We study climate change projections over East Asia under various representative concentration pathway (RCP) scenarios using simulations performed with the Beijing Climate Center Climate System Model version 1.1 (BCC_CSM1.1) for the Coupled Model Intercomparison Project phase 5 (CMIP5). Under all RCPs (including RCP2.6, RCP4.5, RCP6.0, and RCP8.5), East Asian climate is found to be warmer and wetter in the 21st century than the present climatology (1986-2005). For 2080-2099, East Asian mean surface air temperature (precipitation) is higher than that for the present climatology by 0.98°C (4.4%) under RCP2.6, 1.89°C (7.7%) under RCP4.5, 2.47°C (7.1%) under RCP6.0, and 4.06°C (9.1%) under RCP8.5. Such changes in East Asia are all larger than the corresponding global changes, with greater differences under the higher RCPs. In the simulation of RCP4.5, which is extended to the late 23rd century (2280-2299), further warming of 0.34°C relative to 2080-2099 is found in East Asia; this is lower than the global mean warming (0.56°C). Under mitigation scenario RCP2.6, East Asia experiences greater cooling than that experienced globally throughout the 22nd and 23rd centuries. In the late 21st century, East Asian summer mean precipitation increases prominently (by 10-15%) with respect to the present climatology under all RCPs. This increase in precipitation occurs primarily in southern China and northern East Asia and is associated with anomalous southerly flow in the lower troposphere. Drought occurs in the Yangtze River valley for the middle and high RCPs; we attribute this to anomalous subsidence around 30°N associated with the northward shift of the East Asian jet stream. Under the high scenario (RCP8.5), the western Pacific subtropical high extends westward and northward to southern China and is partly responsible for the deficient precipitation in regions to the south of the Yangtze River.
In line with the experimental design for near-term climate prediction toward the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, we perform ensembles of initialized decadal hindcast and near-future projection using three versions of the coupled atmosphere-ocean model MIROC. In the present study, we explore interannual and multiyear predictability of tropical cyclone (TC) activity in the western North Pacific (WNP) using the initialized hindcasts and examine global warming impacts on TC activity in the near-future on the basis of near-future projections up to 2035. The hindcasts of year-to-year variation in TC number capture the observed values reasonably well. Moreover, interannual variability of TC genesis and occurrence frequency associated with the El Niño Southern Oscillation are found to be predictable, mainly through better prediction of sea surface temperature (SST) and large-scale vorticity anomalies in the lower troposphere. These results indicate that the models can reproduce the major basic mechanisms that link TC genesis with large-scale circulation. Skillful prediction of TC number is likely difficult on multiyear timescales, at least based on our hindcasts, but through initializations, the three-year-mean hindcasts from 1998 onward reasonably capture observed major characteristics of TC activity associated with Pacific climate shift during the late 1990s. Near-future projections (2016-2035) suggest significant reductions (approximately 14%) in TC number, particularly over the western part of the WNP, even under scenarios in which projected global warming is less prominent than that at the end of this century. This reduction is likely due to the suppression of large-scale lower tropospheric vorticity and relative humidity and the enhancement of vertical wind shear. The projected SST exhibits a more pronounced warming over the eastern tropical Pacific than over the western region and accompanies the weakening of Walker circulation via redistribution of tropical convection activity, which appears to be responsible for the change in the large-scale fields in the WNP.
In this work, various methods for the estimation of the parameter uncertainty and the covariance between the parameters and the state variables are investigated using the local ensemble transform Kalman filter (LETKF). Two methods are compared for the estimation of the covariances between the state variables and the parameters: one using a single ensemble for the simultaneous estimation of model state and parameters, and the other using two separate ensembles; for the initial conditions and for the parameters. It is found that the method which uses two ensembles produces a more accurate representation of the covariances between observed variables and parameters, although this does not produce an improvement of the parameter or state estimation. The experiments show that the former method with a single ensemble is more efficient and produces results as accurate as the ones obtained with the two separate ensembles method. The impact of parameter ensemble spread upon the parameter estimation and its associated analysis is also investigated. A new approach to the optimization of the estimated parameter ensemble spread (EPES) is proposed in this work. This approach preserves the structure of the analysis error covariance matrix of the augmented state vector. Results indicate that the new approach determines the value of the parameter ensemble spread that produces the lowest errors in the analysis and in the estimated parameters. A simple low-resolution atmospheric general circulation model known as SPEEDY is used for the evaluation of the different parameter estimation techniques.
