Sea surface temperature (SST) predictability in the Pacific on decadal timescales is examined in hindcast experiments using the coupled atmosphere-ocean model MIROC with low, medium, and high resolutions. In these hindcast experiments, initial conditions are obtained from an anomaly assimilation procedure using the observed oceanic temperature and salinity while pre¬ scribing natural and anthropogenic forcing based on the IPCC emission scenarios. Our hindcast experiments show the predictability of SST in the western subtropical Pacific, the Indian Ocean, and the tropics to the North Atlantic. Previous studies have examined the SST predictability in the Indian Ocean and the Atlantic, but SST predictability in the western sub-tropical Pacific has not been evaluated. In the western Pacific, the observed SST anomalies in the subtropics of both hemispheres increased rapidly from the early 1990s to the early 2000s. While this SST warming in the western subtropical Pacific is partly explained by global warming signals, the predic¬tions of our model initialized in 1995 or 1996 tend to simulate the pattern of the SST increase and the associated precipitation changes. This large climate change around the late 1990s may be related to phenomena such as the recent increase in the typhoon frequency in Taiwan and the weakened East Asian monsoon reported by recent studies.
18.6-year period variability has been detected in various aspects of the climate, especially in and around the Pacific Ocean. Although it is believed to be caused by 18.6-year period tidal cycle, no study has directly shown how the tidal cycle regulates such variability. Using a state-of-the-art climate model, we show that the 18.6-year cycle in strong tidal mixing localized around the Kuril Islands induces 18.6-year periodicity in El Nino-Southern Oscillation-like Pacific climate variability. Influence of the tidal mixing propagates along the Pacific western rim as coastally trapped waves. Temperature anomaly is generated in the subsurface western equatorial Pacific, which propagates along the equatorial thermocline and eventually excites 18.6-year periodicity in the equatorial sea surface temperature.
Decadal modulation of El Niño/Southern Oscillation (ENSO) amplitude is reproduced by a long-term simulation with an atmosphere-ocean general circulation model. A frequency of the modulation is in phase with Pacific decadal variability (PDV). A detailed analysis of the budget of sea surface temperature (SST) shows that effect of a change in anomalous zonal advection of mean temperature correlated to ENSO development has a dominant role in low-frequency modulation of ENSO. That is because an equatorial zonal SST gradient changes periodically in the PDV cycle as a change in the background mean state for ENSO. Such characteristics are consistent with an observed climate shift and changes in ENSO characteristics that occurred in mid-1970s.
As reported in former studies, temperature observations obtained by expendablebathythermographs (XBTs) and mechanical bathythermographs (MBTs) appear tohave positive biases as much as they affect major climate signals. These biases havenot been fully taken into account in previous ocean temperature analyses, which havebeen widely used to detect global warming signals in the oceans. This report proposesa methodology for directly eliminating the biases from the XBT and MBT observations.In the case of XBT observation, assuming that the positive temperature biasesmainly originate from greater depths given by conventional XBT fall-rate equationsthan the truth, a depth bias equation is constructed by fitting depth differences betweenXBT data and more accurate oceanographic observations to a linear equationof elapsed time. Such depth bias equations are introduced separately for each yearand for each probe type. Uncertainty in the gradient of the linear equation is evaluatedusing a non-parametric test. The typical depth bias is +10 m at 700 m depth onaverage, which is probably caused by various indeterminable sources of error in theXBT observations as well as a lack of representativeness in the fall-rate equationsadopted so far. Depth biases in MBT are fitted to quadratic equations of depth in asimilar manner to the XBT method. Correcting the historical XBT and MBT depthbiases by these equations allows a historical ocean temperature analysis to be conducted.In comparison with the previous temperature analysis, large differences arefound in the present analysis as follows: the duration of large ocean heat content inthe 1970s shortens dramatically, and recent ocean cooling becomes insignificant. Theresult is also in better agreement with tide gauge observations.
