Bayesian regression model for seasonal forecast of precipitation over Korea

In this paper, we apply three different Bayesian methods to the seasonal forecasting of the precipitation in a region around Korea (32.5°N?C42.5°N, 122.5°E-132.5°E). We focus on the precipitation of summer season (June?CJuly?CAugust; JJA) for the period of 1979?C2007 using the precipitation produced by the Global Data Assimilation and Prediction System (GDAPS) as predictors. Through cross-validation, we demonstrate improvement for seasonal forecast of precipitation in terms of root mean squared error (RMSE) and linear error in probability space score (LEPS). The proposed methods yield RMSE of 1.09 and LEPS of 0.31 between the predicted and observed precipitations, while the prediction using GDAPS output only produces RMSE of 1.20 and LEPS of 0.33 for CPC Merged Analyzed Precipitation (CMAP) data. For station-measured precipitation data, the RMSE and LEPS of the proposed Bayesian methods are 0.53 and 0.29, while GDAPS output is 0.66 and 0.33, respectively. The methods seem to capture the spatial pattern of the observed precipitation. The Bayesian paradigm incorporates the model uncertainty as an integral part of modeling in a natural way. We provide a probabilistic forecast integrating model uncertainty.     

Changes in future precipitation over South Korea using a global high-resolution climate model

We used daily precipitation data from a global high-resolution climate scenario to analyze the features of future precipitation including extreme and heavy rainfall. The scenario shows that the model reproduces the daily precipitation over South Korea well. The projections show an increase in annual precipitation of approximately 18% in the late 21st century, with the highest increase (38%) occurring in winter. The number of days with daily precipitation of less than 5 mm decreases, but that of daily precipitation of more than 5 mm increases slightly in the latter part of the 21st century. The peak of precipitation days shifted from July to August. The number of days with relatively small amounts of precipitation (10 and 30 mm d^?1) increases most substantially in the winter season, but that for large amounts of precipitation (50, 80, 100, and 130 mm d^?1) increases most in the summer season. Events with heavy precipitation rates of 100 and 130 mm d^?1 are expected to occur in the winter season in the late 21st century, although no such events occurred during the winter season in the reference period.     

Statistical multisite simulations of summertime precipitation over South Korea and its future change based on observational data

In this study, a weather generator for summer (May 19 – September 15) precipitation over South Korea is developed. Precipitation data for 33 years (1979–2011) observed at 57 stations of Korea Meteorological Administration (KMA) are used to develop a new weather generator. Using the cyclostationary empirical orthogonal function (CSEOF) technique, the observed precipitation data is described as a linear combination of deterministic evolution patterns and corresponding stochastic amplitude (principal component) time series. An autoregressive-moving average (ARMA) model is used to generate one hundred sets of synthetic amplitude time series for the period of 1979–2061 (83 years) with similar statistical properties of the original amplitude time series. Based on these synthetic time series and the annually repeating evolution patterns, one hundred sets of synthetic summer precipitation were generated. Statistical characteristics of the synthetic datasets are examined in comparison with those of the KMA observational record for the period of the observational record. Characteristic changes of synthetic precipitations for a future period are also examined. The seasonal cycle in the synthetic precipitation is reproduced faithfully with typical bimodal peaks of summer precipitation. The spatial correlation patterns of the synthetic precipitation are fairly similar to that of the observational data. The frequency-intensity relationship of the synthetic precipitation also looks similar to that of the observational data. In the future period, precipitation amount increases except in the precipitation range of (0,10) mm day^?1 with nearly no change in the frequency of no-rain days; frequency increase is particularly conspicuous in the range of (100,500) mm day^?1.     

