ISSN 1000-0526
CN 11-2282/P
CN 11-2282/P
Volume 42,Issue 9,2016 Table of Contents
2016, 42(9):1033-1044.
DOI: 10.7519/j.issn.1000-0526.2016.09.001
Abstract:
Using background error covariance to construct dynamic balance constraints is very important to the design of variational assimilation schemes. It can permit a multivariate analysis, increase the value of each individual observation, and also improve the preconditioning of variational assimilation schemes. In this paper, the basic method and process of incorporating dynamic balance constraints are first surveyed.Based on this, the common problems of constructing the widelyused quasigeostrophic and hydrostatic balance constraints in global and limited domain assimilation schemes are reviewed.Three constructing quasigeostrophic balance constraint methods are discussed, including dynamic balance equation method, statistical method, and dynamicstatistical combined method.The undetermined problems of constructing hydrostatic balance constraint in different vertical grids and their solutions are also surveyed.Finally, the future challenges and opportunities of constructing dynamic balance constraints using background error covariance in special areas (such as tropics), in designing high resolution assimilation schemes and ensemblevariation hybrid schemes are discussed.
Using background error covariance to construct dynamic balance constraints is very important to the design of variational assimilation schemes. It can permit a multivariate analysis, increase the value of each individual observation, and also improve the preconditioning of variational assimilation schemes. In this paper, the basic method and process of incorporating dynamic balance constraints are first surveyed.Based on this, the common problems of constructing the widelyused quasigeostrophic and hydrostatic balance constraints in global and limited domain assimilation schemes are reviewed.Three constructing quasigeostrophic balance constraint methods are discussed, including dynamic balance equation method, statistical method, and dynamicstatistical combined method.The undetermined problems of constructing hydrostatic balance constraint in different vertical grids and their solutions are also surveyed.Finally, the future challenges and opportunities of constructing dynamic balance constraints using background error covariance in special areas (such as tropics), in designing high resolution assimilation schemes and ensemblevariation hybrid schemes are discussed.
2016, 42(9):1045-1057.
DOI: 10.7519/j.issn.1000-0526.2016.09.002
Abstract:
A review of flood forecasting hydrometeorological forecast technology is presented with application of watershed quantitative precipitation forecasts in flood forecasting, hydrological model, and uncertainty analysis of hydrometeorological forecasting. The results show that gridbased watershed quantitative precipitation forecasting with forecasters’ forecasts is an important means of improving the accuracy of watershed quantitative precipitation forecasting in flood forecasting operations, and regional ensemble weather prediction modeling is the main method to improve local heavy precipitation forecasting. The distributed hydrological model, based on control volume method with coupling conceptual runoff generation module and physical routing module, is the developing direction of hydrological model for flood forecasting. Hydrological ensemble prediction is the most popular methods for the uncertainty analysis of hydrometeorological forecasts and the models based on Bayesian theory are one of the most effective solutions to the forcast uncertainty.
A review of flood forecasting hydrometeorological forecast technology is presented with application of watershed quantitative precipitation forecasts in flood forecasting, hydrological model, and uncertainty analysis of hydrometeorological forecasting. The results show that gridbased watershed quantitative precipitation forecasting with forecasters’ forecasts is an important means of improving the accuracy of watershed quantitative precipitation forecasting in flood forecasting operations, and regional ensemble weather prediction modeling is the main method to improve local heavy precipitation forecasting. The distributed hydrological model, based on control volume method with coupling conceptual runoff generation module and physical routing module, is the developing direction of hydrological model for flood forecasting. Hydrological ensemble prediction is the most popular methods for the uncertainty analysis of hydrometeorological forecasts and the models based on Bayesian theory are one of the most effective solutions to the forcast uncertainty.
2016, 42(9):1058-1068.
