ISSN 1000-0526
CN 11-2282/P
CN 11-2282/P
Volume 46,Issue 3,2020 Table of Contents
2020, 46(3):289-300.
DOI: 10.7519/j.issn.1000-0526.2020.03.001
Abstract:
Water cycle, which is significantly affected by climate change， is the most active and important part in the interaction among ocean, land and atmosphere. This paper reviews the impact of climate change on global water cycle, and evaluates the changes of water cycle elements including atmospheric water vapor, precipitation, evapotranspiration and surface runoff and water vapor in China since the 1960s. Climate warming has strengthened the global water cycle, showing that global nearsurface and tropospheric speci fic humidity have increased since 1970s. However, due to the insufficient observation, the trend of various variables in the water cycle still remains considerably uncertain. Moreover, due to interaction between the increase of tropical water vapor and the weakening of the circulation, the influence of climate change on regional water cycle is more uncertain. Obvious changes of regional water cycle in China have been found since 1960s. The atmospheric water vapor and its budget show up an increasing trend after 1980s. No obvious changes in total precipitation, but largely spatial difference of regional precipitation can be found in China since 1960s. Moreover, average evapotranspiration increases slightly, while the mean surface runoff decreases. Since the 21st Century, the precipitation transformed from evapotranspiration has been increa sing, and the internal water cycle is more active than before.
Water cycle, which is significantly affected by climate change， is the most active and important part in the interaction among ocean, land and atmosphere. This paper reviews the impact of climate change on global water cycle, and evaluates the changes of water cycle elements including atmospheric water vapor, precipitation, evapotranspiration and surface runoff and water vapor in China since the 1960s. Climate warming has strengthened the global water cycle, showing that global nearsurface and tropospheric speci fic humidity have increased since 1970s. However, due to the insufficient observation, the trend of various variables in the water cycle still remains considerably uncertain. Moreover, due to interaction between the increase of tropical water vapor and the weakening of the circulation, the influence of climate change on regional water cycle is more uncertain. Obvious changes of regional water cycle in China have been found since 1960s. The atmospheric water vapor and its budget show up an increasing trend after 1980s. No obvious changes in total precipitation, but largely spatial difference of regional precipitation can be found in China since 1960s. Moreover, average evapotranspiration increases slightly, while the mean surface runoff decreases. Since the 21st Century, the precipitation transformed from evapotranspiration has been increa sing, and the internal water cycle is more active than before.
2 Refined Analysis of SpatioTemporal Characteristics of Thunderstorm and Hail over Qinghai Province
2020, 46(3):301-312.
DOI: 10.7519/j.issn.1000-0526.2020.03.002
Abstract:
Based on a comprehensive collection of thunderstorm and hail observation data from the National Meteorological Information Center (NMIC) of China, the hourly and duration datasets which include the onset and end time of thunderstorm and hail were established. Then the characteristics of spatiotemporal distribution, duration and their longterm trends over Qinghai Province from 1981 to 2011 were documented in this study. The results showed as follows. The frequencies of thunderstorm and hail are significantly influenced by the topography, so both of them occur more frequently in the south part than in the north part over Qinghai Province. Generally speaking, the duration of a thunderstorm process does not last for longer than 40 minutes, and the duration of the hailstorm process does not exceed 10 minutes. The frequencies of thunderstorm and hail both present significant monthly and diurnal variations. In a year, they are mainly concentrated in May to September; in the daytime, they are mainly concentrated in the afternoon, but the occurring time shows an obvious decaying from north to south, and the peak time in the north is about 3 h earlier than in the south part of Qinghai Province. With the duration increasing, the number of thunderstorm processes shows an efolding decay derived from the exponential distribution, while the number of hailstorm processes increases first and then decreases according to the efolding decay. The frequencies of thunderstorm and hail both have presented significant decreasing tendencies since 1981, and the climatic tendencies of thunderstorm and hail are -15.0·(10 a)-1 and -2.3·(10 a)-1, respectively. Although the number of thunderstorm process is decreasing, the average duration of each process shows a slightly increasing trend. And the increasingduration stations are mainly located in the northeastern and central parts of Qinghai, which is the most densely populated area, implying that the potential damage caused by thunderstorm is increasing. Different from that of thunderstorm, the number of hail process and its duration both show decreasing trends, which may be related to the increased height of the melting layer.