Diurnal variations of precipitation over the South China Sea

In this study, the diurnal variations of precipitation and related mechanisms over the South China Sea (SCS) are studied using the TRMM and other auxiliary atmospheric data. We have found that: (1) the amplitude and peak time of the diurnal precipitation over SCS exhibit remarkable regional features and seasonal variations. Diurnal variations are robust all the year around over the southern SCS especially over the Kalimantan Island and its offshore area. Over the middle to northern SCS, however, diurnal variations are noticeable only in the summer and autumn; (2) over the northern SCS precipitation peaks in early morning, while over the southern SCS it has two diurnal peaks: one in the early morning and another in the late afternoon; (3) the diurnal variations of precipitation over the SCS are related to the activity of the SCS summer monsoon and the ENSO events. The late afternoon precipitation increases remarkably after the onset of the SCS summer monsoon over the northern SCS. The early-morning rainfall peak is much more significant during La Nina years than during El Nino years; (4) the land–sea breeze is responsible for the diurnal cycle over the Kalimantan Island and its offshore area while the “static radiation–convection” mechanisms may result in the early-morning rainfall peak over the SCS.     

Diurnal cycle of precipitation over the Maritime Continent simulated by a spectral cumulus parameterization

The validity of a spectral cumulus parameterization (spectral scheme) for simulating a diurnal cycle of precipitation over the Maritime Continent (MC) was examined using a regional atmospheric model. The impacts of entrainment parameterization and each type of convective closure, i.e., non-equilibrium (or equilibrium) closure for deep convection, mid-level, and shallow convective closures, were also examined. When vertically variable entrainment and appropriate convective closures were employed, the model adequately simulated a diurnal cycle of precipitation over both land and ocean as compared to the observation. Analysis regarding the entrainment parameterization revealed that variable entrainment parameterization was needed not only for simulating better mean patterns of precipitation, but also for more realistic phases of diurnal cycles. The impacts of convective closures appeared in the differences in the precipitation amplitude. Analysis on diurnal cycles of convective properties and tendencies revealed that the cycles between boundary layer forcing and convective heating determined convection strength and were affected by each type of convective closure. It can be concluded that the spectral scheme with appropriate convective closures is able to simulate a realistic diurnal cycle over the MC.     

区域数值预报系统在北京地区的降水日变化预报偏差特征及成因分析

为了研究北京快速更新循环同化预报系统(BJ-RUCv2.0)在北京地区降水日变化的预报偏差特征及其成因,利用2012—2015年夏季BJ-RUCv2.0系统第2重区域(3 km分辨率)预报结果和北京地区122个自动气象站逐时观测数据以及观象台探空观测资料,分析模式对北京地区降水日变化预报偏差的区域性特征和传播特征,研究模式局地环流预报偏差特征及其对降水预报偏差的可能反馈机制。研究结果表明,BJ-RUCv2.0系统多个更新循环的预报在北京平原地区均存在夜间降水漏报问题,降水预报偏差表现为模式预报降水在西部山区降水偏多,预报降水雨带难以在平原地区增强发展,造成了模式降水在傍晚山区偏多而夜间平原地区降水明显偏少。通过分析模式局地环流预报偏差及其响应机制发现,由于白天平原地区近地层偏暖偏干,山区底层偏冷中层偏湿,造成了山区-平原地区间的温度梯度强度偏强且强温差出现时间提前,西部山区午后降水偏多;由于平原地区地面气温预报持续偏高,入夜后偏北风难以到达平原地区,造成了山区-平原间的地形辐合线位置偏北,影响山区降水雨带向平原地区移动,同时平原地区近地层内水汽持续偏低,抑制降水雨带在东移过程中的发展,造成模式在平原地区夜间降水预报容易出现漏报。模式冷启动所用的GFS资料土壤湿度在北京平原地区明显小于实际观测,是模式预报偏暖偏干的可能原因之一。     

基于WRF的积云对流参数化方案对中国夏季降水预报的影响研究

为了研究WRF(Weather Research and Forecasting)中尺度模式中积云对流参数化方案对夏季降水预报的影响,基于水平分辨率为9 km的WRF模式,采用Kain-Fritsch(KF)、尺度适应的KF、Tiedtke、new Tiedtke和尺度适应的new Tiedtke方案等5种积云对流参数...     