In this study, a regional climate model (WRF-ARW; the Advanced Research Weather Research and Forecasting model) having a resolution of 4.5 km was used to examine the sensitivity of precipitation on the Japan Sea side of Japan to the sea surface temperature (SST) in the Japan Sea during winter. We performed a control simulation (CTL) driven by reanalysis and observational SST datasets. Three sensitivity experiments in which SSTs over the entire domain were 1K, 2K, and − 1K different from the CTL SST were conducted to examine the sensitivity of precipitation to SST. The calculated precipitation on the Japan Sea side increased by 6-12% K−1 of SST warming. Concurrent with the precipitation changes, latent heat flux over the Japan Sea increased by 11-14% K−1 of SST warming. Because the changes in surface relative humidity were very small, the increase can be explained by the Clausius-Clapeyron equation. The deviation from the 7% increase in latent heat flux calculated from this equation can be quantitatively explained by the development of the planetary boundary layer over the Japan Sea, which was related to an increase in sensible heat flux due to the SST warming. This result also implies that the 1 K uncertainty in simulated and projected SST over the Japan Sea among multiple atmosphere-ocean global climate models corresponds to an approximately 10% uncertainty in precipitation on the Japan Sea side of Japan.
Influence of a distantly located typhoon (TY), a weak westerly trough, and orography on the intense rainfall on September 14-15, 1965 over Gifu and Fukui Prefectures (~36°N, 137°E) in central Japan is studied by using European Center for Medium-Range Weather Forecasts 40-year reanalysis data and upper, surface, and local rain-gauge observation data. Prior to the intense rainfall, a long arc-shaped moisture front was intensified on September 13 along the northern periphery of the outer circulation of TY Tix (6524), which was distantly located from Japan. A rainfall of 50-100 mm d−1 occurred in a long rainbelt along the Pacific coast of Japan, associated with the low-level convergence in the moisture front. A small depression formed over the Japan Sea on September 14 under the influence of a weak westerly shortwave trough in the middle troposphere. In the lower troposphere, a Λ-shaped trough extended from the outer circulation of the TY toward the depression and the moisture front protruded locally northward to the depression. Concurrently, with the protrusion of the moisture front, a meso-α-scale (~500 km) intense rainfall area protruded northward. In the meso-α-scale intense rain area, a south-north oriented narrow rainband formed along the east side of Suzuka-Ibuki Mountains on September 14. A maximum precipitation of ~800 mm d−1 occurred in the rainband. This case may fall under “intense rainfall remotely affected by TY.” However, the present case is different from other cases studied in previous articles in that very intense precipitation occurred in a narrow south-north oriented rainband formed along mountains under the influence of the weak westerly shortwave trough.
A database is established that includes microphysical properties of raindrops, cloud droplets, ice nuclei (IN), snow crystals, and the relationship between radar reflectivity (Z) and rainfall rate (R), based on the in situ data reported in the existing literature. The data coverage is divided into East Asia and the other regions (Americas, Europe, Australia, and Africa), and only the datasets obtained over land are considered. The main properties of microphysical variables over East Asia are presented and compared with those over the other regions. The main differences of the properties between those two regions are as follows. The average intercept (N0) of exponential-size distributions is much smaller over East Asia than that over the other regions and slope (λ) is slightly smaller. As for gamma-size distributions, the overall average value of intercept is much smaller over East Asia, and the range of parameters is narrower in East Asia compared with the other regions. In addition, most of the shape parameter γ are negative over East Asia, whereas positive γ appears frequently in the other regions. Compared with the other regions, the average cloud droplet concentration is much smaller in stratiform clouds over East Asia, and cloud liquid water content in East Asia is 0.114 g m−3 lower. There is a higher total mean IN concentration of 20.2 L−1 in East Asia, whereas it is 2.7 L−1 in the other regions. Compared with the other regions, the average N0 for snow-crystal-size distributions over East Asia is much smaller and λ is less than half of that. The Z-R relation shows that the average Z is larger in East Asia than that in the other regions for rains originating from convective clouds with the same R, whereas the average Z is slightly smaller in East Asia for rains originating from stratiform clouds.