Based on a comprehensive collection of thunderstorm and hail observation data from the National Meteorological Information Center (NMIC) of China, the hourly and duration datasets which include the onset and end time of thunderstorm and hail were established. Then the characteristics of spatiotemporal distribution, duration and their longterm trends over Qinghai Province from 1981 to 2011 were documented in this study. The results showed as follows. The frequencies of thunderstorm and hail are significantly influenced by the topography, so both of them occur more frequently in the south part than in the north part over Qinghai Province. Generally speaking, the duration of a thunderstorm process does not last for longer than 40 minutes, and the duration of the hailstorm process does not exceed 10 minutes. The frequencies of thunderstorm and hail both present significant monthly and diurnal variations. In a year, they are mainly concentrated in May to September; in the daytime, they are mainly concentrated in the afternoon, but the occurring time shows an obvious decaying from north to south, and the peak time in the north is about 3 h earlier than in the south part of Qinghai Province. With the duration increasing, the number of thunderstorm processes shows an efolding decay derived from the exponential distribution, while the number of hailstorm processes increases first and then decreases according to the efolding decay. The frequencies of thunderstorm and hail both have presented significant decreasing tendencies since 1981, and the climatic tendencies of thunderstorm and hail are -15.0·(10 a)-1 and -2.3·(10 a)-1, respectively. Although the number of thunderstorm process is decreasing, the average duration of each process shows a slightly increasing trend. And the increasingduration stations are mainly located in the northeastern and central parts of Qinghai, which is the most densely populated area, implying that the potential damage caused by thunderstorm is increasing. Different from that of thunderstorm, the number of hail process and its duration both show decreasing trends, which may be related to the increased height of the melting layer.
2020, 46(3):313-324.
DOI: 10.7519/j.issn.1000-0526.2020.03.003
Abstract:
The warmarea rainstorm has always been a difficult point in operational forecasting, and forecasting the rainstorm in the double rain belt is even more difficult. There was a severe double belt rainstorm from 20:00 BT 19 to 20:00 BT 20 April 2016, which was missing in forecasting. In order to provide lessons for future forecasts, based on the numerical forecast products, NCEP 1°×1 ° global objective analysis data, conventional meteorological data, the severe double rain belt were analyzed. The results are as follows. The predicted 500 hPa trough was far from Yulin, the convergence line at 925 hPa was not obvious, and the rainfall was small. The effect of cooling trough in front of the south branch trough was neglected. The cold air was misjudged into the sea, and the effect of backflow on Yulin was ignored. The CAPE value was not analyzed in detail and it was not found that before the rainstorm Yulin had favorable dynamic conditions for “dry and wet”, which was easy to trigger rainstorm. The energy frontogenesis contributed to the rise of convective instability. The lowlevel jet provided a continuous flow of water vapor for the south rain belt and the large area of water vapor flux and the convergence area of water vapor are concentrated in Yulin, which is conducive to the occurrence and development of rainstorms.
The warmarea rainstorm has always been a difficult point in operational forecasting, and forecasting the rainstorm in the double rain belt is even more difficult. There was a severe double belt rainstorm from 20:00 BT 19 to 20:00 BT 20 April 2016, which was missing in forecasting. In order to provide lessons for future forecasts, based on the numerical forecast products, NCEP 1°×1 ° global objective analysis data, conventional meteorological data, the severe double rain belt were analyzed. The results are as follows. The predicted 500 hPa trough was far from Yulin, the convergence line at 925 hPa was not obvious, and the rainfall was small. The effect of cooling trough in front of the south branch trough was neglected. The cold air was misjudged into the sea, and the effect of backflow on Yulin was ignored. The CAPE value was not analyzed in detail and it was not found that before the rainstorm Yulin had favorable dynamic conditions for “dry and wet”, which was easy to trigger rainstorm. The energy frontogenesis contributed to the rise of convective instability. The lowlevel jet provided a continuous flow of water vapor for the south rain belt and the large area of water vapor flux and the convergence area of water vapor are concentrated in Yulin, which is conducive to the occurrence and development of rainstorms.
4 Statistical Characteristics and Convection Indexes of Thunderstorm and Gale over Western Bohai Sea
2020, 46(3):325-335.
DOI: 10.7519/j.issn.1000-0526.2020.03.004
Abstract:
The statistical characteristics of thunderstorm and gale over western Bohai Sea were analyzed using dense observation data of seven stations at islands， platforms and buoy stations in western Bohai Sea from May to September during 2011-2016. The results showed that the thunderstorm and gale forces in western Bohai Sea are mainly the maganitude of 8-9, which are seen in the more southern part of the western Bohai Sea， mostly in June， then in July and rarely in May. The process in June, mostly occuring from 19:00 BT to 23:00 BT, is related to cold vortex twice as many as highlevel trough, and the most is related to band echo， which accounts for 68.8%. The sea surface temperature is 6℃ per longitude lower than the coastal temperature in 6-12 h before the occurrence of 〖JP2〗thunderstorm and gale, but the sea surface temperature and pressure are higher 4℃ per longitude and lower 2 hPa per longitude than in coastal area in 1-2 h before thunderstorm and gale. At the same time， the index thresholds of thunderstorm and gale were analyzed using corrected sounding datas. The results showed that BCAPE，LI，WINDEX and the 0-3 km vertical wind shear are more obvious in western Bohai Sea than in North China. The convection index thresholds of BCAPE， LI， WINDEX and 0-3 km vertical wind shear which could offer references for the forecast of thunderstorm and gale in western Bohai Sea from June to September are given, under background of highlevel trough and cold vortex.