Diurnal and spatial variabilities of monsoonal CG lightning and precipitation and their association with the synoptic weather conditions over South Korea

The spatial and temporal variations in cloud-to-ground (CG) lightning and precipitation during the summer monsoon months in Korea have been analyzed in relation to the regional synoptic weather conditions. The lightning data used in this study were collected from a lightning detection network installed by the Korean Meteorological Administration, while the precipitation data were collected from 386 Automatic Weather Stations spread over the entire Korean Peninsula during 2000 to 2001. A distinctive morning peak of precipitation is observed over the midwest region of Korea. Along the east coast, little precipitation and CG flash counts are found. Despite the strong afternoon peaks of convective rainfall due to the high elevation over the southern inland region, the south coast shows nocturnal or early morning peaks, which represents a common oceanic pattern of flash counts. In 2000, the nighttime peak for lightning counts dominates over the southern area, while the afternoon peak was strong in the midland during the summer, mainly due to the northward transportation of moisture to the Korean Peninsula. Conversely, the strong afternoon peak for the southern region was confronted with early morning peaks in the midwestern region during 2001. The eastward transport of moisture has been analyzed and was considered to be dominant in 2001. The study of several warm and cold type fronts in 2000 and 2001 indicate that the warm type fronts in 2000 were associated with very little lightning, while the cold type fronts appeared to be responsible for the occurrence of abundant lightning in 2001, thereby, indicating that the warm and cold type fronts were representative of the local lightning distribution in the respective years.     

Statistical guidance on seasonal forecast of Korean dust days over South Korea in the springtime

This study aimed to develop the seasonal forecast models of Korean dust days over South Korea in the springtime. Forecast mode was a ternary forecast (below normal, normal, above normal) which was classified based on the mean and the standard deviation of Korean dust days for a period of 30 years (1981-2010). In this study, we used three kinds of monthly data: the Korean dust days observed in South Korea, the National Center for Environmental Prediction in National Center for Atmospheric Research (NCEP/NCAR) reanalysis data for meteorological factors over source regions of Asian dust, and the large-scale climate indices offered from the Climate Diagnostic Center and Climate Prediction Center in NOAA. Forecast guidance consisted of two components; ordinal logistic regression model to generate trinomial distributions, and conversion algorithm to generate ternary forecast by two thresholds. Forecast guidance was proposed for each month separately and its predictability was evaluated based on skill scores.     

Simulating extreme precipitation with a mesoscale forecast model

Summary ?A statistical analysis of two versions of BOLAM, a numerical model for mesoscale weather forecasting, is performed. The precipitation patterns forecasted by the models are evaluated by means of the standard statistical indexes, using the precipitation values observed in several stations of Liguria and Piedmont Regions (north-west of Italy) during 11 case studies occurred between 1984 and 1994. Results show a remarkable improvement of the performances of the last version of BOLAM, due to the recently implemented features both in the dynamical and the physical schemes of the model. Based on these results, the case study of 22–25 September 1993 (Brig flood) is simulated using integration domains of different size, in order to verify the possible influence of lateral boundary conditions on the model performance. In particular, results show that the use of larger domains improves the capability of the model in localizing the precipitation patterns correctly. In addition, the case study is simulated by introducing a delay in the triggering of downdrafts in the Kain-Fritsh convective parameterization scheme (CPS) as proposed by Spencer and Stensrud (1998). Results show improvements in the capability to forecast heavy rain events, in particular in enhancing the precipitation amounts over the Ligurian area. Received August 3, 2001; revised March 3, 2002     

The variation of evaporation over South China and its relationships to precipitation