DOI: 10.7519/j.issn.1000-0526.2016.09.003
Abstract:
In this study, two land surface schemes (LSSs), Noah LSS and NoahMP LSS, are evaluated over the East Asian Region for January and July in 2013. Through the evaluation of 2m air temperature, 10m wind speed, surface temperature, soil temperature at depths of 5 cm and 70 cm, and soil moisture at depths of 5 cm and 25 cm, the results showed that: (1) Compared to Noah LSS, NoahMP LSS produces less surface sensible heat flux but more latent heat flux in January in most areas. NoahMP LSS produces more sensible heat flux except for Indian continent, western China and its adjacent regions, northeastern China in July, also more latent heat flux in most areas. The magnitude of increase in latent heat flux in July is larger than in January. (2) Compared to Noah LSS, NoahMP LSS improves the overall simulation of soil moisture and soil temperature over the East Asian Region. (3) Compared to Noah LSS, the bias and rootmeansquarederror of simulated 2m air temperature and 10m wind speed with NoahMP LSS decrease, especially the simulation for the 2m air temperature over Indian continent and cold climate regions is improved significantly. This study indicates the advantage of coupled NoahMP over the East Asian Region, and provides the basis for WRF/NoahMP in future operational application.
In this study, two land surface schemes (LSSs), Noah LSS and NoahMP LSS, are evaluated over the East Asian Region for January and July in 2013. Through the evaluation of 2m air temperature, 10m wind speed, surface temperature, soil temperature at depths of 5 cm and 70 cm, and soil moisture at depths of 5 cm and 25 cm, the results showed that: (1) Compared to Noah LSS, NoahMP LSS produces less surface sensible heat flux but more latent heat flux in January in most areas. NoahMP LSS produces more sensible heat flux except for Indian continent, western China and its adjacent regions, northeastern China in July, also more latent heat flux in most areas. The magnitude of increase in latent heat flux in July is larger than in January. (2) Compared to Noah LSS, NoahMP LSS improves the overall simulation of soil moisture and soil temperature over the East Asian Region. (3) Compared to Noah LSS, the bias and rootmeansquarederror of simulated 2m air temperature and 10m wind speed with NoahMP LSS decrease, especially the simulation for the 2m air temperature over Indian continent and cold climate regions is improved significantly. This study indicates the advantage of coupled NoahMP over the East Asian Region, and provides the basis for WRF/NoahMP in future operational application.
2016, 42(9):1069-1078.
DOI: 10.7519/j.issn.1000-0526.2016.09.004
Abstract:
Based on the CMASTI tropical cyclone best track data from 1951 to 2015, the tropical cyclones activities over the Western North Pacific (including the South China Sea) in strong El Ni〖AKn~D〗o years are analyzed. The features of tropical cyclones in 2015 and their corresponding atmospheric circulation characteristics are discussed as well, for the year 2015 is the strongest El Ni〖AKn~D〗o year since 1997. The results show that in strong El Ni〖AKn~D〗o years, there are a larger number of TCs in the first half of the year and smaller number of TCs from June to October, and fewer TCs land in the whole year. The TC genesis is to the southeast of its normal and they tend to recurve from northwestward to northeastward. TC activity in 2015 is in accordance with the laws above. There is a largescale anomalous westerly over the western equatorial Pacific before El Ni〖AKn~D〗o gets mature, and a strong El Ni〖AKn~D〗o weakens the Walker circulation. Both of them suppress the convection. Besides, the subtropical high is located further south, making the ITCZ move southward. All these factors result in fewer TCs in 2015. In strong El Ni〖AKn~D〗o year, the strength of all the TCs is stronger, and this is due to the weaker intensity and smaller area of the subtropical high. There are more TCs forming in the southeast part of the WNP with stronger intensity, and the number and intensity of TCs in the northeast part is in the opposite condition; TCs generating in the SCS is fewer and weaker than normal as well. The TC activity characteristics in 2015 accord with the above rules and they are attributed to the eastward western ridge point of subtropical high and an anomalous northwesterly over the WNP at 500 hPa.
Based on the CMASTI tropical cyclone best track data from 1951 to 2015, the tropical cyclones activities over the Western North Pacific (including the South China Sea) in strong El Ni〖AKn~D〗o years are analyzed. The features of tropical cyclones in 2015 and their corresponding atmospheric circulation characteristics are discussed as well, for the year 2015 is the strongest El Ni〖AKn~D〗o year since 1997. The results show that in strong El Ni〖AKn~D〗o years, there are a larger number of TCs in the first half of the year and smaller number of TCs from June to October, and fewer TCs land in the whole year. The TC genesis is to the southeast of its normal and they tend to recurve from northwestward to northeastward. TC activity in 2015 is in accordance with the laws above. There is a largescale anomalous westerly over the western equatorial Pacific before El Ni〖AKn~D〗o gets mature, and a strong El Ni〖AKn~D〗o weakens the Walker circulation. Both of them suppress the convection. Besides, the subtropical high is located further south, making the ITCZ move southward. All these factors result in fewer TCs in 2015. In strong El Ni〖AKn~D〗o year, the strength of all the TCs is stronger, and this is due to the weaker intensity and smaller area of the subtropical high. There are more TCs forming in the southeast part of the WNP with stronger intensity, and the number and intensity of TCs in the northeast part is in the opposite condition; TCs generating in the SCS is fewer and weaker than normal as well. The TC activity characteristics in 2015 accord with the above rules and they are attributed to the eastward western ridge point of subtropical high and an anomalous northwesterly over the WNP at 500 hPa.