The statistical characteristics of thunderstorm and gale over western Bohai Sea were analyzed using dense observation data of seven stations at islands， platforms and buoy stations in western Bohai Sea from May to September during 2011-2016. The results showed that the thunderstorm and gale forces in western Bohai Sea are mainly the maganitude of 8-9, which are seen in the more southern part of the western Bohai Sea， mostly in June， then in July and rarely in May. The process in June, mostly occuring from 19:00 BT to 23:00 BT, is related to cold vortex twice as many as highlevel trough, and the most is related to band echo， which accounts for 68.8%. The sea surface temperature is 6℃ per longitude lower than the coastal temperature in 6-12 h before the occurrence of 〖JP2〗thunderstorm and gale, but the sea surface temperature and pressure are higher 4℃ per longitude and lower 2 hPa per longitude than in coastal area in 1-2 h before thunderstorm and gale. At the same time， the index thresholds of thunderstorm and gale were analyzed using corrected sounding datas. The results showed that BCAPE，LI，WINDEX and the 0-3 km vertical wind shear are more obvious in western Bohai Sea than in North China. The convection index thresholds of BCAPE， LI， WINDEX and 0-3 km vertical wind shear which could offer references for the forecast of thunderstorm and gale in western Bohai Sea from June to September are given, under background of highlevel trough and cold vortex.
2020, 46(3):336-345.
DOI: 10.7519/j.issn.1000-0526.2020.03.005
Abstract:
In this study, the specific methods of estimating surface longwave radiation fluxes under allsky condition are elucidated. An inverse model for cloudysky surface downward longwave radiation (DLR) flux retrieval is developed based on infrared radiative transfer calculation, statistical regression analysis of simulated data, and atmospheric profiles. After that, we combine the inverse model and GFS data to calculate FY4A cleansky and cloudysky DLR values. Then, on the basis of the Aqua/CERES global DLR product, it validates the corresponding product of FY4A during day or night time. The final results show that RMSE=20.52 W·m-2, R=0.9481， BIAS=3.3 W·m-2 (nighttime); RMSE=25.58 W·m-2，R=0.9096， BIAS=5.4 W·m-2 (daytime). Besides, we also successfully calculate the cloudysky upwelling longwave radiation (ULR) product at surface, which is combined with the operational FY4A clearsky ULR product to make a joint allsky product. The validations of FY4A allsky ULR show that RMSE=10.97 W·m-2，R=0.9762，BIAS=-3.3 W·m-2 (nighttime); RMSE=19.97 W·m-2, R=0.9283，BIAS=5.0 W·m-2 (daytime). The validation results mentionedabove indicate the retrieval methods developed in this research are feasible and highquality, which could be used for retrieving new operational DLR/ULR products for FY4 series satellites in future.
In this study, the specific methods of estimating surface longwave radiation fluxes under allsky condition are elucidated. An inverse model for cloudysky surface downward longwave radiation (DLR) flux retrieval is developed based on infrared radiative transfer calculation, statistical regression analysis of simulated data, and atmospheric profiles. After that, we combine the inverse model and GFS data to calculate FY4A cleansky and cloudysky DLR values. Then, on the basis of the Aqua/CERES global DLR product, it validates the corresponding product of FY4A during day or night time. The final results show that RMSE=20.52 W·m-2, R=0.9481， BIAS=3.3 W·m-2 (nighttime); RMSE=25.58 W·m-2，R=0.9096， BIAS=5.4 W·m-2 (daytime). Besides, we also successfully calculate the cloudysky upwelling longwave radiation (ULR) product at surface, which is combined with the operational FY4A clearsky ULR product to make a joint allsky product. The validations of FY4A allsky ULR show that RMSE=10.97 W·m-2，R=0.9762，BIAS=-3.3 W·m-2 (nighttime); RMSE=19.97 W·m-2, R=0.9283，BIAS=5.0 W·m-2 (daytime). The validation results mentionedabove indicate the retrieval methods developed in this research are feasible and highquality, which could be used for retrieving new operational DLR/ULR products for FY4 series satellites in future.
2020, 46(3):346-356.
