"I look forward to integrating AI into space weather forecast"

Kathmandu — Nirakar Sapkota, a Princeton University graduate, is making significant achievement in space weather research. His work has notably advanced the prediction and mitigation of solar storm impacts on Earth’s technologies.

Sapkota has developed methods to safeguard satellites from solar particles that can cause costly damage and service interruptions. His research predicts these particle influxes, helping operators schedule maintenance and adjust satellite positions effectively.

Since graduating at the top of his class in 2017 from St. Xavier’s College, Sapkota's innovative approaches have garnered global use, influencing space weather models and protecting crucial technologies across the world.

Sapkota enjoys playing table tennis and performing magic tricks in his spare time.

In an email interview with Binod Dhakal for Ratopati, he covered a wide range of topics like space weather prediction, solar storms, his research methods, and the connection between polar cap potentials and field-aligned currents.

What thing in particular, be it any moment or a project, attracted you to the research of space weather?

The idea that events on the Sun, millions of miles away, could directly affect life on Earth fascinated me. Realizing that an invisible solar event could disrupt modern infrastructure highlighted the urgency of understanding and predicting space weather. This need for preparedness drove my passion for this research.

How can your work in solar storm forecast and the impact it has on technologies change disaster preparedness or response?

My research focuses on identifying when geomagnetic disturbances are most likely to generate geomagnetically induced currents (GICs) that can overload power grids. Using wavelet analysis, I detect sudden spikes in GICs during storms, pinpointing when infrastructure is at risk. This real-time insight allows grid operators to take targeted actions, like shutting down only the most vulnerable parts of the grid, instead of widespread shutdowns.

What are some of the major differences between your weather predictions models and hitherto existing conventional models?

The main difference is in data analysis. My models use wavelet analysis to catch sudden changes or "spikes" in the data, like when a solar storm intensifies. Traditional models often look at broader patterns and might miss these sharp variations. Focusing on these moments, we can make more accurate, real-time predictions about storm timing and strength.

How does your graduation from Princeton shape your research methodology and your approach in dealing with the complexities of space weather?

Princeton encouraged multidisciplinary learning, allowing me to explore beyond physics. I got involved in research on augmented reality systems and machine learning, which introduced me to advanced computational techniques. This cross-disciplinary exposure shaped my problem-solving approach by combining the precision of physics with the adaptability of computer science and machine learning.

Briefly enlighten us on the rigorous process of developing your prediction model. What are the key factors or data you take into account during the Process?

Developing my prediction model involved analyzing extensive interplanetary datasets, including solar wind speed, interplanetary magnetic field, and geomagnetic indices. I applied continuous wavelet transforms to detect time-sensitive patterns in these variables. The process requires understanding the relationship between field-aligned currents, polar cap potentials, and GICs to accurately predict when solar storms will impact Earth's infrastructure.

You claim that your research has helped to minimize the effects of solar storms on critical communication infrastructures like power grids. Could you please give some examples in the context.

For instance, during intense geomagnetic storms, my research on GICs can provide insights into the timing and severity of electrical disturbances on the ground that can affect power grid networks. Predicting these disruptions with greater accuracy can help with preemptive measures, such as grid isolation or energy rerouting, reducing the likelihood of blackouts and equipment damage.

Satellite protection from the harmful effects of solar storms is one of the achievements of your research. What are the major challenges and how have you overcome them?

One of the significant challenges is the unpredictability of solar storm intensity and its variable effects on different satellite systems. With a focus on the relationship between polar cap potentials and field-aligned currents, my work can help to better predict when and where satellite disruptions are likely to occur. These models help satellite operators take precautionary steps, such as switching to lower energy modes or altering orbits to avoid damage.

Can you tell us something about your future research undertakings or anything specific that has been drawing your attention?

I’m looking forward to integrating AI into space weather forecasting, particularly to improve real-time predictions of solar storms and their effects on Earth’s magnetic field. I’m also looking into how space weather can impact newer technologies like autonomous vehicles and 5G networks, which are sensitive to electromagnetic interference.

How often do you collaborate with other scientists in the field?

Collaboration is a necessity in my work. Depending on the expertise needed for specific projects, I can collaborate with researchers from any part of the world. Once you reach a certain level of recognition in the field and have works that scientists around the world have used, people want to work with you. The early years are hard, but now I can pretty much collaborate with any professor from most universities in the world.

Does your work contribute to the broader scientific community?

Yes, my research contributes to the scientific community, particularly in understanding the coupling between solar wind and Earth’s magnetosphere. My works have been published extensively in journals and cited by researchers around the world.

Space weather phenomenon is full of complexities. What are the challenges you face to make your research understandable to common people having little or no scientific knowledge?

Space weather is inherently complex as the charts can look scary. But there are ways to simplify things for the public. For example, I often compare geomagnetic storms to normal storms we have on earth, drawing parallels between their destructive capabilities. Simplifying these concepts helps me communicate the importance of preparedness to broader audiences. And explaining what kind of damage these events can do gives the public a clearer picture. As an example, how would your life change if you were without internet and electricity for weeks? That is a real possibility if we get hit by a big enough geomagnetic storm.

Do you have any words of advice to aspiring scientists or researchers in the field?

My advice is to remain curious and interdisciplinary. In today’s world, the boundaries between fields are becoming blurred. Having a strong foundation in one field while being open to learning from others—like how I combined computer science with space weather —can lead to breakthroughs in research.

Lastly, where do you see yourself ten years down the line?

 Looking ahead, I'm excited about the potential developments in space weather research. While it's challenging to predict the specifics, I will continue contributing to the field of space weather in meaningful ways. I'm particularly interested in the integration of AI and advanced technologies in space weather forecasting. My goal is to be part of collaborative efforts that enhance our understanding and prediction capabilities, ultimately improving our resilience to space weather events. Whatever role I find myself in, I'm committed to advancing this crucial area of science and its practical applications for society.