2016, 42(9):1079-1089.
DOI: 10.7519/j.issn.1000-0526.2016.09.005
Abstract:
Cold vortex is identified based on the definition of cold vortex, and 73 cold vertexes during 2008-2013 are identified. According to the screening criteria of squall line, the qualified squall lines are screened out. Digital radar mosaics of composite radar reflectivity are used in this study to perform the survey. Detailed analyses are carried out on the spatiotemporal distribution, moving characteristics, formation, dissipating way of squall lines and the relationship between cold vortex and squall lines. The results show that (1) 73 cold vortexes and 17 squall lines during 2008-2013 are identified, which mainly formed in Huaihe River Basin, North China and Northeast China. (2) Squall lines have obvious features of monthly and diurnal variations, with nearly 58.8% in July, and 52.9% from afternoon to evening. (3) Squall lines form mainly in the south of the cold vortex. During the different stages of cold vortex, squall line develop in different positions. (4) Some squall lines move towards eastsouth influenced by the midlatitude westerlies weather system, and others move from northwest to southeast. Due to the influence of the cold vortex system, squall lines move fast. (5) Most squall lines under the background of cold vertex are accompanied by heavy rainfall. The main forming mode is embedded areal (EA), the main dissipation mode is reversed broken areal (RBA), and EA type squall lines are often seen in highly humid environment.
Cold vortex is identified based on the definition of cold vortex, and 73 cold vertexes during 2008-2013 are identified. According to the screening criteria of squall line, the qualified squall lines are screened out. Digital radar mosaics of composite radar reflectivity are used in this study to perform the survey. Detailed analyses are carried out on the spatiotemporal distribution, moving characteristics, formation, dissipating way of squall lines and the relationship between cold vortex and squall lines. The results show that (1) 73 cold vortexes and 17 squall lines during 2008-2013 are identified, which mainly formed in Huaihe River Basin, North China and Northeast China. (2) Squall lines have obvious features of monthly and diurnal variations, with nearly 58.8% in July, and 52.9% from afternoon to evening. (3) Squall lines form mainly in the south of the cold vortex. During the different stages of cold vortex, squall line develop in different positions. (4) Some squall lines move towards eastsouth influenced by the midlatitude westerlies weather system, and others move from northwest to southeast. Due to the influence of the cold vortex system, squall lines move fast. (5) Most squall lines under the background of cold vertex are accompanied by heavy rainfall. The main forming mode is embedded areal (EA), the main dissipation mode is reversed broken areal (RBA), and EA type squall lines are often seen in highly humid environment.
2016, 42(9):1090-1095.
DOI: 10.7519/j.issn.1000-0526.2016.09.006
Abstract:
Applications of vertical velocity in forecast are studied by analyzing the vertical velocity data from Zhejiang wind profiler radars (WPR). Based on falling speed data of precipitation particles in different precipitation phases, the vertical speed characteristics of different precipitation phases are established. In terms of the analysis of radiosonde and surface observation data of 4 winters during 2011-2014 in Zhejiang Province, 2 m temperature, 1000 hPa temperature, height difference between 850 hPa and 1000 hPa and the height of 0℃ layer are suggested as the best identification indexes for identifying winter precipitation types. And the best identification criterion is given at which TS score can be greater than 0.8. The TS scores of these indexes are poor to use, but when combined with index of middlelevel warm layer, they can be improved obviously.