DOI: 10.7519/j.issn.1000-0526.2020.03.006
Abstract:
Combining the gauged precipitation and the predicted precipitation in 24 h and 48 h of regional numerical model named GRAPES during 2016 to 2017, GRAPES was quantitatively evaluated in the four provinces （regions） of Northwest China, using the indices including mean error, root mean square error, correlation coefficient and TS score. The results showed that the forecast accuracy of rain probability was higher than 0.7, daily spatial correlation coefficient was 0.2-0.4. The highest bias appeared in summer, the mean errors of 24 h and 48 h forecast were 4 mm·d-1 and 6 mm·d-1, the root mean square errorr were 6 mm·d-1 and 8 mm·d-1, respectively. The TS of heavy rain and above was less than 0.1, TSs of light rain and moderate rain were 0.2-0.5 and 0.1-0.2, respectively. Spatially, the 24 h and 48 h forecast accuracies of rain probability were higher than 0.6 in most regions, the correlation coefficients in eastern Gansu, middle and southern Shaanxi were higher than 0.6. The highest mean error of 24 h forecast appeared in the southern part of Qinghai, Gansu and Shaanxi, which reached to 2-4 mm·d-1. The highest mean error of 48 h forecast reached to 5~8 mm·d-1 and appeared in the southern Shaanxi. The mean error of 48 h forecast was 1~2 mm·d-1 higher than 24 h forecast in other regions, The TS score of 24 h forecast for each precipitation level was obviously better than that of 48 h, 24 h forecast could predict heavy rain and rainstorm, while 48 h forecast showed poor ability for moderate rain and above.
Combining the gauged precipitation and the predicted precipitation in 24 h and 48 h of regional numerical model named GRAPES during 2016 to 2017, GRAPES was quantitatively evaluated in the four provinces （regions） of Northwest China, using the indices including mean error, root mean square error, correlation coefficient and TS score. The results showed that the forecast accuracy of rain probability was higher than 0.7, daily spatial correlation coefficient was 0.2-0.4. The highest bias appeared in summer, the mean errors of 24 h and 48 h forecast were 4 mm·d-1 and 6 mm·d-1, the root mean square errorr were 6 mm·d-1 and 8 mm·d-1, respectively. The TS of heavy rain and above was less than 0.1, TSs of light rain and moderate rain were 0.2-0.5 and 0.1-0.2, respectively. Spatially, the 24 h and 48 h forecast accuracies of rain probability were higher than 0.6 in most regions, the correlation coefficients in eastern Gansu, middle and southern Shaanxi were higher than 0.6. The highest mean error of 24 h forecast appeared in the southern part of Qinghai, Gansu and Shaanxi, which reached to 2-4 mm·d-1. The highest mean error of 48 h forecast reached to 5~8 mm·d-1 and appeared in the southern Shaanxi. The mean error of 48 h forecast was 1~2 mm·d-1 higher than 24 h forecast in other regions, The TS score of 24 h forecast for each precipitation level was obviously better than that of 48 h, 24 h forecast could predict heavy rain and rainstorm, while 48 h forecast showed poor ability for moderate rain and above.
2020, 46(3):357-366.
DOI: 10.7519/j.issn.1000-0526.2020.03.007
Abstract:
Using four-dimensional variational Doppler radar assimilation system (VDRAS) simulation and radar mosaic data, a weak squall line influenced by northeast cold vortex on 16 May 2012 was analyzed. The result showed that strong convergence line on the surface was an important factor causing convective occurrence under the weak hydrostatic instability, and was also an important reason for the quickly developing of the system at the early stage. Cold air was continuously transported to Yangtze-Huaihe River Valley under the control of northeast cold vortex. Because of precipitation dragging by convection within squall line, the dry and cold air was entrained into convection and continuously moved downward, causing strong cold pool as the result of evaporation. The main reason for weak squall line is that after the formation of cool pool, vertical wind shear is not perpendicular to the squall line with weak magnitude, causing imbalance between them. Besides, weak 〖JP2〗hydrostatic instability and low relative humidity are part of reasons for the weak squall line. The major difference between this case and other squall lines in eastern China is the strong convergence line and relative dry environment, which means less moisture content is delivered to middle level and less latent heating is released. Finally, upward motion is more dependent on balance between cool pool and vertical wind shear.
Using four-dimensional variational Doppler radar assimilation system (VDRAS) simulation and radar mosaic data, a weak squall line influenced by northeast cold vortex on 16 May 2012 was analyzed. The result showed that strong convergence line on the surface was an important factor causing convective occurrence under the weak hydrostatic instability, and was also an important reason for the quickly developing of the system at the early stage. Cold air was continuously transported to Yangtze-Huaihe River Valley under the control of northeast cold vortex. Because of precipitation dragging by convection within squall line, the dry and cold air was entrained into convection and continuously moved downward, causing strong cold pool as the result of evaporation. The main reason for weak squall line is that after the formation of cool pool, vertical wind shear is not perpendicular to the squall line with weak magnitude, causing imbalance between them. Besides, weak 〖JP2〗hydrostatic instability and low relative humidity are part of reasons for the weak squall line. The major difference between this case and other squall lines in eastern China is the strong convergence line and relative dry environment, which means less moisture content is delivered to middle level and less latent heating is released. Finally, upward motion is more dependent on balance between cool pool and vertical wind shear.
2020, 46(3):367-380.
