应用气象学报

Advances of the Short-range Climate Prediction in China

Jia Xiaolong, Chen Lijuan, Gao Hui, Wang Yongguang, Ke Zongjian, Liu Changzheng, Song Wenling, Wu Tongwen, Feng Guolin, Zhao Zhenguo, Li Weijing

2013, 24(6): 641-655.

[FullText HTML] (1537) [PDF 566KB](1722)

Abstract:
Through the past 60-year development, the short-range climate prediction operation has made great progress in China in terms of the technology and methodology, undergoing the stages from the simple experiential statistic methods to numerical model. Especially in recent years, many objective prediction techniques and methods are well developed and applied in real-time operation. Meanwhile, many improved understanding and new knowledge of the climate system are also gradually used by climate forecasters.The ARGO (Array for Real-time Geostrophic Oceanography) global ocean data are applied in the global ocean data assimilation systems in NCC (NCC_GODAS), which enhances the monitoring and analyzing capability for the global ocean. The NCC_GODAS is integrated with coupled atmosphere-ocean models of NCC_CGCM, which increases the forecast skills for the short-term climate prediction. ARGO data are also applied for improving physical parameterization schemes in oceanic models, and the model capability of describing the real oceans and forecasting El Nio/Southern Oscillation is improved. The second-generation short-range climate forecast model, which upgrades many aspects of the resolution and physical process, exhibits a higher prediction skill comparing to the first-generation system. A preliminary evaluation indicates that the second-generation system shows a certain capability in predicting the pentad, ten-day, monthly, seasonal and inter-annual climate variability. The downscaling methods based on dynamical climate model are extensively used in operation including monthly prediction, seasonal prediction and extreme climate event prediction, improving the prediction skill of model production. Due to the limited predictability of a single model, multi-model ensemble (MME) is efficiently employed. Based on four operational dynamical models from NCC, NCEP, ECMWF and TCC, a multi-model ensemble system (MODES) is developed in NCC in 2011, in which downscaling technique is introduced and added to the ensemble prediction system. At present, this forecast system can issue monthly and seasonal ensemble prediction products and is applied by regional climate center. Based on the dynamical-statistical integration forecasting method, a forecasting system for seasonal precipitation (FODAS1.0) is developed, which has already been in quasi-operational use, showing stable prediction skill. The application of the intra-seasonal oscillation in the operational extended range forecast has made a great progress including that a MJO monitoring, prediction and application operational system is built up, and several forecast methods based on the intra-seasonal oscillation are applied. New knowledge and research achievements are gradually introduced into operation by forecaster, for example, in addition to the sea surface temperature (SST) in the equatorial mid-east Pacific, the SST in the Indian Ocean and the Atlantic Ocean are also seriously considered. In addition, the sea ice, snow cover over the Eurasia and the climate system in the Southern Hemisphere are also considered as the important impact factors in seasonal prediction.

Recent Progress on the Objective and Quantifiable Forecast of Summer Precipitation Based on Dynamical-statistical Method

Feng Guolin, Zhao Junhu, Zhi Rong, Gong Zhiqiang, Zheng Zhihai, Yang Jie, Xiong Kaiguo

2013, 24(6): 656-665.

[FullText HTML] (1283) [PDF 1972KB](1297)