Applications of vertical velocity in forecast are studied by analyzing the vertical velocity data from Zhejiang wind profiler radars (WPR). Based on falling speed data of precipitation particles in different precipitation phases, the vertical speed characteristics of different precipitation phases are established. In terms of the analysis of radiosonde and surface observation data of 4 winters during 2011-2014 in Zhejiang Province, 2 m temperature, 1000 hPa temperature, height difference between 850 hPa and 1000 hPa and the height of 0℃ layer are suggested as the best identification indexes for identifying winter precipitation types. And the best identification criterion is given at which TS score can be greater than 0.8. The TS scores of these indexes are poor to use, but when combined with index of middlelevel warm layer, they can be improved obviously.
2016, 42(9):1096-1104.
DOI: 10.7519/j.issn.1000-0526.2016.09.007
Abstract:
The WRF/Chem model coupled with anthropogenic atmospheric pollution emission data from NCAR is used to simulate characteristics of one pollution event in Beijing on 31 August 2012. The results show that ozone (O3) concentrations have obvious diurnal variation characteristics. The high and low peaks appear in the afternoon and night respectively. The spatial distribution and flow field are related. PM2.5 also has the diurnal variation, but the spatial distribution is relatively stable and mainly in urban areas. Local pollutants have horizontal transport characteristics but the total amount is about zero, and it has not increased the local accumulation of pollutants. Long distance transportation from southern Beijing is one of the main sources of the pollutant. Sensitivity test analysis shows that the O3 pollution of the process is mainly from Beijing, and the fine particles are mainly generated locally. The sensitivity of O3 pollution to the nitrogen oxide (NOx) is relatively stable, but varies greatly to volatile organic compounds, especially from wee hours to morning.
The WRF/Chem model coupled with anthropogenic atmospheric pollution emission data from NCAR is used to simulate characteristics of one pollution event in Beijing on 31 August 2012. The results show that ozone (O3) concentrations have obvious diurnal variation characteristics. The high and low peaks appear in the afternoon and night respectively. The spatial distribution and flow field are related. PM2.5 also has the diurnal variation, but the spatial distribution is relatively stable and mainly in urban areas. Local pollutants have horizontal transport characteristics but the total amount is about zero, and it has not increased the local accumulation of pollutants. Long distance transportation from southern Beijing is one of the main sources of the pollutant. Sensitivity test analysis shows that the O3 pollution of the process is mainly from Beijing, and the fine particles are mainly generated locally. The sensitivity of O3 pollution to the nitrogen oxide (NOx) is relatively stable, but varies greatly to volatile organic compounds, especially from wee hours to morning.
2016, 42(9):1105-1113.
DOI: 10.7519/j.issn.1000-0526.2016.09.008
Abstract:
The features and formation of a severe convection process are studied by using the conventional observation data, automatic weather station data, GFS analysis field data with special resolution of 0.5° and temporal resolution of 6 h, Doppler radar, radar wind profiler and microwave radiometer data collected in Zhejiang Province on 26 July 2014. The results suggest that the severe convection process near the edge of subtropical high was due to the penetration of cold air caused by the east movement of the upper trough and accumulation of unstable energy, as well as the ambient background of dry and cold in upper layer but wet and warm in low layer, such unstable stratification condition and background was favorable for occurrence of the severe convection. Good predictability was provided when lowlevel 925 hPa mesoscale convergence line overlaid the vertical upward region in the middle of troposphere at 700 hPa. This means that severe convection could happen near the mesoscale convergence line in next 6 h under such situation. The Ningbo mesoscale convergence line, which was produced by confluence of southerly flow and northeasterly flow, triggered the severe convection in Ningbo. What’s more, the convergence line already existed under the orographic influence of coastal line and Hangzhou Bay. In contrast, the Hangzhou mesoscale convergence line, resulting from the convergence of ambient northeasterly current and easterly airflow enhanced by surface outflow of Ningbo thunderstorm, gave rise to the outbreak of the severe convection around Hangzhou.
The features and formation of a severe convection process are studied by using the conventional observation data, automatic weather station data, GFS analysis field data with special resolution of 0.5° and temporal resolution of 6 h, Doppler radar, radar wind profiler and microwave radiometer data collected in Zhejiang Province on 26 July 2014. The results suggest that the severe convection process near the edge of subtropical high was due to the penetration of cold air caused by the east movement of the upper trough and accumulation of unstable energy, as well as the ambient background of dry and cold in upper layer but wet and warm in low layer, such unstable stratification condition and background was favorable for occurrence of the severe convection. Good predictability was provided when lowlevel 925 hPa mesoscale convergence line overlaid the vertical upward region in the middle of troposphere at 700 hPa. This means that severe convection could happen near the mesoscale convergence line in next 6 h under such situation. The Ningbo mesoscale convergence line, which was produced by confluence of southerly flow and northeasterly flow, triggered the severe convection in Ningbo. What’s more, the convergence line already existed under the orographic influence of coastal line and Hangzhou Bay. In contrast, the Hangzhou mesoscale convergence line, resulting from the convergence of ambient northeasterly current and easterly airflow enhanced by surface outflow of Ningbo thunderstorm, gave rise to the outbreak of the severe convection around Hangzhou.