DOI: 10.7519/j.issn.1000-0526.2020.03.008
Abstract:
GRAPES_3 km (Global/ Regional Assimilation and Prediction System） is a convection-permitting model, which provides an important objective basis for short-term forecast of severe weather. In this paper, the performance of GRAPES_3 km model in severe weather forecasting was comprehensively evaluated using traditional pixel-versus-pixel threat score, neighborhood and object-based methods. The applicability and differences of the traditional and new spatial verification methods for high resolution model assessment were analyzed, and the results were compared with those of GRAPES_Meso model. The results are showed as follows. Through case analysis, it is found that the characteristics of convective storm forecast and the evolution of convection 〖JP2〗can be comprehensively and objectively evaluated by various verification methods. GRAPES_ 3 km is superior to GRAPES_Meso in forecasting convective storms, especially severe storms over 50 dBz. The latest forecast of the initial forecast time is the best. Neighborhood TS method takes space-time deviation into account.The forecasting skill is the highest when the time neighborhood of GRAPES_3 km model is 1 h for 20 dBz and 35 dBz, and 3 h for 50 dBz. The fractions skill score (FSS) shows that GRAPES_3 km model can achieve the lowest forecast skill scale for convective storms with different thresholds, while GRAPES_Meso model usually fails to reach the lowest skill scale for storms above 35 dBz. Method for object-based diagnostic evaluation (MODE) can be used to evaluate the forecast of convective storm attributes. GRAPES_3 km model is consistent with the actual number of storms at all scales, but the area is obviously underestimated. The model can predict the shape and location of meso-β scale convective storms, while for meso-γ scale convective storms, the forecast scale is larger, the shape is more circular and the axis angle is smaller, but conversely for meso-α scale convective storms. The traditional pixel-versus-pixel verification method and new spatial verification methods have the same conclusion for convection-permitting model. Comparatively, the new spatial verification methods can provide more detailed information of convective storms.
GRAPES_3 km (Global/ Regional Assimilation and Prediction System） is a convection-permitting model, which provides an important objective basis for short-term forecast of severe weather. In this paper, the performance of GRAPES_3 km model in severe weather forecasting was comprehensively evaluated using traditional pixel-versus-pixel threat score, neighborhood and object-based methods. The applicability and differences of the traditional and new spatial verification methods for high resolution model assessment were analyzed, and the results were compared with those of GRAPES_Meso model. The results are showed as follows. Through case analysis, it is found that the characteristics of convective storm forecast and the evolution of convection 〖JP2〗can be comprehensively and objectively evaluated by various verification methods. GRAPES_ 3 km is superior to GRAPES_Meso in forecasting convective storms, especially severe storms over 50 dBz. The latest forecast of the initial forecast time is the best. Neighborhood TS method takes space-time deviation into account.The forecasting skill is the highest when the time neighborhood of GRAPES_3 km model is 1 h for 20 dBz and 35 dBz, and 3 h for 50 dBz. The fractions skill score (FSS) shows that GRAPES_3 km model can achieve the lowest forecast skill scale for convective storms with different thresholds, while GRAPES_Meso model usually fails to reach the lowest skill scale for storms above 35 dBz. Method for object-based diagnostic evaluation (MODE) can be used to evaluate the forecast of convective storm attributes. GRAPES_3 km model is consistent with the actual number of storms at all scales, but the area is obviously underestimated. The model can predict the shape and location of meso-β scale convective storms, while for meso-γ scale convective storms, the forecast scale is larger, the shape is more circular and the axis angle is smaller, but conversely for meso-α scale convective storms. The traditional pixel-versus-pixel verification method and new spatial verification methods have the same conclusion for convection-permitting model. Comparatively, the new spatial verification methods can provide more detailed information of convective storms.
ZHANG Tianhang
,
CHI Xiyuan
,
ZHANG Bihui
,
ZHANG Hengde
,
JIANG Qi
,
WANG Jikang
,
RAO Xiaoqin
,
XIE Chao
,
LYU Mengyao
,
AN Linchang
,
NAN Yang
2020, 46(3):381-392.
DOI: 10.7519/j.issn.1000-0526.2020.03.009
Abstract:
To decrease the forecast uncertainties of single models and improve the refinement of multi-model ensemble air quality forecast system, the gridding observed pollutant concentration with resolution of 0.25°×0.25° was firstly established by using Cressman interpolation method. Then, combined with four numerical air quality forecast models, the mean, weighted and multiple linear regression ensembles were established in each grid, respectively. Finally, based on the evaluation results of single models and ensemble methods in previous 50 days, an optimal ensemble was established. The evaluation results of PM2.5 concentrations during a heavy pollution process in 19-22 December 2018 showed that in the case of heavy pollution, the NMB values between the optimal ensemble forecast and observations could also be maintained between -20% and 40%. And the forecast coverage area with good and above pollution by the optimal ensemble was closer to observation than those of single models. During the whole process, the NMB, root mean squared error (RMSE) and R values between forecasted PM2.5 concentrations by the optimal ensemble and observation were from -20% to 20%, from 35 to 75 μg·m-3 and higher than 0.4, respectively, in most polluted areas. Among all single models and ensemble methods, number of girds over China with high total scores was the largest in optimal ensemble. In the eight cities located in the most polluted region, the average onset and end times of the pollution process by optimal ensemble forecast was 1.8 and 6.9 h earlier than observation, respectively. Therefore, we propose that pollutant concentrations retrieval by satellite and surface observation should be fused to improve the refinement of gridding observed pollutant concentrations. And the methods of scale reduction, subjective and objective fusion and rolling correction should be used to further improve the forecast accuracy of gridding multi-model ensemble air quality forecast.