Abstract:
Short-term climatic prediction, which mainly aims at monthly, seasonal and annual time scales, is very important for the public and government decision making. The trend of summer flood and drought distribution is one of the most important contents in operational forecast. Generally, there are two types of forecasting methods, including statistical method and dynamical method, which both have advantages and disadvantages. Therefore, the general consensus is to let them learn from each other, merging and developing. During recent 50 years, the Dynamical-Statistical Integration Forecasting Method (DSIFM) has made great progresses in dealing with the complex scientific issue of summer precipitation forecasting in China and abroad.The research results in early period and the development about DSIFM are briefly reviewed, as well as the two forms of dynamical-statistical integration forecasting method. And then, the principle, processes and programs of Dynamical-Statistical Objective Quantitative Forecasting (DSOQF) in recent operational forecast are systematically introduced. Based on the Coupled Global Circulation Model (CGCM) of National Climate Center and two types of prediction scheme of DSOQF, a dynamical-statistical integrated forecasting system for seasonal precipitation (FODAS1.0) is set up, which fully assimilates existing research and profession achievements, especially forecaster diagnostic techniques and forecasting experience from national, regional and provincial climate centers. Suitable regional climate characteristics prediction scheme is also developed based on the theory and methods of DSOQF. By now, FODAS1.0 achieves quasi-operational trial in National Climate Center, 8 regional climate centers and Guangxi, Shandong and other provincial climate centers.Experimental predictions are carried out for the summer rainfall in China from 2009 to 2012 with the method of DSOQF. The predictive score (PS) from 2009 to 2012 are 79, 72, 70 and 70, respectively. The anomaly correlation coefficient (ACC) from 2009 to 2012 are 0.38, 0.10, 0.12 and 0.03. For abnormal years such as 2010 and 2011, diagnostic analysis is performed. Overall, the forecast results are ideal, but it still needs further improving.The problems in DSIFM and the solutions are also discussed. The forecasting skills can be improved by strengthening diagnostic analysis of the relationship between precipitation and its main factors, improving the physics processes and parameterization scheme of short-term climate models, and developing the targeted regional climate models. The DSIFM will be more useful in the future.

Major Advances of China Climate Monitoring and Diagnosis Operation in Recent 20 Years

Li Qingquan, Sun Chenghu, Yuan Yuan, Si Dong, Wang Dongqian, Wang Yanjiao, Guo Yanjun, Liu Yanju, Ren Fumin, Zhou Bing, Wang Pengling

2013, 24(6): 666-676.

[FullText HTML] (1213) [PDF 708KB](1819)

Abstract:
Climate monitoring and diagnosis is an important method to know about the changes of climate system and their causes, playing an important role in climate operation. The history and current status of the climate monitoring and diagnosis operation in China are reviewed and some new technologies, methods and mechanisms developed and applied in operation in recent years are introduced.In 1990, the operational system of monthly climate monitoring is established in China for the first time. After the developments of more than 20 years, the climate monitoring and diagnosis operation in China makes great progresses in the aspects of theory, method, technology, and operational system, which go through the stages from the simple analysis of station observations to the monitoring of the multiple layers of climate system and the dynamic diagnosis of climate anomalies.At present, a set of climate monitoring and diagnosis operation system on multiple temporal and spatial-scales is set up and continuously improved, and some new theories, technologies, and approaches are promoted and applied. In the aspect of atmospheric circulation, besides basic characteristic variables related to height, wind and so on, the monitoring and diagnosis operational products of stratospheric process, northeast cold vortex, Walker circulation, Hadley circulation, water vapor transportation are also developed. Furthermore, the monitoring and diagnosis operation of the extreme weather and climate events over China and the world is gradually set up and improved. In the aspect of ocean, besides the physical system in the tropical and extra-tropical ocean and atmosphere related to ENSO, the monitoring and diagnosis technologies and products in light of the sea temperature anomalies of Indian Ocean and Atlantic Ocean with the major modes of oceanic and atmospheric system are developed. In the aspect of sea ice and snow, the monitoring and diagnosis technologies and products of the Antarctic and Arctic sea ice concentration and days as well as snow days, snow areas, and snow depths of the Northern Hemisphere and China are developed and set up. In the aspect of land surface, the monitoring technologies and products on the basis of station observations of soil temperature and humidity are developed. At the same time, the research is enhanced on the key factors, such as atmosphere, ocean, land surface, sea ice, snow and so on, as well as the mechanisms how they influence climate anomalies in China.

Multi-model Downscaling Ensemble Prediction in National Climate Center

Liu Changzheng, Du Liangmin, Ke Zongjian, Chen Lijuan, Jia Xiaolong, Ai Wanxiu

2013, 24(6): 677-685.