2016, 42(9):1114-1123.
DOI: 10.7519/j.issn.1000-0526.2016.09.009
Abstract:
Using ECMWF Ensemble data, the performances of ensemble forecasting about largescale circulation and heavy rain processes in the midlower reaches of the Yangtze River during the drought and flood transition period in 2011 are examined. The results show that ensemble average forecasting has a good performance on the adjustments of largescale circulation in the whole troposphere in extendedrange period of validity. The forecasting period is up to 10-15 days. During different forecasting periods, ensemble average forecasting displays stability of forecasting the main affecting systems during the heavy rain processes. With the shortening of the forecasting period, the dispersion of various ensemble members to weather systems gradually decreases. The forecasting of individual ensemble member including ensemble average and control member has not obvious regularity in different period of validity. The forecast of a few ensemble members is in accord with the real weather condition about main impact systems during different forecasting periods. Occurrence of lowlevel stronger winds is closely related to heavy rainfall in midlower reaches of the Yangtze River. The small probability information of ensemble forecasting is meaningful in the extendedrange forecasting period. It can provide a reference for extendedrange heavy rainfall forecast.
Using ECMWF Ensemble data, the performances of ensemble forecasting about largescale circulation and heavy rain processes in the midlower reaches of the Yangtze River during the drought and flood transition period in 2011 are examined. The results show that ensemble average forecasting has a good performance on the adjustments of largescale circulation in the whole troposphere in extendedrange period of validity. The forecasting period is up to 10-15 days. During different forecasting periods, ensemble average forecasting displays stability of forecasting the main affecting systems during the heavy rain processes. With the shortening of the forecasting period, the dispersion of various ensemble members to weather systems gradually decreases. The forecasting of individual ensemble member including ensemble average and control member has not obvious regularity in different period of validity. The forecast of a few ensemble members is in accord with the real weather condition about main impact systems during different forecasting periods. Occurrence of lowlevel stronger winds is closely related to heavy rainfall in midlower reaches of the Yangtze River. The small probability information of ensemble forecasting is meaningful in the extendedrange forecasting period. It can provide a reference for extendedrange heavy rainfall forecast.
2016, 42(9):1124-1134.
DOI: 10.7519/j.issn.1000-0526.2016.09.010
Abstract:
To increase the efficiency in artificial hail suppression operation, a storm autoclassification technology is developed based on the storm tracking and recognition algorithm using the new generation Doppler weather radar data in this study. The characteristic indices of storm structure are firstly automatically extracted by using the radar and sounding data and the SCIT algorithm. And then according to the intensities, the storms are classified into weak thunderstorm, singlecell storm, multicell storm and severe storm by adopting the automatic classification technology of decision tree model. Finally, the early warnings on the downstream direction of the storm or working parameters near the operating location are automatically performed according to properties of the storm such as storm intensity, height, location and GIS information. From the analysis of 182 hail cases during the 31 hail weather processes over Chongqing, Dalian of Liaoning and Sanmenxia of Henan from 2006 to 2014, the storm autoclassification technology developed in current study can significantly increase (decrease) the hit (false alarm) rate of storm tracking and recognition. The hit rate can reach 100%, and the false alarm rate is only 11.4%. The result suggests that this storm autoclassification technology can enhance the automation of artificial hail prevention and contribute to the decision making for the operation of artificial hail suppression.