To decrease the forecast uncertainties of single models and improve the refinement of multi-model ensemble air quality forecast system, the gridding observed pollutant concentration with resolution of 0.25°×0.25° was firstly established by using Cressman interpolation method. Then, combined with four numerical air quality forecast models, the mean, weighted and multiple linear regression ensembles were established in each grid, respectively. Finally, based on the evaluation results of single models and ensemble methods in previous 50 days, an optimal ensemble was established. The evaluation results of PM2.5 concentrations during a heavy pollution process in 19-22 December 2018 showed that in the case of heavy pollution, the NMB values between the optimal ensemble forecast and observations could also be maintained between -20% and 40%. And the forecast coverage area with good and above pollution by the optimal ensemble was closer to observation than those of single models. During the whole process, the NMB, root mean squared error (RMSE) and R values between forecasted PM2.5 concentrations by the optimal ensemble and observation were from -20% to 20%, from 35 to 75 μg·m-3 and higher than 0.4, respectively, in most polluted areas. Among all single models and ensemble methods, number of girds over China with high total scores was the largest in optimal ensemble. In the eight cities located in the most polluted region, the average onset and end times of the pollution process by optimal ensemble forecast was 1.8 and 6.9 h earlier than observation, respectively. Therefore, we propose that pollutant concentrations retrieval by satellite and surface observation should be fused to improve the refinement of gridding observed pollutant concentrations. And the methods of scale reduction, subjective and objective fusion and rolling correction should be used to further improve the forecast accuracy of gridding multi-model ensemble air quality forecast.
2020, 46(3):393-402.
DOI: 10.7519/j.issn.1000-0526.2020.03.010
Abstract:
Based on historic casualty loss records of rainstorm that occurred at county level in Guangxi from 2009 to 2017, seven factors were selected as explanatory variables by comprehensively considering the trigger factors, disaster formative environment and exposure units, and the prediction model of casualty loss caused by rainstorms was built up by using random forest algorithms. The refined grid precipitation analysis and forecast products were used to drive the model to predict loss of life. The results showed that the classification accuracies are both above 90% in training and testing samples. Disastertriggering factors (precipitation) are the most significant explanatory variables. The importances of these precipitation variables in turn are the anomaly percentage of accumulated precipitation over the previous 10 days, the maximum daily precipitation, the maximum hourly precipitation and the frequency of shorttime severe rainfall. By applying the intelligent grid precipitation products, several rainstorm processes in Guangxi in recent two years were used to verify the model, showing that prediction accuracies are above 70%.
Based on historic casualty loss records of rainstorm that occurred at county level in Guangxi from 2009 to 2017, seven factors were selected as explanatory variables by comprehensively considering the trigger factors, disaster formative environment and exposure units, and the prediction model of casualty loss caused by rainstorms was built up by using random forest algorithms. The refined grid precipitation analysis and forecast products were used to drive the model to predict loss of life. The results showed that the classification accuracies are both above 90% in training and testing samples. Disastertriggering factors (precipitation) are the most significant explanatory variables. The importances of these precipitation variables in turn are the anomaly percentage of accumulated precipitation over the previous 10 days, the maximum daily precipitation, the maximum hourly precipitation and the frequency of shorttime severe rainfall. By applying the intelligent grid precipitation products, several rainstorm processes in Guangxi in recent two years were used to verify the model, showing that prediction accuracies are above 70%.
2020, 46(3):403-411.
DOI: 10.7519/j.issn.1000-0526.2020.03.011
Abstract:
This study conducted experiments based on the hourly visibility forecast products with spatial resolution of 3 km by RMAPS-CHEM. The data in 2016 were taken as a sample, and the model forecasts and observations from each site were compared. The bias was corrected step by step, by considering diffe- rent forecasting errors in different regions, periods and levels. 〖JP2〗The data for 2017 were used for verification. The results show that statistical bias correction has a good correction effect on visibility forecast for 2017, which could not only improve the visibility overestimation in high altitude areas, but also better predict the low visibility phenomenon. Taking January 2017 as an example, the average deviation and root mean square error of Beijing Guanxiangtai Station were reduced, and the accuracy of 0-24 h grading forecasting was improved. Also, the optimized results were reasonably interpolated and applied to the Beijing Integrated Grid Analysis Prediction Systems (iGrAPS) to provide the visibility forecast product in Beijing for 0-96 h with 1 km spatial resolution, which could better support the forecast of low-visibility weather phenomena such as fog and haze.