[FullText HTML] (1241) [PDF 22631KB](1077)

Abstract:
Dynamic model is the dominant tool for the seasonal prediction operation in most climate prediction centers of the world. But now, for any single model, the predictability to seasonal precipitation and temperature is quite limited. Therefore, two kinds of techniques (i.e., multi-model ensemble and downscaling) are developed efficiently to access better prediction ability. Multi-model ensemble can reduce model error and then bring higher prediction skills. Meanwhile, as the model predictability of circulation is better than that of precipitation and temperature, downscaling improves the prediction of temperature and precipitation via regional model or statistic methods.Due to the complex physical mechanism, the seasonal prediction to China climate is much a challenge. China National Climate Center (NCC) develops a new kind of prediction technique combining multi-model ensemble and downscaling. At present, the output variables from four seasonal models from WMO GPCs (including ECMWF, TCC, NCEP and NCC) are used as predictors and four statistic downscaling methods (EOF-ITE, BP-CCA, Optical Subset Regression, Regress Ensemble of High Correlation Factors) are used to set prediction model. Every model output and every downscaling method are used so that 16 model-downscaling components are available. Besides, two methods (equal-weighted average, classic super-ensemble) are employed to access the ensemble result, respectively. As an index showing the prediction ability, the mean PS scores are computed for the reforecast of recent five years for every model-downscaling and ensemble component. The component with highest mean PS score is chosen as the best prediction result.In NCC, the Multi-model Downscaling Ensemble Prediction System (MODES) are set up to realize the above ideas and the operational application of monthly and seasonal temperature with precipitation over China. Reforecast and operational application are carried out. The present reforecast and operational application for seasonal climate indicates that MODES has achieved quite good prediction skills for temperature and also improved precipitation prediction. The real-time application for monthly climate prediction for from September 2012 to July 2013 is assessed with NCC traditional PS methods. For monthly mean temperature, MODES holds the mean and median PS score of 76 and 81, respectively, showing much good prediction ability. Meanwhile, for the monthly precipitation of MODES, the mean and median PS scores are both 68, higher than mean scores of the operational prediction product of NCC. The reforecast and operational application indicate that MODES is a useful tool for the short-term climate operation prediction.

A Verification Approach for the Assessment of Extended-range Process Event Prediction

Du Liangmin, Ke Zongjian

2013, 24(6): 686-694.

[FullText HTML] (1053) [PDF 1355KB](1251)

Abstract:
Based on the features of forecast and assessment for extended-range weather and climate events, a verification approach named PPS (process-event prediction score) for process event forecast is proposed, which is combined with the actual requirements of extended-range forecast operation. This approach considers not only the criteria of event forecast scores including hit rate, false alarm rate commonly used in weather forecast operation, but also the advantages of other approaches such as Euclidean distance and dynamic time warping distance.As the forecast period is relatively long, it is very difficult to forecast a process event completely and accurately. Therefore, it is terrifically valuable for adjacent hit, denoting the forecast with one to two days lead or lag, in operational application. Based on the above-mentioned features, the periods of forecast and observation for process events are extended, respectively, and the virtual events are transformed into similar imaginary events. In terms of the accordance extent between forecast and observation, classified score table is constructed. Moreover, weight is used to show the influence of false alarm on forecast score.The features of PPS approach are assessed by couples of cases including "no false alarm but missing" and "missing and false alarm", and the relations of PPS to hit rate and false alarm rate are analyzed. Under the condition of "no false alarm but missing", scores of PPS and hit rate increase with the correct forecast number of days. The PPS score is generally higher than hit rate score, which indicates the increasing score effect from the expansion for process events of observation and forecast. In the case of missing and false alarm, PPS scores are higher than the hit rate score when false alarm rate is low. However, PPS scores will be lower than the hit rate score when false alarm rate significantly increases, which shows the influence of false alarm to PPS score. Combined with the features of process event forecast and the possible influence of false alarm on forecast skill, PPS score objectively reflects the actual skill of forecast. Compared with hit rate and false alarm rate, it is more efficient to represent the process event information involved in a forecast. Therefore, it is more applicable for assessing the skills of process event forecast.By this approach, skills of operational cold air process forecast are assessed during winters from 1999 to 2010. The results show that the PPS score reflects the accuracy of cold air process forecast well. Moreover, the verification actually indicates relatively low accuracy of extended-range forecast today. Above all, this approach can be used to assess extended-range process forecast and shows good prospect for operational application.

Staged Meteorological Drought Index Based on Boltzmann Function

Hou Wei, Yang Jie, Zhao Junhu

2013, 24(6): 695-703.

