초록 ▼

- 해양 기후변화 대응과 적응 방안 수립을 위하여 국지적 고해상도 모델 개발: 북태평양 전체 약 50km, 한반도 주변해 약 10km, 울산/광양만 최소 약 300m, 태화강 및 섬진강 하구역 최소 약 10m의 격자크기.
- Downscaling, CSEOF 분석 방법을 이용하여 IPCC AR4 A1B 이산화탄소 배출 시나리오에 따른 2050년 수층별 수온, 염분, 해수면, 해류 등 해양환경 변화 예측 수행.
- 예측 결과는 현 상태를 포함하여 해역별 격자별 DB 자료로 구축하여 manual과 함께 납품 - Non-Bou

- 해양 기후변화 대응과 적응 방안 수립을 위하여 국지적 고해상도 모델 개발: 북태평양 전체 약 50km, 한반도 주변해 약 10km, 울산/광양만 최소 약 300m, 태화강 및 섬진강 하구역 최소 약 10m의 격자크기.
- Downscaling, CSEOF 분석 방법을 이용하여 IPCC AR4 A1B 이산화탄소 배출 시나리오에 따른 2050년 수층별 수온, 염분, 해수면, 해류 등 해양환경 변화 예측 수행.
- 예측 결과는 현 상태를 포함하여 해역별 격자별 DB 자료로 구축하여 manual과 함께 납품 - Non-Boussinesq 모델 결과에 IPCC에 의한 육빙 녹음 기여분을 고려하여 예측된 2001~2050년 사이 동해, 황해, 동중국해 해수면 상승은 31.0cm, 31.8cm, 31.8cm. 장기자료 기반 이차함수와 지수함수적 증가를 가정한 통계적 해수면 상승은 한반도 주변해 평균 20.8cm, 43.8cm를 보여, 모델에 의한 예측값은 (31.5 cm) 두 값의 중간값을 보인다.
- 한반도 주변해 정밀 모델 결과에 의한 2001~2050년 표층 수온 상승폭은 동해 2.3℃, 황해 2.7℃, 동중국해 1.4℃.
- 울산 연안역과 광양만 해역의 2001년과 2050년 표층수온은 각각 1.5℃, 1.3℃ 증가하는 것으로 예측.
- 해수면 자료와 표층수온 자료를 이용하여 최초로 전 지구적 1900년부터 100년간 해수면 변동 재구성. 해수면 상승율은 1950~2009년 사이 1.92 mm/yr.

Abstract ▼

Ⅳ. Results
1. Development of ocean climate change scenario
1.1 Future projections of climate change in the North Pacific
Sea level rises in the North Pacific and the East Asian Regional Seas are projected for 2050 relative to 1995, based on the IPCC AR4 SRES A1B scenarios predicted by the t

Ⅳ. Results
1. Development of ocean climate change scenario
1.1 Future projections of climate change in the North Pacific
Sea level rises in the North Pacific and the East Asian Regional Seas are projected for 2050 relative to 1995, based on the IPCC AR4 SRES A1B scenarios predicted by the three different climate system models (GFDL Cm².1, ECHAM5-MPIOM and MIROC3.2(hires)). Multi-model ensemble means of relative sea level rise for 1995-2050 are predicted to be 15.46 cm (±6.89 cm) for the North Pacific, 15.88 cm (±6.87 cm) for the East Sea, 16.70 cm (±8.65 cm) for the Yellow Sea and 16.66 cm (±6.88 cm) for the East China Sea, respectively. Thus, the rates of mean sea level rise during the same period are estimated to be 2.81 mm/yr (±1.25 cm/yr) for the North Pacific, 2.89 mm/yr (±1.25 mm/yr) for the East Sea, 3.04 mm/yr (±1.57 mm/yr) for the Yellow Sea and 3.03 mm/yr (±1.25 mm/yr) for the East China Sea.