Band-shaped precipitation systems are frequently observed as back-building (BB) types with most convective cells successively initiating on the upwind side. Usually, nearly neutral stratifications of moist static energy (h*) are observed in the convection areas, whereas conditionally unstable stratifications remain in the surroundings. To understand the difference in features between these two areas and the frequent occurrence of the BB type, heating, whose types are (i) a point source, (ii) an infinitely extending line, and (iii) a line with an end, is prescribed in environmental fields with uniform zonal wind (U*) and damping (γ*), and the linear responses of buoyancy are studied. The response fields for heating types (ii) and (iii) are obtained by applying a solution of buoyancy for (i) heating. Here it is assumed that the neutralization of h* occurs by temperature adjustment with a deviation having an intensity of ΔT*. In the U*= γ*=0 case, the temperature deviations are ΔT*/2 at most in the influence region far from the end, because two oppositely propagating gravity waves are induced for the response. This solution corresponds to response one on the Cartesian coordinate system. It is concluded that the neutralization of h* in the convection areas and its response in the surroundings induce the different stratifications, maintaining the conditionally unstable stratifications in the surroundings even though intense precipitation occurs. In the U*≠0 cases, frequent formations of the BB types are anticipated as one possibility, because the influence of the response on the upwind side of precipitation is small, and new convection is likely to occur.
Algorithms for estimating precipitation rates from spaceborne radar observations of apparent radar reflectivity depend on attenuation correction procedures. The algorithm suite for the Ku-band precipitation radar aboard the Tropical Rainfall Measuring Mission satellite is one such example. The well-known problem of nonuniform beam filling is a source of error in the estimates, especially in regions where intense deep convection occurs. The error is caused by unresolved horizontal variability in precipitation characteristics such as specific attenuation, rain rate, and effective reflectivity factor. This paper proposes the use of vertical decorrelation for correcting the nonuniform beam filling error developed under the assumption of a perfect vertical correlation. Empirical tests conducted using ground-based radar observations in the current simulation study show that decorrelation effects are evident in tilted convective cells. However, the problem of obtaining reasonable estimates of a governing parameter from the satellite data remains unresolved.
An operational real-time adjusted radar-automatic weather station (AWS) rain rate (RAD-RAR) system, using 10 radars of the Korea Meteorological Administration (KMA), has been developed for the South Korea region. The procedure of the RAD-RAR system consists of four steps for real-time operation: 1) the quality control of volumetric reflectivity for each radar, 2) the computation of the rain-gauge rain rate every 10 min. within each radar, 3) the real-time (updated every 10 min.) rainfall estimation by the Z-R relationship minimizing the difference between the 10-min constant altitude plan position indicator and rain-gauge rain rate based on window probability matching method (WPMM) and by the real-time bias correction of RAD-RAR conducted at 10 min. intervals for each radar by making the bias, and 4) the composition of estimated rainfall data of the 10 radars. It is noted that this RAD-RAR system is available only for summer rainfall cases with the absence of bright band around 1.5 km in height, as the system does not include bright band correction. The performance of RAD-RAR was examined for the 10 heavy rainfall events of 2006, and we obtained results suggesting that the real-time Z-R adjustment of RAD-RAR is better in terms of the agreement with the rain-gauge rain rate than that of the previously fixed Z-R relationship, and the additional bias correction of RAD-RAR yields slightly better results. A square of correlation coefficient R2=0.84 was obtained between the daily accumulated observed and RAD-RAR estimated rainfall.