[FullText HTML] (1054) [PDF 1229KB](1271)

Abstract:
In the practical research and business, it's often needed to evaluate the overall meteorological drought degree of a site within a certain period of time. Based on the daily meteorological drought index, the accumulative probability distribution of the number of days for different drought and flood degrees at a time scale can be obtained, and through solving the accumulative probability distribution by Boltzmann function, the standardized staged meteorological drought intensity index (I_SD) and staged meteorological drought discrepancy index (I_SDD) is designed.I_SD and I_SDD of Kunming Station at the monthly scale are constructed as an example. Through comparing and analyzing the precipitation anomaly percentage, I_SD, standardized precipitation index (I_SP) and I_SDD in the different periods, and combined with the day by day evolution of the rainfall and multi-scale standardized precipitation index (I_MSP), the effectiveness of I_SD and I_SDD index are verified. A basic principle is that the rainfall at a certain moment could affect the drought and flood states after that moment, but has no impact on the drought and flood states before.I_SD makes full use of daily meteorological drought index which contains prophase precipitation information to synthesize drought and flood characteristics of a certain period of time. The bigger I_SD is, the more serious the drought is, and vice versa. The larger (smaller) I_SDD means the drought or the flood is more concentrated (scattered). Moreover, these two indices can be calculated in different time scales for any site, and in the practical application, the other meteorological drought index also can be used by a forementioned method to build I_SD and I_SDD. A theoretical algorithm to get I_SD and I_SDD on any time scale is proposed, but the most applicable time scale should be 10 days, month and season, and on the inter-annual and inter-decadal scale, the proposed approach and other methods such as precipitation anomaly percentage and I_SP can also be used.In view of the complexity of drought, no single index can fully express its intensity, harm and the potential impact, so the drought index is still being continuously explored and improved.

The Phase Features of a Cold Vortex over North China

Gao Yanna, He Lifu

2013, 24(6): 704-713.

[FullText HTML] (1081) [PDF 4741KB](1370)

Abstract:
Using the conventional weather data, disastrous weather data, hourly precipitation data observed by automatic weather station and NCEP/NCAR reanalysis data, the phase features over North China during 12—20 July in 2011 are analyzed. The result indicates that the precipitation of the vortex is located in the northeast of Mongolia, North China and the south of Northeast China. Thunderstorm, gale and hail occur in the development stage, while short-time strong rainfall occurs in the weakening stage. Strong warm ridge is located at 850 hPa, a full jet stream exists at 200 hPa, and cold-core structure is presented on the whole troposphere in the development stage. The strong warm ridge isn't obvious in the weakening stage, the east wind is enhanced at the lower layers, and the cold-core also increases. The vortex is baroclinic at 700 hPa, the ascending motion is located in the east area and becomes stronger in the weakening stage, and the southeast wind is also speeded up. The relative humidity is larger at 300 hPa and 700 hPa, but it's dry at 500 hPa, and the south of cold vortex is invaded by dry and cold air in the development stage. The relative humidity is large at the whole layer in the weakening stage, as the vapor is brought by east wind. A northeast—southwest θ_e frontal zone exists in both stages of the cold vortex. The gradient of θ_e becomes significantly strong in the development stage, and the structure is unstable with dry and cold at the upper layer while warm and wet at the lower layer in the vertical direction. The θ_e frontal zone and the unstable structure weakens in the weakening stage. The cold advection which is located at the low and middle layers and the vorticity advection which is located at the middle and high layers play main roles for the cold vortex in the development stage. As the cold advection at the low layer enters the cold vortex center, the cold advection becomes weak at the middle layer, the positive vorticity advection recedes at the medium-to-high level, and the cold vortex weakens.

Estimating Tropical Cyclone Vertical Gradient Parameter and Its Relationship with TC Intensity

Wang Xin, Fang Xiang, Liu Nianqing

2013, 24(6): 714-722.