Assuming the linear sea level change for the estimation of land ice melting, mean sea level rises with land ice melting for 1995-2050 are 33.4 cm for North Pacific, 33.8 cm for East Sea, and 34.6 cm for Yellow and East China Seas. The corresponding rates of sea level rise are 6.1 mm/yr, 6.1 mm/yr and 6.3 mm/yr for the North Pacific, East Sea, and Yellow and East China Seas, respectively.
1.2 Future projections of climate change in the marginal seas around the Korean Peninsula
Regional climate modeling was performed in the seas around the Korean Peninsula based on the dynamic downscaling. Prior to the downscaling, water temperature from the reanalysis field obtained from a data assimilation model is analyzed. AR4 A1B (MIROC3.2 (hires), ECHAM5/MPI-OM, HadCM3) and AR5 RCP 4.5(Hadgem²_AO_kma), RCP6.0(GFDL-CM3) scenarios of the IPCC were selected for downscaling experiments. Climate change signals were calculated from the GCMs using CSEOF (Cyclostationary Empirical Orthogonal Function) analysis and added to the present lateral open boundaries and the surface boundary from 2001 to 2050. The surface warming trends calculated from the AR4 A1B, AR5 RCP 4.5, and AR5 RCP 6.0 downscaling models were 0.02–0.05 ℃/yr, 0.01–0.02 ℃/yr, and 0.02–0.04 ℃/yr. The warming rate decreases with depth with about 0.01℃/yr at 500 m. The effect of wave on the temperature change was examined by including mixing effect of wave. The simulation with wave effect results in deepening of the surface mixed layer but no significant difference in the temperature increase rate compared to the simulation without the wave effect.
2. Prediction of response of coastal areas and estuaries to changes in offshore forcing
2.1 Response of coastal areas
(1) High-resolution modeling offshore of Ulsan and in Gwangyang Bay based on nesting technique
In order to evaluate changes in the coastal environment due to changes in offshore forcing, coastal environments in Ulsan and Gwangyang areas are predicted quantitatively by dynamic downscaling based on results from the marginal sea projections
(2) Results of coastal prediction.
Results of the present (2001) and future (2050) states, which are 10-year-long annual means from 2001 to 2010 and from 2041 to 2050, respectively, are presented. The Ulsan area shows an increase in area-averaged surface temperature of about 1.5℃, and the Gwangyang area shows an increase of about 1.3℃.
2.2 Response of estuaries
(1) Future projections of esturaries by downscaling
For the estuarine change analysis, downscaling modeling is performed. Study areas are Gwangyang Bay and Ulsan Bay. Results of structured coastal model, EFDC and RIAMOM, are used to set the initial condition and time-varying data to unstructured model, FVCOM. According to the simulation results, the impacts of temperature increase and sea level rise mainly appear in the river. The result of Seomjin river area suggests that the surface temperature of the future projection (case for 50 cm of sea level rise in 2050) is higher about 0.1℃ than present surface temperature. In Taehwa river case, the surface temperature of the future projection (case for 50 cm of sea level rise in 2050) is higher by about 0.8~0.1℃ than present surface temperature. The upstream intrusion of salt wedge was strengthened about 2~4 km because of the sea level rise. These results indicated that downscaling model is important and essential for the prediction of the response of complex coastal areas to the climate change.
(2) Changes in stratification and salt wedge
In order to analyze stratification and salt wedge change of estuary, an accurate numerical simulation of coastal area and estuary is required. Therefore, a downscaling based on the high resolution data is required to produce accurate results. The result of simulation indicates that mixing by tide becomes weaker when sea level rises, and stratification and salt wedge intrusion become stronger. Not only the estuary but also the river stratification become strong due to 1 m sea level rise. Especially, well stratified river Potential Energy Anomaly increases to be 40% stronger, the salt wedge length grows about 3 km toward upstream. Such changes could result in the ecological and socio-economic change near estuary.
(3) Socio-economic scenario
Climate change adaptation framework consists of two parts: assessment of climate change impact and procedure for adaptation planning. Firstly, climate exposure and its impacts are identified. Secondly, providing five steps for planning procedure. Climate change prediction scenario that considers the impact of sea level rise and typhoon increase has very high uncertainty at present, and this study suggests specific adaptation options and adaptation planning methods for responding to the climate impact of study area.
