

• Mini-array

20

Implemented with Pytorch

Compare with AOFlagger

• AOFlagger-default

(too aggressive)

​

• AOFlagger-local-RMS

(passing through most of the burst)

​

• ConvRFI

(have some left-over RFI)

​

• Hybrid

(best precision)

21

Result on observation

• Before and after

22

Result on simulation data

• Add weak RFI with ‘hera_sim’

23

Speed benchmark

• Implemented with pytorch

​

• Speed benchmark with 7.6 GB

​

24

>300MB/s

Flagging correctness

25

Implementation

• Dynamic spectrum pipeline

26

Summarize

• Need to consider RFI in all ground-based radio observation
• Source of RFI: instrument, natural source, human activities
• Mitigation:
• Instrument level
• Data processing
• ConvRFI is open-source: github.com/peijin94/ConvRFI

​

​

27

7/7/23

(C) IDOLS: Incremental development of �LOFAR for spaceweather

28

7/7/23

Performance of radio telescope for spaceweather monitor

• Sensitivity
• Capability to distinguish weak signals

​

• Resolution
• Capability to Resolve fine structures

​

• Coverage
• Capability to monitor spaceweather

29

IDOLS

IDOLS

• Incremental development of LOFAR for spaceweather

​

• Station CS032

​

• 96 LBA
• 48 HBA

30

[van Haarlem 2013]

IDOLS observation

• Day time:
• Solar observation

​

​

​

​

• Resolution:

0.01s, 12kHz

31

(a)

(b)

IDOLS observation

• Night time:
• Ionosphere
• IPS

​

​

​

​

• Resolution:

0.01s, 12kHz

32

https://spaceweather.astron.nl/SolarKSP/data/IDOLSarchive/idols.html

Sensitivity

• Callisto and IDOLS

33

Sensitivity

• Callisto and IDOLS

34

Sensitivity

35

• Background astronomical radio emission

​

• TODOList: use sky-model to remove the slowly varying background

Daily summary of 6 days

RFI handling

• Algorithm-based RFI flagging

​

• Feature matching the lines and edges

36

[Zhang et al 2022 manuscript in preparation]

Event detection

• Algorithm:
• Binarization
• Hough transform
• Obtain Hough peaks
• Modeling the frequency drift line

​

• Open sourced

https://github.com/peijin94/type3detect

37

[Zhang et al 2018]

Event detection

Too many

38

Overall summary

• New tools developed for data processing: FastQSL
• Features updating in: LOFAR-Sun-Tools
• Scientific results form LOFAR observations: Quiet Sun

39

7/7/23