[FullText HTML] (1027) [PDF 2639KB](1160)

Abstract:
There are many challenges in tropical cyclone (TC) intensity theoretical study and forecasting, and diagnostic analysis and mechanism study will help improve the understanding and predication of TC intensity evolution. And revealing TC internal structure is one of efficient methods. It is satisfying to find that TC structure and its environmental field can even be described by satellite measurements. Especially, for the applications with some microwave instruments can provide TC vertical temperature profiles to help detecting the TC internal structure, such as NOAA/AMSU-A (advanced microwave sounding unit-A) on board the polar satellites. By using NOAA/AMSU-A retrieved temperature data, more details of TC vertical structure evolution could be presented, Therefore, an objective method describing the vertical symmetry is originally proposed, which focuses on how to calculate a vertical structure index with the combined information from observations, which including the TC track and strength from CMA-STI data.In this method, the tropical cyclone vertical gradient parameter (F_TC-VGP) is defined, and several prominent features for the TC upper warm-core variation in aspects of altitude, temperature anomaly and the warm core center location are considered for calculation. This parameter of F_TC-VGP describe the TC upper warm-core declined from their low level circulation, it reflects the three-dimensional thermal structural information. And for this calculating method of TC internal structure, typical TCs occurs in the South China Sea and the Northwest Pacific during 2009—2011 are selected as the examples, F_TC-VGP time series of all TC samples are calculated, and while the relationship between F_TC-VGP and TC intensity evolution are inspected, results can be divided into three main parts.One of results demonstrates that F_TC-VGP is well fitted to the TC intensity and indicating the TC intensity evolution process, anomalous and abrupt points of TC intensity in the time sequence can be found by F_TC-VGP. That is, F_TC-VGP abnormal inflection point indicates TC strength mutation. The second, intensity analysis result is about the relationship between F_TC-VGP and environmental circulation. When the variation between TC intensity and F_TC-VGP is found opposite trend, the intensity evolution cause is considered from other influences. It shows that sometimes intensity decrease is not caused by vertical structure changes, but by environmental cold air. Another important conclusion shows that F_TC-VGP has a fast adaptive adjustment process to contribute to TC intensity, for example, during the TC strengthening process, the F_TC-VGP increasing and decreasing generate the TC development. The result also presents the intensity changes with their internal cloud structure. In brief, F_TC-VGP can give a quantitative description of TC internal three-dimensional structure characteristics, showing important reference value for grasping TC intensity and trends accurately, as well as TC monitoring and forecasting.

Light Response Characteristics of Summer Maize at Different Growth Stages Under Drought

Wu Wei, Jing Yuanshu, Ma Yuping, E Youhao, Sun Linli, Zheng Tengfei

2013, 24(6): 723-730.

[FullText HTML] (1504) [PDF 1377KB](1397)

Abstract:
The field experiment of drought on summer maize growth is carried out by using large electric water proof and irrigation installations. First of all, the diurnal variations of photosynthesis and photosynthesis-light response curves of summer maize leaves are measured. And then, the different models are used to fit light response curve to determine the optimal model and extract the photosynthetic parameters. Finally, the impact of soil moisture on the photosynthetic characteristics of summer maize leaves at different growth stages is discussed. The comparison of light response curve fitting by different models shows that comparing to the non-rectangular hyperbolic model and exponential model, the simulation result of modified rectangular hyperbola model is better. In particular, it can effectively simulate the downward trend of light saturated net photosynthetic rate with light intensity increased, which is more common under drought conditions. In addition, the use of modified rectangular hyperbola model can extract the quantum efficiency of the light compensation point which is the numerical uniqueness indicator of evaluation of crop light use. The photosynthetic parameter analysis shows that both light saturation point (LSP) and maximum net photosynthetic rate (P_max) decline in different growth stages, and quantum efficiency of light compensation point (CQY) and light compensation point (LCP) are insignificantly affected under slight drought. With the aggravation of drought, LSP and P_max has a further decrease and CQY has a significant decline while LCP had a great increase under severe drought condition. The comparison of different growth stages show that LSP and P_max decline largest in jointing stage, second in tasselling stage and least in milky maturity stage under slight drought. LSP and P_max decrease by 24.1% to 43.7% and 9.3% to 46.1%. LSP and P_max decline largest in tasselling stage, the second in milky maturity stage, the least in jointing stage under severe drought. LSP and P_max decrease by 12.3% to 33.6% and 48.5% to 62.2%. In addition, observations show that photosynthetic and transpiration rate of summer maize leaves at different growth stages both decline under drought. The comparison of different growth stages show that photosynthetic and transpiration decline largest in tasselling stage, second in jointing stage and least in milky maturity stage under slight drought. With the aggravation of drought, photosynthetic and transpiration still decline largest in tasselling stage, but second in milky maturity stage, and least in jointing stage. Water use efficiency of maize leaves at different growth stages are relatively large under suitable soil water condition (2.8—4.5 μmol·mmol^-1), and slight drought (2.6—4.2 μmol·mmol^-1). Relative to tasselling and milky maturity stage, water use efficiency of maize leaves in jointing stage is the largest.