3. Understanding the dynamics of ocean climate change
3.1 IPCC AR4/AR5 analysis
(1) Assessment of present climate simulation capability for the North Pacific Ocean
Both AR4 and AR5 models project that the MLD in the 21st century will decrease in most regions of the North Pacific, whereas the spatial pattern of the MLD is nearly unchanged. The overall shoaling seems to be related with intensified upper-ocean stratification caused by both surface warming and freshening. A significant MLD decrease (> 20 %) is found in the Kuroshio extension (KE), which is predominantly driven by reduced surface cooling, caused by weakening of wind. Conversely, the MLD increases in the narrow-banded region between 40 and 45.8˚N and west of 160.8˚W, just north of the KE front. This deepening appears to be largely driven by a northward shift of the KE.
We evaluated the performance of AR5 global climate models by comparing their historical Pacific Decadal Oscillation (PDO) simulations with observational data. Statistical analyses showed that the spatial patterns of PDO simulated by AR5 models were improved compared with those by AR4 models, mainly due to better representations of teleconnection between the tropics and mid-latitudes, as well as the improvement in precipitation sensitivity to tropical sea surface temperature.
(2) Analyses of variability, teleconnection, and future projections in the North Pacific Ocean
PDO variability of North Pacific consists of WPDO and CPDO, which are the western and eastern parts of PDO, respectively. It is found that the intensity of the WPDO mode remains unchanged, but CPDO would enhance in the future warmer climate state. This enhanced CPDO mode seems to be associated with teleconnection changes due to ENSO characteristic changes, with no clear connection with Aleutian Low variability. Many model support increased precipitation over central Pacific near equator in the future climate, which leads to enhanced stationary Rossby wave activity and increased variability of the western North subtropical high. It also leads enhanced CPDO mode.
(3) Variability of thermocline and the Intermediate Water distribution in the North Pacific Ocean
The first EOF mode of thermocline depth in the North Pacific is related to the North Equatorial Current bifurcation latitude. Comparison of distribution of chemical tracer (CFC) at the North Pacific Intermediate Water density range simulated by earth system model (CESM) with observation implies that the earth system model poorly simulates the transport of water from the subpolar region to the subtropical region.
(4) Energy budget change due to global warming
IPCC AR4 model datasets under the A1B scenario were used to investigate the change in the atmospheric and oceanic energy budgets in the tropical region due to global warming. CSEOF analysis was applied to extract the global warming signal. Then, other variables particularly those related to radiations were regressed onto the global warming signal. It was found that incoming energy is increased due to the reduction of cloud fraction. Further, longwave radiation exchange between the surface and the atmosphere has increased significantly; which is a positive feedback process between increased water vapor and warmer surface. Such changes affect meridional heat transport in the atmosphere and the oceans. It appears that energy transport toward higher-latitude decreases due to global warming.
(5) Analysis of future atmospheric and oceanic variability over the North Pacific
IPCC AR4 datasets based on the A1B scenario were analyzed to understand future change in oceanic and atmospheric variability over the North Pacific. The second CSEOF mode represents global warming and atmospheric temperature is expected to rise 4K by 2100. Ocean temperatures along the Kuroshio Current are also sensitive to global warming and sea level height increases (decreases) and anomalous anticyclonic (cyclonic) circulation develops where sea surface temperature increases (decreases).
(6) Origin of the two types of El Nino
CSEOF analysis was applied to both observational and model datasets to investigate two types of El Nino. The first CSEOF mode denotes the canonical El Nino, whereas the second CSEOF mode depicts a dipole-like structure in the tropical Pacific. A comparison of the index time series created from a 20-year moving average of NINO4 and NINO3 SST difference with the PC time series of the two modes via sliding correlation indicates that natural variability of the CP El Nino is primarily determined by the phase relationship of the two modes, which in turn is strongly connected with the long-term variability of the tropical Pacific.