The evaporation rate over South China is estimated based on the Climate Prediction Center Merged Analysis of Precipitation(CMAP)data and the NCEP/DOE reanalysis II data from 1979 to 2007. The temporal variation of evaporation over South China and its relationship to precipitation are discussed. Climatologically,the evaporation rate over South China is the largest in July and smallest in March.In spring and summer,the evaporation rate is approximately one half of the precipitation rate.However,the evaporation rate is approximately equal to the precipitation rate in fall and winter.The year-to-year variation of the evaporation rate over South China is quite in phase with that of the precipitation rate in the period from February to May but out of phase with that of the precipitation rate in early winter.Over South China there is a pronounced decreasing trend in the evaporation in colder seasons and a positive correlation between the evaporation variation and the rainfall variation in spring.In summer,the abnormality of rainfall over South China is closely related to the anomalous evaporation over the northeastern part of the South China Sea and its eastern vicinity.In winter,the rainfall variation in South China has a close linkage with the evaporation variation in a belt area covering the eastern Arabian Sea,the Bay of Bengal,the southeastern periphery of the Plateau,the southern part of South China Sea and the central part of Indonesia.     

中国暖季短时强降水分布和日变化特征及其与中尺度对流系统日变化关系分析

利用2010—2018年河南省371个气象观测站（包含122个国家站和249个骨干区域站）逐时降水资料,对河南省暖季（5—9月）小时极端降水时空分布特征进行了统计分析。主要结果如下：（1）河南省暖季第99.9百分位小时极端降水阈值、强度、频次和贡献率的局地差异明显,其高值区主要分布在伏牛山南部、黄淮平原东部和淮河流域西南部。（2）河南小时极端降水事件主要发生7、8月,其中7月最多,且有1/4以上为区域性极端降水事件;全省小时极端降水频次日变化表现为明显的双峰型,主峰值出现在傍晚;80 mm·h^-1以上小时极端降水频次日变化呈多峰结构,主峰值出现在夜间。（3）山地、丘陵、城市和平原四类下垫面区域的小时极端降水指标存在差异,城市小时极端降水强度最大,频次最低;山地小时极端降水强度最低,频次最高。（4）四类下垫面小时极端降水日变化虽均表现为双峰型,但也存在明显差异：山地其峰值以夜间为主,傍晚为辅;丘陵其峰值夜间、傍晚并存,且峰值强度接近;平原以及城市则以午后峰值为主,其中城市午后峰值强度更高。

     

2004年主汛期我国强对流天气灾害统计特征

利用1961-2015年江西省83个国家级气象观测站逐时降水资料,采用经验正交函数（EOF）和Morlet小波变换等方法,分析江西省短历时强降水时空分布特征,探讨当地短历时强降水与暴雨的关系。结果表明：（1）江西省短历时强降水极值与发生频次分布不匀,呈东多西少分布;全省短历时强降水发生频次变化一致,其中浙赣铁路沿线地区是其频次异常敏感区。（2）近55 a江西短历时强降水频次增加趋势明显,与江西区域性增暖呈正响应关系,该时段短历时强降水频次存在21 a和14 a左右年代际周期变化。（3）江西短历时强降水主要出现在4-9月,6月最多;其日变化4-7月呈双峰型,8-9月呈单峰型,强降水频次在16-17时最多。（4）江西短历时强降水与暴雨空间分布的高频中心相同,其相似系数达0.98;54.6%的短时强降水当日出现暴雨,尤其是凌晨至上午发生的短时强降水当日,暴雨概率高达69.2%~76.9%。

     

1961—2010年中国华南地区夏季降水结构变化分析

利用1961—2010年华南地区64个气象站的逐日降水资料,通过计算降水集中度指数Q,分析了华南夏季降水的结构。结果表明：夏季华南地区北部（南部）大部分地区降水集中度较小（大),表明该地区降水较为分散（集中)。在趋势变化上,近50年华南大部分地区夏季降水量和降水集中度都是增多的。北部和南部的降水量也均呈增加的趋势,北部增加更明显。另外,降水集中度在华南北部和南部也均呈增加的趋势,即降水呈现更集中的趋势,尤其是华南南部降水集中度增加更明显。此外,无论降水量为1 mm以上、25 mm以上还是50 mm以上的降水,持续1 d降水的雨日都在减少,而超过1 d的持续性降水过程都在增多。在空间分布上,华南大部分地区1 mm以上降水的雨日呈减少的趋势,而25 mm以上和50 mm以上的持续性降水过程呈增加的趋势。     