Analog Bias Correction of Numerical Model on Wind Power Prediction

Xu Jingjing, Hu Fei, Xiao Ziniu, Li Jun

2013, 24(6): 731-740.

[FullText HTML] (1170) [PDF 1031KB](2407)

Abstract:
A new post-processing method is proposed to reduce numerical weather prediction's systematic and random errors. The method overcomes a difficulty of a post-processing algorithm inspired by Kalman filtering and a 7-day running-mean correction in dealing with sudden changes of the forecast error that could be caused by rapid weather transitions. The analog forecast for a given location and time is defined as a past prediction that matches selected features of the current forecast. The method is the weighted average of observations that verifies when the best analogs are valid. The method is tested for 70-m wind speed prediction from Weather Research and Forecasting (WRF) model, with observations from one wind farm sited at Yanchang, Shaanxi Province for 3 months.The analog bias correction method is able to produce skillful corrections of the raw forecasts, even with large day-to-day changes in forecast error, and thus the method can predict drastic changes in forecast error. Moreover, being a prediction based solely on observations, it results in an efficient downscaling procedure that eliminates representativeness discrepancies between observations and predictions.Also, it is able to reduce random errors, therefore improving the predictive skill of raw forecast. The correction method is much better, with average improvement of 9.3% and 9.8% measured by root mean square error (RMSE) and centered root mean square error (CRMSE), respectively. Meanwhile, the method shows a better pattern of correspondence between predictions and observations.Moreover, the correction method for middle wind speed (5—12 m·s^-1), which plays the most important role on wind power prediction, is much better, with average improvement of 12.3% and 21.7% measured by RMSE and CRMSE, respectively. Thus the analog bias correction method is very suitable for wind power prediction.The analog bias correction method is based purely on verifying observations of past predictions that are similar to the forecast (i.e., the analogs), which provide physically based insight about the atmospheric state, thus improving the predictive skill. And it also has the potential to be applied to other prediction systems and variables.

Diagnostic Analysis on the First Summer Rainstorm Process of Central Yunnan in 2012

Zhou Hong, You Hong, Li Fan, Cai Aiping

2013, 24(6): 741-752.

[FullText HTML] (1034) [PDF 9222KB](1015)

Abstract:
Based on intensive observations, hourly FY-2E infrared TBB data, Doppler radar echo data and analysis data of NCEP (1°×1°, 4 times a day), the first rainstorm process in central Yunnan from 1 June to 2 June in 2012 are diagnostically analyzed using meso scale filtering method and generalized moist potential vortices theories (GMPV).The result shows that this strong precipitation process is caused by cold front and sheer, which is typical in central Yunnan. Shear line, mesoscale convergence line and meso-β-scale low vortex are significant at 700 hPa after mesoscale filtering, but they are not obvious in largescale original stream fields. So the direct causes for this rainstorm process are mesoscale weather systems. It seems apparent that the rainstorm always happens at the side which TBB gradient is higher in the convective cloud clusters by hourly FY-2E infrared TBB data. After analysis on Doppler radar echo data, there is a large area of flocculent echoes at the strong precipitation region, and then some convective clouds develop in these flocculent echoes. Distribution of rainfall is not uniform in space and the efficiency of rainfall is high because of uneven distribution of echoes in space, low height and dense structure of echoes. The source region of water vapor is the Bay of Bengal. The water vapor convergence zones have a good correlation with the position of surface cold front, shear line, mesoscale convergence line and meso-β-scale low vortex at 700 hPa. The ground precipitation strengthens when the center of vapor convergence area at 700 hPa and 850 hPa are superimposed.The positive anomaly of GMPV at mid-low layers over strong rainfall area can reflect characteristics of high water vapor convergence. Vertical distribution and change of GPMV at the low layer of single station show good indicative significance in this strong rainfall process. The rainfall is intensified when the positive anomaly of GPMV at the low layer of single station increase, and vice versa. The GMPV at 800 hPa has an indicative effect on the location of heavy rainfall. The area of GPMV positive anomaly is always located in the center of strong precipitation and its surrounding area, but the center of strong precipitation is not coincided with the center of positive anomaly of GMPV completely. The forecast of this process will be better if the circulation patterns are analyzed synthetically, and the generalized moist potential vorticity theories are used as well.