3.2 Long-term ocean-atmosphere data analysis
(1) Sea level reconstruction
Sea level is important in understanding climate change and was extended back to 1950 based on tide gauge data. The rate of sea level rise is about 1.92 mm yr-1 between 1950 and 2009 and is 2.8 mm yr-1 between 1993 and 2009. By using sea surface temperature together with tide gauge, sea level reconstruction can be improved.
(2) Analysis and prediction of sea level
Analysis of sea level change during the recent 20 years, Pacific Decadal Oscillation (PDO) appears to contribute significantly to the long-term variation of the global average sea level change. PDO is an important source of sea level variability and influences strongly on the acceleration of sea level trend. By removing PDO variability, a more accurate estimate of sea level trend can be obtained. Change in the sea level rise in the equatorial Pacific remains to be significant even after removing the PDO contribution.
Based on the sea level reconstruction data, by 2060, the rate of sea level rise for 50 years from 2005 to 2055 is expected to be [5.28, 4.87, 6.18] mm/yr for the Yellow (124.5-126.5°E, 33.0-38.0°N), South (126.0-130.0°E, 33.0-35.0°N), and East (129.0-131.0°E, 35.0-38.0°N) Seas respectively. As a result, it is estimated that the sea level in 2055 will rise by [248, 229, 290] mm above the sea level in 2005. Mean sea level rise in the seas around the Korean Peninsula (125.0-140.0°E, 35.0-45.0°N) in 2055 is expected to be about 28 cm as compared to that in 2005, which is about 1.7 times faster that the global sea level rise of about 17 cm during the same period.
(3) Long-term variability of water temperature and upper ocean heat content
Analysis reveals that interannual variability of the Korea Strait cold water is strongly linked with anomalous meridional wind along the coast of the Korean peninsula. Further, warming signal was extracted from heat content in the upper 400 m in the East Sea, which is intimately linked with the warming pattern in the northwestern Pacific.
(4) Dynamics of oceanic and atmospheric interactions and their recent changes
The dynamical link between the tropical Pacific and North Pacific has changed around the late 1990s. The period of 1980-1998 is characterized by Eastern Pacific (EP) El Nino, and the period of 1999-2010 is characterized by the Central Pacific (CP) El Nino. The sea surface temperature change in the North Pacific during the CP El Nino resembles the North Pacific Oscillation (NPO) pattern. The atmospheric bridge mechanism is prevalent in the earlier period and the seasonal footprinting mechanism is prevalent in the latter period. These findings are consistent with earlier studies.
Analysis indicates pronounced reduction of sea ice in the Bering Sea in winter and in the Chukchi Sea in summer. The increased heat flux in the Bering Sea due to global warming induces the NPO atmospheric pattern and appears to explain increased connectivity between the tropical CP El Nino and the NPGO in the North Pacific.
CSEOF analysis was employed to separate the global warming mode, the low-frequency mode, and the biennial mode from the sea surface temperature in 1871-2010. The sum of these three modes explain the majority of tropical sea surface temperature variability. Based on these modes, it was found that the CP El Ninos in the early 1990s are due to the low-frequency mode, but those in the early 2000s are due to global warming; the tropical warming in the two periods has different physical mechanisms.
Due to the shift from EP El Nino to CP El Nino in the late 1990s, patterns of anomalous sea surface temperatures in the Southern Hemisphere have also changed significantly. The pattern of the Southern Annual Mode (SAM) and that of sea surface temperature has also changed around the 1990s. The relative dominance of the two mechanisms of sea surface temperature variability, i.e., SAM and El Nino, appear to explain a significant portion of sea surface temperature variability in the Southern Hemisphere.
(5) Oceanic-atmospheric interaction in association with the East Asian winter monsoon
East Asian winter monsoon was extracted from reanalysis data for the purpose of understanding long-term variability of the EAWM. With the 850-hPa temperature as the target variable, the first CSEOF mode represents the seasonal cycle and the second mode seasonal-scale interannual variability. Then, several physical variables including thermal advection and heat fluxes were analyzed in conjunction with the first two CSEOF modes. It was found that the impact of the EAWM is strong in the variation of surface temperature, circulation, thermal advection and turbulent heat flux particularly over the Kuroshio region (near the continent).
