MSV-2035 Astronomy Document - Inside Design - FINAL - FINAL

Astronomy & Astrophysics 92 information about the Universe at scales not witnessed before. This is expected to increase in the future with data from various large projects. Such data are challenging our present ability to analyse the data, and also our theoretical understanding, requiring better simulations.Also, it is important to make the large volumes of data accessible to larger communities through databases. These require efficient and high performance computing facilities, large data servers and implementation of machine learning and artificial intelligence techniques. Discussed below are some specific areas and facilities that have special computing needs. 5.7.1 Computational Centre for SpaceWeatherForecasting Space-based observations of solar activity (through remote imaging instruments) and (in-situ) measurements of local variations in space provide important constraints for space weather phenomena. These observations, however, do not generate the comprehensive information that results in a holistic understanding of cause and effect, i.e., origin-to- impacts. An observation is made when a space weather event (say, flare) has just occurred or is occurring, or an interplanetary disturbance (such as a CME) is already near-Earth. This does not provide the necessary time window for space weather forecasts that can be translated to mitigation strategies by national space agencies, defense establishments and space-reliant industries. Thus, space- and ground-based observations need to be supplemented by data-driven computational magnetohydrodynamic models (i.e., physics based simulations) andArtificial Intelligence and Machine Learning algorithms to return actionable space weather forecasts. This is already realized by advanced space-faring nations, with the US having established the NOAA Space Weather Prediction Center, UK Met Office hosting the UK SpaceWeather Unit, the European Space Agency hosting the Office of Space Weather and China establishing the State Key Laboratory for SpaceWeather. Indian expertise in the domain of computational space weather forecasting, while still limited, is growing with the multi-institutional Center of Excellence in Space Sciences India at IISER-Kolkata taking the first initiatives to develop predictive space weather models. The solar, space sciences and heliophysics communities recommend rapid escalation of these efforts towards creation of a national space weather modelling and forecasting center, which relying upon multi-institutional expertise, can generate indigenous space weather forecasting capabilities for India. This translational activity will immensely benefit multiple sectors such as space, defense, communications and civil aviation. 5.7.2 ForRadioAstronomy – SKA The SKAtelescope, when it becomes operational, will produce several hundred petabytes of scientific data every year. This prodigious volume of data will require a complete paradigm shift in the working model of radio astronomers. They will no longer be able to download data to their desktop for analysis and interpretation. SKA data analysis will require the incorporation of what is presently considered specialized, rare and centralised High-Performance Computing (HPC) and data analytics systems into the very fabric of operational facilities. Alongside this, considerable innovation will be needed in algorithms and methodologies for the analysis of these huge datasets. Advanced tools will need to be developed to enable visualisation of large data volumes as well as for finding rare astronomical objects whose discovery may be paradigm-changing. India is uniquely positioned to leverage its well- MEGA SCIENCE VISION-2035