Research

The solar corona is a million Kelvin hot and dynamic upper atmospheric layer of the Sun. It is highly structured due to the dominance of the magnetic field at this height (we find lots of beautiful looking loops in EUV corona). One of the long-standing problems of solar physics is to examine why the corona is so hot (FYI, photosphere is at ~5770K). This is somewhat counter-intuitive as one expects to have cooler temperatures as you move away from the source (i.e., center of the Sun). After decades of research on this problem, two major sources have been identified as potential solutions to this 'coronal heating' problem. These two sources are: 1. MHD waves, 2. Magnetic reconnection. Interestingly, it is quiet possible to have wave-modulated reconnection or reconnection-driven waves and hence, the added complexity.

Magnetic reconnection is a process where magnetic field lines in the solar atmosphere rearrange and release vast amounts of energy. This phenomenon occurs when oppositely directed magnetic fields come into close contact, breaking and reconnecting in a highly dynamic event. It plays a key role in driving solar flares, coronal mass ejections, and heating the solar corona. By converting magnetic energy into kinetic and thermal energy, magnetic reconnection helps explain some of the Sun's most energetic and mysterious processes, impacting space weather and Earth's magnetosphere.  

To better understand the characteristics of magnetic reconnection, I analyze the evolution of various solar features, such as jets, and explain the observations using magnetic reconnection theories. As mentioned earlier, magnetic reconnections play a major role in addressing the coronal heating problem; however, reconnection and waves are not mutually exclusive phenomena. Therefore, my research on waves is also an integral part of my quest to understand the properties of reconnection. Since not all reconnection events exhibit the same signatures, it is important to explore phenomena through multi-wavelength data. Lastly, I routinely compare the observations with predictions from state-of-the-art numerical simulations.

Physical processes in the Sun occur at variety of time scales, ranging from few seconds to few tens of years. Solar cycle often refers to the cyclic variation in sunspot number or in sunspot area with a period of ~11 years. Sunspots are the proxy for the magnetic fields of the Sun and hence, such variation related to the sunspots will obviously mean a variation of solar magnetic fields. Thus, almost all the solar phenomena will be affected by this solar cycle including number of solar flares and occurrences CMEs (CMEs or Coronal Mass Ejections are large scale plasma eruption from the Sun). These two things (flares & CMEs) are particularly important to us as they affect the space-weather (Sun-Earth inter-space) heavily.

In this research are, I focus on studying the Sun's magnetic activity using century-long solar observational data from various observatories around the globe. One of my key works involves compiling a sunspot area database spanning nearly 90 years (1921–present) using data from the Kodaikanal Solar Observatory. This publicly available dataset has enabled several important studies on solar active longitudes and sunspot size variations. Additionally, my research includes constructing a calibrated sunspot area catalogue by combining data from different observatories. My catalogue, the most comprehensive of its kind, provides daily sunspot area measurements and is widely used to study solar irradiance, sunspot group distribution, and other sunspot properties. You can access both datasets through this link (click).

 Retrieving information from old historical sunspot observations requires special feature detection techniquies (and lots of patience). In any given historical catalogue, images are often found to be in-homogeneous e.g., change in drawing style (pencil to pen & black-and-white to color), change in capturing device (change in paper), aging effects (dust in photograpic plates), miss-handeling of images (leading to artifacts like scratches) etc. Moreover, these catalogues often contain tens of thousands of images. Hence, we need suitable feature detection methods that are not only accurate but also fast enough to complete the detection in a reasonable time. I am involved in generating such algorithms for detection of sunspots, filaments and plages. Recently, I have started exploring the usage of neural networks in such detection and even applied one kind of network onto the sunspot drawings from Purple mountain observatory, China and the preliminary results indeed look very promising.