(6) Investigation of ocean-atmospheric interactions due to atmospheric Rossby waves
Rossby waves are prevalent atmospheric features observed in the upper troposphere in Northern Hemisphere. In order to understand the influence of Rossby waves on the lower troposphere and ocean surface, Rossby waves have been extracted from the 30-year winter daily potential vorticity in the upper troposphere. For positive Rossby waves, atmospheric column becomes colder and drier with a downward motion on the downstream side of Rossby waves. As a result, atmospheric column becomes less cloudy and net radiation increases. Latent and sensible heat fluxes also increase. Due to the increased turbulent heat flux, divergence and anticyclonic circulation strengthens in the lower troposphere.
(7) Comparison and evaluation of global reanalysis turbulent heat flux products in the northwestern Pacific
Latent heat fluxes (LHFs) are calculated with COARE3.0 algorithm based on meteorological variables acquired from eleven buoy stations over the northwestern Pacific marginal seas and the Kuroshio region. The calculated fluxes are compared with those archived from five reanalysis products (ERA-Interim, NCEP1, NCEP2, CFSR, MERRA) and an objectively analyzed data (OAflux) to validate their performance in the region considered. While temporal correlation between buoy LHFs and product LHFs are high, mean values and magnitudes of variation of product fluxes are overestimated or underestimated. Error analyses of mean LHFs show that the large mean biases and amplitudes of LHF are mainly caused by warm SST bias in the coastal regions. On the other hand, errors in wind speed and specific humidity mainly result in the large mean LHF errors in the Kuroshio region. In general, ERA-I and CFSR show relatively better performance in the study area.
(8) Correlation analysis between SST in the northwestern Pacific Ocean and climate indices of East Asian Winter Monsoon and North Pacific Oscillation
The East Asian winter monsoon (EAWM) and North Pacific Oscillation (NPO) constitute two outstanding surface atmospheric circulation patterns affecting the winter sea surface temperature (SST) variability in the western North Pacific. During the 1973-1987 strong winter monsoon epoch, the EAWM and NPO were significantly correlated to each other (-0.89), but their correlation practically vanished (-0.11) during the 1988-2002 weak winter monsoon epoch. Their impacts on the SST are also nonstationary and regime-dependent with a sudden change around 1988, suggesting major contribution to the interannual SST variability is changed from atmosphere (or air-sea interaction) to ocean dynamics, such as advection.
3.3 Observations of major processes relating to ocean-climate change
(1) Observations of ocean-atmosphere variables using ocean buoys and calculation of surface heat fluxes
Characteristics of ocean-atmosphere variables are analyzed and monthly mean surface heat fluxes are calculated using COARE 3.0 flux algorithm based on data from an ocean buoy (YSROB) deployed outside the summer tidal front in the southeastern Yellow Sea between July 2011 and October 2013 together with the Korea Meteorological Agency buoy (CB) located inside the front. The difference between sea surface temperature and air temperature is less than 1 ℃ from May to August and more than 3.8 ℃ from November to February, but the difference between saturation specific humidity by sea surface temperature and air specific humidity is less than 1.7 g/kg from January to July and more than 3.0 g/kg from August to October. The net shortwave radiation flux (Q₁) shows the maximum in May (237.7 W/m²) rather than summertime between June and September. Heat losses from sea surface to atmosphere occur through the net longwave radiation flux (Q_b), the sensible heat flux (Q_h), and the latent heat flux (Q_e) with the largest contribution from the Q_e of about 51 % of the total heat loss. The sea surface loses heat to atmosphere by Q_e throughout the year, and also through Q_h except during May and July when SST is lower than air temperature. The net heat flux (Q_n) shows maximum in May (144.3 W/m²) when Q_i is maximum, and minimum in December when all the heat flux components show their annual minimum.