To increase the efficiency in artificial hail suppression operation, a storm autoclassification technology is developed based on the storm tracking and recognition algorithm using the new generation Doppler weather radar data in this study. The characteristic indices of storm structure are firstly automatically extracted by using the radar and sounding data and the SCIT algorithm. And then according to the intensities, the storms are classified into weak thunderstorm, singlecell storm, multicell storm and severe storm by adopting the automatic classification technology of decision tree model. Finally, the early warnings on the downstream direction of the storm or working parameters near the operating location are automatically performed according to properties of the storm such as storm intensity, height, location and GIS information. From the analysis of 182 hail cases during the 31 hail weather processes over Chongqing, Dalian of Liaoning and Sanmenxia of Henan from 2006 to 2014, the storm autoclassification technology developed in current study can significantly increase (decrease) the hit (false alarm) rate of storm tracking and recognition. The hit rate can reach 100%, and the false alarm rate is only 11.4%. The result suggests that this storm autoclassification technology can enhance the automation of artificial hail prevention and contribute to the decision making for the operation of artificial hail suppression.
2016, 42(9):1135-1145.
DOI: 10.7519/j.issn.1000-0526.2016.09.011
Abstract:
Based on daily data of precipitation, air pressure, temperature, average vapor pressure and average relative humidity in 622 stations released by the National Meteorological Information Centre of China during 1960-2013 and the data collected at 02:00, 08:00, 14:00 and 20:00 BT from 25 August to 25 September 2013 by 249 stations, the algorithms of China Meteorological Administration, Barnes and Georgakakos are applied to calculate the height of lifting condensation level to approximate height of precipitation cloud base, and then the spatiotemporal distribution characteristics of cloud base height in China are analyzed. The results indicate that (1) the spatial distribution of precipitation cloud base heights on the annual, seasonally and hourly bases show consistent with precipitation levels Ⅰ and Ⅱ. The cloud base height gradually increases from southeast to northwest, and the interannual difference is not significant. (2) The regional difference is significant. In term of four nature zones in China (North, South, Northwest and the Tibetan Plateau), the precipitation cloud base heights of the Northwest and the Tibetan Plateau are higher than the national mean. The base height decreases gradually year by year except the Tibetan Plateau. (3) Seasonally, the precipitation cloud base height in spring and summer is higher, and the lowest is in winter. (4) Diurnal variation of the precipitation cloud base height is significant and the lowest at 08:00 and the highest is at 14:00. (5) By using the three algorithms, the correlation coefficients between the cloud base height and precipitation are -0.47, -0.46 and -0.44, and between the cloud base height and relative humidity are -0.81, -0.81 and -0.79, respectively.
Based on daily data of precipitation, air pressure, temperature, average vapor pressure and average relative humidity in 622 stations released by the National Meteorological Information Centre of China during 1960-2013 and the data collected at 02:00, 08:00, 14:00 and 20:00 BT from 25 August to 25 September 2013 by 249 stations, the algorithms of China Meteorological Administration, Barnes and Georgakakos are applied to calculate the height of lifting condensation level to approximate height of precipitation cloud base, and then the spatiotemporal distribution characteristics of cloud base height in China are analyzed. The results indicate that (1) the spatial distribution of precipitation cloud base heights on the annual, seasonally and hourly bases show consistent with precipitation levels Ⅰ and Ⅱ. The cloud base height gradually increases from southeast to northwest, and the interannual difference is not significant. (2) The regional difference is significant. In term of four nature zones in China (North, South, Northwest and the Tibetan Plateau), the precipitation cloud base heights of the Northwest and the Tibetan Plateau are higher than the national mean. The base height decreases gradually year by year except the Tibetan Plateau. (3) Seasonally, the precipitation cloud base height in spring and summer is higher, and the lowest is in winter. (4) Diurnal variation of the precipitation cloud base height is significant and the lowest at 08:00 and the highest is at 14:00. (5) By using the three algorithms, the correlation coefficients between the cloud base height and precipitation are -0.47, -0.46 and -0.44, and between the cloud base height and relative humidity are -0.81, -0.81 and -0.79, respectively.
2016, 42(9):1146-1153.