Sea ice has a large impact on climatic system and its variability. A good reproducibility of the past state of the sea ice in global climate models will reduce uncertainty in future projection. Here, we present sea-ice simulations for new versions of atmosphere-ocean coupled general circulation models, the Model for Interdisciplinary Research on Climate version 4h (MIROC4h) and version 5 (MIROC5), and assess the reproducibility of the sea ice prior to the future projection. The horizontal resolution of MIROC4h is signifi-cantly high for a coupled climate model, although its sea-ice component is based on the previous version. MIROC5 employs some improved schemes including subgrid-scale ice thickness distribution. Hindcast simulations of twentieth-century climate by the new models are compared with observations and with the results of previous versions of MIROC. For the Northern Hemisphere, Arctic sea-ice simulations are improved in both MIROC4h and MIROC5 compared with previous mod-els. MIROC5 generally agrees well with observational data, whereas in MIROC4h, the Arctic sea ice is smaller in summer extent and in thickness.Employment of the ice thickness distribution, which allows large heat exchange through subgrid-scale thin ice regardless of the grid-averaged thickness, and relatively high albedo parameterscontribute to reproduce more realistic ice thickness in MIROC5 compared with that in MIROC4h. For the Southern Hemisphere, MIROC4h well reproduces the observed ice edge, especially in winter, while MIROC5 underestimates sea-ice extent. Both models indicate decreasing trends in Arctic sea ice in the late twentieth century. A heat budget analysis of theMIROC5 Arctic Ocean suggests that intensification of ice-albedo feedbackaccelerates the rate of Arctic ice decline.
In line with the experimental design for near-term climate prediction toward the 5th Assessment Report of the Intergovernmental Panel on Climate Change (IPCC-AR5) and the Coupled Model Intercomparison Project Phase 5 (CMIP5), we perform ensembles of initialized decadal hindcast experiments using two recent versions of the Model for Interdisciplinary Research On Climate (MIROC): MIROC4h (T213L56 AGCM and 1/6-1/4 deg. 48 level OGCM) and MIROC5 (T85L40 AGCM and 0.56-1.4 deg. 50 level OGCM). We analyze sets of 10-year-long 9-ensemble hindcasts (3 members by MIROC4h and 6 members by MIROC5) with initialization every five years after 1961 and explore the predictability of decadal climate changes.The most predictable variation on decadal timescales is the global warming signal due to the favorable response of the models to external forcing. The results of these initialized hindcast experiments using MIROC5 validate our ability to enhance decadal predictability primarily through the initialization, particularly of the Pacific Decadal Oscillation (PDO) for a few years and of the Atlantic Multidecadal Oscillation (AMO) for almost a decade. The initialization has large impacts on the upper ocean temperature hindcasts over the mid- and high latitudes of the North Pacific and the high latitudes of the North Atlantic, where the PDO and AMO signals are observed to be strongest. In contribution to process and assessment studies in IPCC-AR5 and CMIP5, further analysis of our hindcast data (and near-term prediction data) using MIROC4h and MIROC5 is worthwhile. We note that the initialized hindcasts using MIROC4h have predictive skill inferior to the MIROC5 results and that at this stage, fully significant discussions may not be possible due to the small number of ensembles with limited computational resources.
Decadal-scale climate variations over the Pacific Ocean and its surroundings are strongly related to the so-called Pacific decadal oscillation (PDO) which is coherent with wintertime climate over North America and Asian monsoon, and have important impacts on marine ecosystems and fisheries. In a near-term climate prediction covering the period up to 2030, we require knowledge of the future state of internal variations in the climate system such as the PDO as well as the global warming signal. We perform sets of ensemble hindcast and forecast experiments using a coupled atmosphere-ocean climate model to examine the predictability of internal variations on decadal timescales, in addition to the response to external forcing due to changes in concentrations of greenhouse gases and aerosols, volcanic activity, and solar cycle variations. Our results highlight that an initialization of the upper-ocean state using historical observations is effective for successful hindcasts of the PDO and has a great impact on future predictions. Ensemble hindcasts for the 20th century demonstrate a predictive skill in the upper-ocean temperature over almost a decade, particularly around the Kuroshio-Oyashio extension (KOE) and subtropical oceanic frontal regions where the PDO signals are observed strongest. A negative tendency of the predicted PDO phase in the coming decade will enhance the rising trend in surface air-temperature (SAT) over east Asia and over the KOE region, and suppress it along the west coasts of North and South America and over the equatorial Pacific. This suppression will contribute to a slowing down of the global-mean SAT rise.
