2017년도 이전 세미나
I: Mathematical Approach for climate change studies, II: Climate change: finding evidences and causes
|Category||2016 Fall Math Colloquium|
|날짜||2016-12-09||시간||15:50:00 ~ 18:00:00|
|장소||Math Sci Bldg 404||Host||Kunwoo Kim|
|Speaker||Jong-Seong Kug, Seung-Ki Min||소속||POSTECH|
|TOPIC||I: Mathematical Approach for climate change studies, II: Climate change: finding evidences and causes|
|소개 및 안내사항||국종성 교수님(Prof.Jong-Seong Kug)
Title: Mathematical Approach for climate change studies
Abstract: Final goals of climate studies are to understand our climate system and its physical processes, and to predict future changes of the climate system on time scales of about 10 days to a few hundred years. In general, there can be four approaches to pursue the two goals of climate sciences to a large extent. The first is observation and monitoring for climate system. Various mathematical approaches are being applied to develop algorithms of in-situ and remote observations and their calibration. The second is to find meaningful signals throughout big-data analyses, and interpret them physically and dynamically based on our current knowledges. To extract the signals, many linear and nonlinear statistical tools are utilized, which come from mathematical formulas. Still, there is unceasing efforts to develop a new tool including artificial intelligence. The third is to utilize numerical models, which consist of huge partial differential equation sets, to mathematically represent physical and dynamical processes of the climate system. In order to express physical and dynamical processes probabilistically to mathematical formulas and improve computational efficiency, various mathematical approaches are being applied. Lastly, there is theoretical approaches to understand climate phenomena conceptually. By extracting key processes, the huge partial different equations for climate system can be simplified, or sometimes linearized, and then analytical solution can be derived. The analytical solutions can be much helpful for understanding characteristics of climate phenomena and predicting their future changes. In this talk, these mathematical approaches will be introduced, and future collaborating with Math people will be suggested.
민승기 교수님(Prof.Seung-Ki Min)
Title: Climate change: finding evidences and causes
Abstract: This seminar introduces the basics of climate change science and associated mathematical approaches, focusing on human impacts on climate. Reliable prediction of future climate changes is fundamental information to better cope with the global and regional impacts of climate changes. First step for this is to identify causes of the past observed climate changes. In particular, quantifying human contribution, due mainly to the increase in greenhouse gases (GHG) from the Industrial Revolution, is required, and for this purpose, climate scientists have been using climate models. Climate models are mathematical equations representing the Earth climate system components and their interactions, which are governed by the physical conservation laws. The climate models are integrated under various conditions of external forcing factors such as the observed GHG increases and aerosols, and the simulation results are compared with the real observations. This observation-model comparison is more rigorously assessed using a mathematical approach based on multiple linear regressions. This approach enables us to effectively find evidences for human influence (signal) on climate given the ranges of natural climate variability (noise). The signal-to-noise problem is called ‘detection and attribution’ of climate change, and some details will be explained with recent studies on the Arctic sea-ice melting and the heavy precipitation intensification. Finally, challenging questions will be discussed in terms of the role of mathematics.