DOI: 10.7519/j.issn.1000-0526.2016.09.012
Abstract:
The European Centre for MediumRange Weather Forecasts (ECMWF) highresolution deterministic forecasts (HDet) and probability forecasts from the ensemble prediction system (EPS) are evaluated and compared for the heavy rainfall from May to September 2012 in Huaihe River Basin, China. For the EPS, the daily precipitation extreme forecast index (EFI) and the calibrated probability calculated by the modified Bayesian model averaging (MBMA) are used. Relative operating characteristic (ROC) verification which is independent of forecast bias is used. The TS vs. false alarm rate (FAR) and the relative economic value are measured. The results indicate that the MBMA has the best performance, followed by the HDet and then the EFI. For the forecast within lead times of 2 days, the HDet is close to MBMA. As the lead times prolong, the advantage of MBMA relative to HDet keeps increasing. Under different criteria, the heavy precipitation forecast thresholds for the three methods are estimated. It is shown that the MBMA is the best and the EFI is the worst. However, the advantage of MBMA is accompanied by higher forecast bias. If the forecast bias is set to be 1.37, which is equal to the subjective forecast of National Meteorological Centre (NMC) forecasters, the MBMA is almost similar to HDet.
The European Centre for MediumRange Weather Forecasts (ECMWF) highresolution deterministic forecasts (HDet) and probability forecasts from the ensemble prediction system (EPS) are evaluated and compared for the heavy rainfall from May to September 2012 in Huaihe River Basin, China. For the EPS, the daily precipitation extreme forecast index (EFI) and the calibrated probability calculated by the modified Bayesian model averaging (MBMA) are used. Relative operating characteristic (ROC) verification which is independent of forecast bias is used. The TS vs. false alarm rate (FAR) and the relative economic value are measured. The results indicate that the MBMA has the best performance, followed by the HDet and then the EFI. For the forecast within lead times of 2 days, the HDet is close to MBMA. As the lead times prolong, the advantage of MBMA relative to HDet keeps increasing. Under different criteria, the heavy precipitation forecast thresholds for the three methods are estimated. It is shown that the MBMA is the best and the EFI is the worst. However, the advantage of MBMA is accompanied by higher forecast bias. If the forecast bias is set to be 1.37, which is equal to the subjective forecast of National Meteorological Centre (NMC) forecasters, the MBMA is almost similar to HDet.
2016, 42(9):1154-1160.
DOI: 10.7519/j.issn.1000-0526.2016.09.013
Abstract:
The main characteristics of the general atmospheric circulation in June 2016 are as follow： the circulation of polar vortex in the Northern Hemisphere was in a singlepole pattern, which is stronger than normal years. The 500 hPa geopotential height presented the distribution of a fourwave pattern in the high latitude of Northern Hemisphere. The subtropical high of Northwest Pacific was stronger and larger than the climatological normal. The monthly mean temperature was 20.7℃， 0.7℃ higher than the corresponding period of normal years. The mean precipitation was 117.0 mm, 17.2% more than normal 99.8 mm. The temperature and precipitation are both the third highest values since 1961. Seven rainfall processes occurred this month. Torrential rainfalls occurred frequently in the regions of Yangtze River and Huaihe River as well as the northern part of South Yangtze River Region, causing many places in the middle and lower reaches of Yangtze River to have suffered from serious floods. Meanwhile, severe convective weather was seen in North China, the regions of Yellow River and Huaihe River, and the northern part of Yangtze River. One severe tornado disaster occurred in Yancheng of Jiangsu Province, causing serious casualties. None tropical cyclone was initiated over the Northwest Pacific.
The main characteristics of the general atmospheric circulation in June 2016 are as follow： the circulation of polar vortex in the Northern Hemisphere was in a singlepole pattern, which is stronger than normal years. The 500 hPa geopotential height presented the distribution of a fourwave pattern in the high latitude of Northern Hemisphere. The subtropical high of Northwest Pacific was stronger and larger than the climatological normal. The monthly mean temperature was 20.7℃， 0.7℃ higher than the corresponding period of normal years. The mean precipitation was 117.0 mm, 17.2% more than normal 99.8 mm. The temperature and precipitation are both the third highest values since 1961. Seven rainfall processes occurred this month. Torrential rainfalls occurred frequently in the regions of Yangtze River and Huaihe River as well as the northern part of South Yangtze River Region, causing many places in the middle and lower reaches of Yangtze River to have suffered from serious floods. Meanwhile, severe convective weather was seen in North China, the regions of Yellow River and Huaihe River, and the northern part of Yangtze River. One severe tornado disaster occurred in Yancheng of Jiangsu Province, causing serious casualties. None tropical cyclone was initiated over the Northwest Pacific.
Mobile website
® 2024 All Rights Reserved
Supported by:Beijing E-Tiller Technology Development Co., Ltd.
ISSN:1000-0526 CN:11-2282/P