A new high-resolution atmosphere-ocean coupled general circulation model named MI-ROC4h has been developed, and its performance in a control experiment under conditions present in 1950 is examined and compared with simulations of preindustrial conditions carried out by the Intergovernmental Panel on Climate Change Fourth Assessment Report (IPCC AR4) using the previous high- and medium-resolution versions of the model, called MIROC3h and MIROC3m, respectively. A major change in MIROC4h is a doubling of the resolution of the atmospheric component to 0.5625º, compared to 1.125º for MIROC3h. The oceanic compo-nents of MIROC4h and MIROC3h are eddy-permitting, with a horizontal resolution of 0.28125° (zonal) × 0.1875° (meridional). In MIROC3m, the horizontal resolution is 1.40625° (zonal) × 0.56° -1.4° (meridional). Compared with MIROC3h and MIROC3m, many improvements have been achieved; for example, errors in the surface air temperature and sea surface temperature are smaller, there is less drift of the ocean water temperature in the subsurface-deep ocean, and the frequency of heavy rain is comparable to observations. The fine horizontal resolution in the atmosphere makes orographic wind and its effects on the ocean more realistic than those of the former models, and the treatment of coastal upwelling motion in the ocean has been improved. Phenomena in the atmosphere and ocean related to the El Niño and southern oscillation are now closer to observations than was obtained by MIROC3h and MIROC3m. The effective climate sensitivity for CO2 doubling is calculated to be about 5.7 K, which is much larger than the value obtained using the IPCC AR4 models, and this is mainly due to a decrease in the low-level clouds at low latitudes.
Initialization based on data assimilations using historical observations possibly improves near-term climate predictions. Significant volcanic activity in the future is unpredictable and not assumed in future climate predictions. To examine the possible influence of unpredictable future volcanic activity on the decadal potential predictability of the natural variability, we performed a 2006-2035 climate prediction experiment with the assumption that the 1991 Mt. Pinatubo eruption would take place again in 2010. The Pinatubo forcing induced not only significant cooling responses but also considerable noises in the natural variability. The errors due to the Pinatubo forcing grew faster than that arising from imperfect knowledge of the observed state, leading to a rapid reduction of the decadal potential predictability of the natural variability.
This is the first study to examine whether human contributions to changes in extreme temperature indices have larger amplitudes than natural variability in near future (up to 2030) climate prediction. We performed 10 runs of the initial condition perturbed ensemble of a coupled atmosphere-ocean general circulation model. In the near future, over most land areas, all 10 runs predict more frequent occurrences of warm nights and warm days, and less frequent cold nights and cold days, suggesting that human influences have become larger than natural variability. The fraction of areas where all runs agree on the direction of changes over land is less sensitive to ensemble sizes (for warm nights, 96% and 93% for 4 runs and 10 runs, respectively). The changes in the frequency of warm and cold extremes are mainly due to shifts in seasonal mean temperatures. Additionally snow cover affects the frequency of cold extremes in some areas.
We show the near-term predictions of mean and extreme precipitation up to the year 2030 by analyzing ten-member ensemble runs with perturbed initial conditions of the MIROC model. Mean and extreme precipitation increase in high latitudes and the tropics, and decrease in the subtropics in the ensemble mean. Uncertainty due to natural variability was also examined. Most of the ten runs predict positive anomalies in high latitudes and some parts of the tropics. Changes in parts of the subtropics are uncertain due to natural variability. Thermodynamic changes mainly explain robust total increases in mean precipitation in high latitudes and the tropics. Thermodynamic changes of mean precipitation are uncertain in the subtropics, resulting in a large uncertainty in total changes. It is suggested that small signal-to-noise rations of thermodynamic changes in the subtropics are induced by regional decreases in relative humidity at lower troposphere, which counteract the effects of increased column-integrated water vapor.