This study conducted experiments based on the hourly visibility forecast products with spatial resolution of 3 km by RMAPS-CHEM. The data in 2016 were taken as a sample, and the model forecasts and observations from each site were compared. The bias was corrected step by step, by considering diffe- rent forecasting errors in different regions, periods and levels. 〖JP2〗The data for 2017 were used for verification. The results show that statistical bias correction has a good correction effect on visibility forecast for 2017, which could not only improve the visibility overestimation in high altitude areas, but also better predict the low visibility phenomenon. Taking January 2017 as an example, the average deviation and root mean square error of Beijing Guanxiangtai Station were reduced, and the accuracy of 0-24 h grading forecasting was improved. Also, the optimized results were reasonably interpolated and applied to the Beijing Integrated Grid Analysis Prediction Systems (iGrAPS) to provide the visibility forecast product in Beijing for 0-96 h with 1 km spatial resolution, which could better support the forecast of low-visibility weather phenomena such as fog and haze.
2020, 46(3):412-419.
DOI: 10.7519/j.issn.1000-0526.2020.03.012
Abstract:
The moving mean bias correction method and the historical deviation correction method are used to correct the error of the extended-range 2 m temperature forecast of the ECMWF numerical model. Through the study of the moving training period, the optimal training period length of the moving mean bias correction method is 25-30 days for daily temperature forecast from 11 to 15 days. The verification of corrected temperature forecast in 2018 shows that the application of these two temperature deviation correction methods can correct the deviation that the temperature forecast in model is significantly lower than observation, and improve the forecast accuracy by 30% at least. The mean absolute error of the corrected temperature forecast is basically less than 2℃ from June to October, which has certain reference value and could provide good support for forecasters. There is no obvious difference between the two methods in the bias correction effect of the extended-range forecast within 15 days. With the extension of forecast lead time, the advantages of the historical deviation correction method show up gradually.
The moving mean bias correction method and the historical deviation correction method are used to correct the error of the extended-range 2 m temperature forecast of the ECMWF numerical model. Through the study of the moving training period, the optimal training period length of the moving mean bias correction method is 25-30 days for daily temperature forecast from 11 to 15 days. The verification of corrected temperature forecast in 2018 shows that the application of these two temperature deviation correction methods can correct the deviation that the temperature forecast in model is significantly lower than observation, and improve the forecast accuracy by 30% at least. The mean absolute error of the corrected temperature forecast is basically less than 2℃ from June to October, which has certain reference value and could provide good support for forecasters. There is no obvious difference between the two methods in the bias correction effect of the extended-range forecast within 15 days. With the extension of forecast lead time, the advantages of the historical deviation correction method show up gradually.
ZHAO Ruixia
,
DAI Kan
,
JIN Ronghua
,
WEI Qing
,
ZHANG Hong
,
GUO Yunqian
,
LIN Jian
,
WANG Yu
,
TANG Jian
2020, 46(3):420-428.
DOI: 10.7519/j.issn.1000-0526.2020.03.013
Abstract:
The performance of three statistical postprocessing methods and their combinations for 3 h precipitation forecasts out to 72 h from May to September are compared in this paper. They are optimal threat score (OTS) correction, model output statistics (MOS) and MOS with priorspatial observation predictors (OMOS). The 3 h precipitation forecasts of ECMWF model output (DMO), MOS, OMOS, and their OTS correction forecasts (DMOOTS, MOSOTS, OMOSOTS) are evaluated. The results show that MOSOTS method has the best performance in the short term forecast. At the same time, for the heavy precipitation forecast, MOSOTS also obviously outperforms the operational guidance (GD) forecast which integrates subjective and objective predictions. In the first 3 h precipitation forecast, OMOSOTS is the best method. For the first 3 h precipitation forecast, the TS scores of OMOSOTS for thresholds of 0.1, 3 and 10 mm per 3 h are improved about 73%, 198% and 483% than DMO respectively. And the bias score of OMOSOTS is close to 1. In the daily variation during summer time, the first 3 h precipitation forecast from OMOSOTS outperforms both MOSOTS and GD forecast in most days evaluated by TS and Bias scores for the threshhold of 0.1 mm per 3 h.
The performance of three statistical postprocessing methods and their combinations for 3 h precipitation forecasts out to 72 h from May to September are compared in this paper. They are optimal threat score (OTS) correction, model output statistics (MOS) and MOS with priorspatial observation predictors (OMOS). The 3 h precipitation forecasts of ECMWF model output (DMO), MOS, OMOS, and their OTS correction forecasts (DMOOTS, MOSOTS, OMOSOTS) are evaluated. The results show that MOSOTS method has the best performance in the short term forecast. At the same time, for the heavy precipitation forecast, MOSOTS also obviously outperforms the operational guidance (GD) forecast which integrates subjective and objective predictions. In the first 3 h precipitation forecast, OMOSOTS is the best method. For the first 3 h precipitation forecast, the TS scores of OMOSOTS for thresholds of 0.1, 3 and 10 mm per 3 h are improved about 73%, 198% and 483% than DMO respectively. And the bias score of OMOSOTS is close to 1. In the daily variation during summer time, the first 3 h precipitation forecast from OMOSOTS outperforms both MOSOTS and GD forecast in most days evaluated by TS and Bias scores for the threshhold of 0.1 mm per 3 h.
