연구보고서
- 저자
- 윤선권 박사
- 작성일
- 2016.01.23
- 조회
- 372
- 요약
- 목차
The mechanism of major climate phenomena such as El Niño and La Niña has not yet been revealed, also it is impossible to completely simulate using the physical climate model, where the global environment of the large scale climate phenomenon, and its weather conditions. Furthermore, analyzing the climatic phenomena through the physical model to exist in limited areas, predicting involves considerable difficulties. Based on the results of the physical model, when applied to hydrologic circulation system in a small region such as Korean Peninsula, it is possible to follow a problem in scientific credibility and understanding. Therefore, this study analyzed non- linear behavior links with atmospheric teleconnections between hydrologic variable and climate index using statistical model over the Korean Peninsula with the ocean-related major climate factors such as ENSO (El Niño- Southern Oscillation) and IOD (Indian Ocean Dipole) in the tropical regions. The statistical approach has the disadvantage that it is difficult to expect the results of significance level due to the limited number of observations, it is one of the important way that can be utilized to complement to the prediction result of the physical model.
This study is used to the SSTA (Sea Surface Temperature Anomaly) data provided by the NOAA (National Oceanic and Atmospheric Administration). And then we divided different conditions of SST of the traditional CT (Cold- Tongue) El Niño phase and a new type of WP (Warm-Pool) El Niño phase in the Tropical Pacific Ocean (TPO), and also positive (+) and negative (-) IOD phases in the Tropical Indian Ocean (TIO). The analysis of the lag time correlation with climate indices (e.g. ENSO and IOD) and hydrologic variables (e.g. precipitation and runoff) in the 5-Major River Basin was performed by principle component analysis (PCA) using a singular spectrum analysis (SSA), which is widely used in low frequency analysis for time series data. Using this result has estimated to the joint probability density function by kernel density estimation. As well, the SSA results are used to analysis with the atmospheric teleconnections derived from linear and non- linear correlations by Mutual Information (MI) techniques.
The composite anomalies of JJAS (June to September) precipitation CT/WP El Niño decaying years and La Niña decaying years based on global observed data from the NCEP (National Centers for Environmental Prediction) / NCAR (National Center for Atmospheric Research) reanalysis 2 over the East Asia and the Korean Peninsula have expected to clearly decreasing tendency during CT El Niño year compared to the normal years. Particularly, during WP El Niño year has expected to increasing tendency then normal years. In addition to La Niña year shows tended to slightly increasing tendency. Warm season anomaly precipitation of positive (+) IOD and negative (-) IOD years, it shows slightly decreasing tendency and normal state, respectively.
The singular spectrum analysis (SSA) of the eigenvalue analysis of precipitation and runoff over the Korean Peninsula have explained 54.8%, and 63.4% by empirical orthogonal function (EOF) 1-12 mode. And also the climate indices (such as, NCT, NWP, and IOD) have explained 65.4%, 71.4%, and 61.0% by empirical orthogonal function (EOF) 1-6 mode, respectively. Therefore, we used the reconstructed component time series data from the SSA techniques by selecting the RC 1-12 and the RC 1-6 corresponding to a relatively short period. Results of correlation analysis between hydrological variable and climate index has came out high correlation with non-linear correlation analysis using the MI technique more then linear correlation analysis. In case of NCT index and NWP index shows that there are lag correlations of several month. But for the case of IOD index shows direct correlation. As a result from this study, extreme climate pattern changes and their hydrologic local impacts over the Korean Peninsula, it shows significant increasing tendency during the WP El Niño decaying year, the CT El Niño years are significant decreasing tendency, and during the La Niña years have analyzed as to maintain the average state compared to the normal years. In addition, during the positive IOD years have clearly appeared decreasing tendency of precipitation and runoff to compare with the normal years. But for the negative IOD years are analyzed that keeping with the normal year state and slightly decreasing tendency.
Finally, to carry out utilizing of research results and policy support, we analysed that the relevant provisions, guidelines, and the business work for the government affiliated ministry and agencies. As a result of this study, it is believed that can be utilized to provide long-range river flood forecasting, reservoir operation, and water resources management during rainy season by development of the long-range flood forecasting system. The results of this study, it is also possible to take advantage of support premium determination of flood insurance, and as a material for notification of flood risk zone, and also it may be utilized to support the flood risk policy and to obtain the publicity effect for the public. The "Guidelines of district, city, and country group management plan" is to run the disaster vulnerability analysis of climate change every five years, since it should be reflected in the plan, if to perform every year the climate change disaster vulnerability analysis, it is possible to take advantage of the research results of long-range flood forecasting. In the future, extends of this study, it is necessary to clarify a mechanism over the Korean Peninsula between climate factors and hydrologic variables links with possible teleconnections in the large scale environments, and also it is possible to take advantage for the long-range flood forecasting during warm season to build a statistic-dynamic time series model using the APCC (APEC Climate Center) MME (Multi Model Ensemble) data.