A Predictable Component Analysis (PrCA) has been applied to near-term forecasts of ocean-atmosphere natural variability within a global warming scenario. The most predictable pattern in the North Pacific region on decadal time scales for the period 1965-1999 is a clockwise movement of heat content anomalies, which is composed of the eastward propagation along the Kuroshio-Oyashio extension (hereafter KOE) region and the subsequent southwestward propagation along the subduction pathway through the subtropical region. Our results also suggest the important role played by the westward traveling Rossby wave north of the KOE region in determining the phase change of the Pacific Decadal Oscillation (PDO) during this period. A comparative study with different initial conditions demonstrates the importance of proper thermal constraints on the ocean subsurface, in conjunction with the Rossby wave dynamics, for the initiation of predictable patterns in the heat content anomaly.
This paper documents the procedure of ocean data assimilation that initializes the climate models MIROC3m, MIROC4h, and MIROC5 for decadal climate predictions following the CMIP5 protocol, and summarizes the performance of the climate models using this data assimilation. Only anomalies of observed ocean hydrographic data are assimilated using the incremental analysis update method in order to prevent model climate drifts during predictions. In the case of MIROC4h, which has an eddy-permitting ocean model, a spatial smoother is used in calculating analysis increments so that oceanic mesoscale eddies can be free from observational constraints and that they are generated and decay physically in response to the assimilated background state. Globally, the decadal-scale variations of ocean temperatures in the assimilation runs are highly correlated with the observations. Variations of surface air temperature over oceans are also consistent with the observations, but this is not the case in some regions over continents. Atmospheric responses to the SST variations corresponding to the Pacific Decadal Oscillations (PDO) and the Atlantic Multi-decadal Oscillation (AMO) are better represented in MIROC4h and MIROC5 than in MIROC3m. The high resolution of MIROC4h and new cloud parameterizations in MIROC5 may contribute to this improvement. Root-mean-squared amplitudes of sea surface height variations associated with oceanic eddies (hereafter, eddy activity) are not suppressed undesirably in the MIROC4h assimilation run and these are comparable with those in the uninitialized runs. In the Kuroshio-Oyashio confluence zone, eddy activity is modulated on a decadal timescale. This modulation is reasonably represented in the assimilation run compared with the observations. In the hindcast experiments, significant decadal prediction skills are found for the North Atlantic, the subtropical North Pacific, and the Indian Ocean. The decadal climate predictions are expected to contribute to the IPCC AR5 and political decision-making for the coming decades.
We have investigated the effects of assimilating sea iceconcentration (SIC) data on a simulation of Arctic Ocean climateusing an atmosphere-ocean-sea ice coupled model. Our resultsshow that the normal overestimation of summertime SIC in theEast Siberian Sea and the Beaufort Sea in simulations withoutsea-ice data input can be greatly reduced by assimilating seaicedata and that this improvement is also evident in a followinghindcast experiment for 3-4 years after the initialization of theassimilation. In the hindcast experiment, enhanced heat storage inboth sea ice and in the ocean surface layer plays a central role inimproving the accuracy of the sea ice distribution, particularly insummer. Our detailed investigation suggests that the ice-albedofeedback and the feedback associated with the atmospheric pressurepattern generated by the improved estimation of SIC workmore effectively to retain the heat signal after initialization for acoupled atmosphere-ocean-sea ice system prediction. In addition,comparison with field observations confirms that the model failsto produce a realistic feedback loop, which is (presumably) dueto inadequacies in both the ice-cloud feedback model and thefeedback via the Beaufort Gyre circulation. Further developmentof coupled models is thus required to better define Arctic Oceanclimate processes and to improve the accuracy of their predictions.
The influence of the expendable bathythermograph (XBT) depth bias correction on decadal climate prediction is presented by using a coupled atmosphere-ocean general circulation model called MIROC3. The global mean subsurface ocean temperatures that were simulated by the model with the prescribed anthropogenic and natural forcing are consistent with bias-corrected observations from the mid-1960s onward, but not with uncorrected observations. The latter is reflected by biases in subsurface ocean temperatures, particularly along thermoclines in the tropics and subtropics. When the correction is not applied to XBT observations, these biases are retained in data assimilation results for the model's initial conditions. Hindcasting past Pacific Decadal Oscillations (PDOs) is more successful in the experiment with the bias-corrected observations than that without the correction. Improvement of skill in predicting five-year mean vertically average ocean subsurface temperature is also seen in the tropical and the central North Pacific where PDO-related signals appear large.