14 Typhoon Vulnerability of Typical Low-Rise Buildings and Application of Intelligent Grid Forecast
2020, 46(3):429-440.
DOI: 10.7519/j.issn.1000-0526.2020.03.014
Abstract:
Lowrise buildings are most severely affected by typhoons. In this paper, based on the reliability design of building structures, MonteCarlo simulation method is used with a probabilistic framework to simulate the wind vulnerability on component of typical doublesloping roofs with small blueandtile bricks (lowrise building). The wind vulnerability of different components of low buildings is as follows: side of roof > eaves > windward roof > roof ridge > wall corner > leeward roof. A probability model of typhoon damage to lowrise houses is constructed which can quantify the risk of wind disasters to lowrise buildings. The ground roughness is classified by using DMSP/OLS data, and the difference of the vulnerability of lowrise buildings with different ground roughness is analyzed. Based on the intelligent grid forecast, the risk of lowrise building induced by Hato, the 13th typhoon in 2017, is estimated. The results are consistent with the actual disaster in the central and western Guangdong. However, partial area of false prediction is generated in the eastern coastal areas of Guangdong, which may be caused by excessive large speed of intelligent grid wind speed forecast products on the east side of Typhoon Hato. The TS score of the estimated results is 0.28, the false alarm 〖JP2〗ratio is 0.62, and prediction omission is 0.48. So, this 〖JP〗model could estimate the risk of wind damage to lowrise buildings to a certain extent.
Lowrise buildings are most severely affected by typhoons. In this paper, based on the reliability design of building structures, MonteCarlo simulation method is used with a probabilistic framework to simulate the wind vulnerability on component of typical doublesloping roofs with small blueandtile bricks (lowrise building). The wind vulnerability of different components of low buildings is as follows: side of roof > eaves > windward roof > roof ridge > wall corner > leeward roof. A probability model of typhoon damage to lowrise houses is constructed which can quantify the risk of wind disasters to lowrise buildings. The ground roughness is classified by using DMSP/OLS data, and the difference of the vulnerability of lowrise buildings with different ground roughness is analyzed. Based on the intelligent grid forecast, the risk of lowrise building induced by Hato, the 13th typhoon in 2017, is estimated. The results are consistent with the actual disaster in the central and western Guangdong. However, partial area of false prediction is generated in the eastern coastal areas of Guangdong, which may be caused by excessive large speed of intelligent grid wind speed forecast products on the east side of Typhoon Hato. The TS score of the estimated results is 0.28, the false alarm 〖JP2〗ratio is 0.62, and prediction omission is 0.48. So, this 〖JP〗model could estimate the risk of wind damage to lowrise buildings to a certain extent.
2020, 46(3):441-448.
DOI: 10.7519/j.issn.1000-0526.2020.03.015
Abstract:
The main characteristics of the general atmospheric circulation in December 2019 are as follows. There were two polar vortex centers in the Northern Hemisphere. The circulation presented a threewave pattern in middlehigh latitudes. The East Asian trough behaved weakly while the south branch trough was stronger. And the subtropical high was located more westward than that in the same period of normal years. Monthly mean precipitation over China was 11.2 mm, which is 6.7% higher than normal (10.5 mm). Monthly mean temperature over China was －2.7℃, which is 0.5℃ higher than normal (－3.2℃). Three relatively severe rainfall events and three cold air processes with moderate strength occurred in December. In addition, two largescale continuous foghaze weather appeared in the central and eastern part of China during 7-10 and 20-26 December.
The main characteristics of the general atmospheric circulation in December 2019 are as follows. There were two polar vortex centers in the Northern Hemisphere. The circulation presented a threewave pattern in middlehigh latitudes. The East Asian trough behaved weakly while the south branch trough was stronger. And the subtropical high was located more westward than that in the same period of normal years. Monthly mean precipitation over China was 11.2 mm, which is 6.7% higher than normal (10.5 mm). Monthly mean temperature over China was －2.7℃, which is 0.5℃ higher than normal (－3.2℃). Three relatively severe rainfall events and three cold air processes with moderate strength occurred in December. In addition, two largescale continuous foghaze weather appeared in the central and eastern part of China during 7-10 and 20-26 December.
Mobile website
® 2024 All Rights Reserved
Supported by:Beijing E-Tiller Technology Development Co., Ltd.
ISSN:1000-0526 CN:11-2282/P