High-performance Computing System Performance Evaluation Method and Its Application

Wei Min, Sun Jing, Shen Yu, Xiao Huadong, Li Juan

2013, 24(6): 753-760.

[FullText HTML] (1201) [PDF 676KB](2714)

Abstract:
Meteorological numerical simulation has a high requirement on hardware structure, software realization and system performance of high performance computing system. It's one of the main applications of high performance computing (HPC). With the development of HPC, the scale of the high performance computers is expanded rapidly. The peak performance of computers increases in a continuous and rapid way. But the sustained performance achieved by the real applications does not increase in the same scale as the peak performance does. As the computer architectures and program structures are becoming much more complex, more and more factors may affect the performance of programs.Combined with the real meteorological numerical model, the technical research work of effective performance evaluation to high performance computing system has more and more important guiding significance of each stage in system's procurement, installation and performance optimization.In order to promote the ability of dealing with climate change science and technology support, China Meteorological Administration starts a new generation of national meteorological high performance computing power construction. The high performance computing system performance evaluation technology research during this period is carried out.The computing features of meteorological numerical models and its demand for high performance computing systems are analyzed in detail. Combined with the existing high performance computing system architectures and the attributes of software and hardware, the performance evaluation scheme of high performance computing system is proposed. Adopting the actual measurement, benchmark test and evaluation model methods, the comprehensive performance test of high performance computing system is carried out. The test is divided into two stages. In system selection stage, the evaluation model method and benchmark test method are used. In the system installation and debugging stage, the actual measurement method is used. The research results are applied to the practical work, conducting reasonable inspection, performance and function analysis and contrasting to the test results of mainstream high performance computing systems. Through quantitative marking, the high performance computing systems are evaluated reasonably and objectively. The practice result shows that the evaluation scheme design is reasonable and evaluation model research has achieved good results, ensuring the new generation of high performance computing system bidding work successfully completed.With the rising requirements of meteorological numerical model and the development of high performance computing environment, the test program set, evaluation method and evaluation model will be optimized and improved continuously. And on this basis, scalable high performance computing system performance evaluation model for heterogeneous system will be built.

Reliability Design of Meteorological Broadband Network at Provincial Level

Chen Xiaoyu, Zhang Yonghua, Wu Zhaoxiong, Wang Jia, Sun Peng

2013, 24(6): 761-768.

[FullText HTML] (1058) [PDF 1644KB](1303)

Abstract:
With the rapid development of information technology, the demand of connectivity keeps rising and the scale of network keeps expanding. Design and implementation of a reliable wide-area-network are critical for information transmission. With the development of meteorological services and growth of social needs, observations and forecast data gradually become larger and call for timeliness. Simple network infrastructure is no longer a guarantee for reliable transmission and highly reliable network is desired, not only for mass data transmission, but also for higher reliability. At present, the reliability of broadband network at provincial level is normally realized by taking a downgrade backup. But the bandwidth of backup link is narrow, so the transmission performance cannot catch up with the original link. Besides, it's unable to switch links automatically, and manual intervention is still needed.An original network reliability designing method is proposed based on BFD (bidirectional forwarding detection), NQA (network quality analysis) network reliability designing scheme, and strategy of routing technology. Through summarizing the network reliability research status, analyzing the meteorological broadband network reliability, the design and realization of this new meteorological broadband network is discussed from three aspects: The network backup, the flow sharing and the automatically switching. The application of the scheme has improved the network structure and bandwidth of the meteorological Province-City-County network system, and also has realized network backup, flow sharing and automatic switching on faults, which improves the reliability level greatly. There are 3 challenges in the design and implementation of this network. The first is the implementation of redundancy network with heterogeneous transmission links, the second is the routing redundancy of physical link testing, and the last is the real-time switching at business level of task types. The scheme has been tested strictly and successfully implemented in meteorological broadband network of Guangdong Province, and it's also suitable for meteorological network design of other provinces.

Vol.24, NO.6, 2013