Record-length mean air-sea variables at YSROB are characterized by 1.4 ℃ higher in air temperature, 2.6 ℃ higher in sea surface temperature, and 0.9 m/s stronger in surface wind as compared to those at CB. Surface pressure and humidity mean values are not significantly different at two station. Annual mean Q_n at YSROB is estimated to be 1.9 W/m², which is negligible considering instrumental errors in calculating Q_n(max. ±19 W/m²). On the other hand, the annual mean Q_n is 59 W/m² at CB, indicating significant heating of ocean surface by air-sea interaction. The implication of the net heat gain at CB deserves future study.
The differences between common-length mean Q_e from reanalysis products (MERRA and ERA-Interim) and estimated mean Q_e at YSROB are small, -6.1 W/m² for MERRA and –1.9 W/m² for ERA-Interim. On the other hand, the products yield large biases of Q_e of about 30 W/m² at CB due to their coarse horizontal resolutions not being able to resolve the contrast of air-sea variables in the local frontal region.
(2) Investigation of the importance of double diffusive vertical mixing
Based on time series measurements of ocean’s microstructures at 0°N and 156°E in 2012 when the ENSO state was normal, turbulent dissipation rates of kinetic energy and temperature variance are calculated, and vertical diffusivities and vertical fluxes of temperature, salt, and buoyancy are estimated using turbulence and double diffusion models. Double diffusive mixing is considered because turbulence level was low in the thermocline possibly due to the development of strong upper stratification caused by an eastward extension of fresh and warm surface water to the study area which typically happens usually during the El Nino period. The observations indicate that salt fingering dominates shear-driven turbulence below 100 m. Vertical heat and salt fluxes due to double diffusion are 10^-6~10^-7 K ms^-1, 10^-5~10^-6 psu ms^-1, which are about 10―100 times larger than those from the turbulence model. Vertical density flux due to double diffusion is downward to enhance stratification with a magnitude of about 10^-7 m²s^-1. Vertical diffusivity of density due to double diffusion is 10^-5~10^-6 m²s^-1, which is comparable to the estimated diffusivity when the turbulent dissipation rate is 10^-8~10^-9 m²s^-3 at the equator. This study shows, for the first time, the potentially large contribution of double diffusion to vertical mixing when turbulence level is low in the western tropical Pacific. Under weak turbulence conditions, the turbulence model significantly underestimates the magnitude of the vertical buoyancy flux with the flux having the wrong sign, and fluxes induced by double diffusive mixing need to be considered. The results suggest a consideration of double diffusion in the parameterization of vertical mixing in numerical models used for future climate projections. Vertical structure of water properties in the East Sea shows a favorable condition for double diffusion with feeble deep velocity, hence double mixing may have to be considered in future projections of the East Sea using numerical models.
4. Construction ocean climate change scenarios DB and data visualization
4.1 DB and visualization
- Model-predicted data on sea level, and temperature/salinity/current in multiple vertical levels at the present state and in 2050 together with their differences were collected and put into DB
- Visualization of the products in areas of the North Pacific, the seas around the Korean Peninsula, coastal areas, and estuaries
- DB consists of two classes, ‘General users’ and ‘Professional users’, to provide more user-friendly web service for each group.
- Preparation of a manual for the use of DB
5. Korea-China cooperative research in the field of climate change
5.1 Holding Korea-China workshop
- Date: 24 June ~ 26 June, 2014
- Venue: Ramada Plaza Jeju Hotel
- Objective: Materialization of collaborative research and initial preparation of related topics on ocean climate change between Korea and China
- Participants: 72 scientists and students from Korea and China
- Presentations: 26 scientific oral presentations and 22 proposals for cooperative research
5.2 Integrated cooperative research topics between two countries in the field of ocean climate change
- The role and response of the Yellow Sea and the East China Sea to regional climate change
- Origin, distribution, and dynamics of fine-grained deposits in the Yellow Sea and East China Sea with regard to climate changes
- Ecosystem dynamics their response to natural and anthropogenic forcings in the Yellow Sea and the East China Sea