Daily precipitation forecast of ECMWF verified over Iran

In this paper, the performance of the Centre for Medium Range Weather Forecast (ECMWF) model (t?+?27 h to t?+?51 h) in predicting precipitation is discussed. This model is the first, which has been verified over Iran. The spatial resolution of the model is 0.351° and the 24-h forecasts are compared with daily observations. The study concentrates on year 2001 and the precipitation measurements were collected from the data of 2,048 rain gauges in Iran. The accuracy of four different interpolation methods (nearest neighborhood, inverse distance, kriging, and upscaling) was investigated. Using cross-validation, the inverse distance method (IDM) with minimum mean error was applied. Verification results are given in terms of difference fields (mean error?=?0.46 mm/day), rank–order correlation coefficients (0.70), as well as accuracy scores (false alarm ratio?=?0.50 and probability of detection?=?0.60) and skill scores (true skill statistics [TSS]?=?0.45) in year 2001. The position of the rain band was only partly captured by the ECMWF model; however, the position of maximum precipitations agrees with the observations well. The results show that the high values of TSS are associated with the large amounts of precipitation (over 25 mm). Slight to moderate precipitation events have been underforecasted by the model (bias?<?1) and it leads to a small value of TSS for these thresholds (5–25 mm/day). The ECMWF model has better performance in high and mountainous regions than over flat terrain and in deserts. Comparing TSS over the Alborz and the Zagros Mountains, it is obvious that the ability of the model to predict the convective precipitation events needs some improvement. The amount of daily precipitation has been also slightly overestimated over Iran. From the beginning of January up to 21 March 2001, the ECMWF time series indicates an obvious phase shift of 1 day, although in other months, no phase shift is noticed.     

Development of updateable model output statistics (UMOS) system for air temperature over South Korea

In this study, a Updateable Model Output Statistics (UMOS) system has been developed for the forecast of 3-h temperature over South Korea using two significantly different models’ (Regional Data Assimilation and Prediction System (RDAPS) and Korea Meteorological Administration (KMA) Weather Research and Forecasting (WRF) model (KWRF)) outputs based on the Canadian UMOS system (Wilson and Vallee, 2002; 2003). The UMOS system is designed to consider the local climatology and the model’s forecasting skills. The 20 most frequently selected potential predictors for each season, station, and forecast projection time from the 67 potential predictors of the Model Output Statistics (MOS) system, were used as potential predictors of the UMOS system. The UMOS equations are developed by a weighted blending of the new and old model data, with weights chosen to emphasize the new model data while including enough old model data in the development to ensure stable equations and a smooth transition to dependency on the new model. The UMOS equations were updated regularly at a predefined time interval to consider the changes of covariance structure between the new model output and observations as the new model data increase. The validation results showed that seasonal mean bias, Root Mean Square Error (RMSE), and correlation coefficients for the total forecast projection times are ?0.379～0.055°C, 1.951～2.078°C, and 0.741～0.965, respectively. Although, the forecasting skills of UMOS system are very consistent without regard to the season and geographic location, the performance is slightly better in autumn and winter than in spring and summer, and better in coastal regions than in inland region. When we take into account the significant differences of the RDAPS and KWRF, the UMOS system can be used as a supplementary forecasting tool of the MOS system for 3-h temperature over South Korea. However, the UMOS system is very sensitive to the selected number and/or types of predictors. Therefore, more work is needed to enable the use of the UMOS system in operation, including tuning of the number and types of potential predictors and automation of the updating processes of the UMOS equations.