2013 Vol. 19, No. 1
2013, 19(1): 1-15.
Abstract:
Due to the higher temporal and spatial resolution and the better integrality of long-term satellite infrared (IR) Brightness Temperature (TBB) data, a climatology of deep convection during summer over South China and the adjacent waters is presented in this paper based on the 1-hourly infrared IR TBB data during June-August of 1996-2007 (except 2004). The results show that the geographic distribution of deep convection denoted by TBB ≤ -52°C over South China and the adjacent waters are basically consistent with previous statistical results based on surface thunderstorm observations and low-orbit satellite lightning observations. The monthly, ten-day, five-day and diurnal variations of deep convection in this region are focused on in this paper. There are 5 active deep-convection areas in June-August. The monthly variations of the deep convection are closely associated with the large-scale atmospheric circulations. The deep convection over the land areas of South China is more active in June while that over the South China Sea is more active in July and August. The development of deep convection is prominently intermittent and its period is about 3 to 5 five-day periods. However, the deep convection over the coastal areas in South China remains more active during summer and has no apparent intermittence. The ten-day and five-day variations of deep convection show that there are different variations of deep convection over different areas in South China and the adjacent waters. The tendency of deep convection over the land areas of South China is negatively correlated with that over the South China Sea. The diurnal variations of deep convection show that the sea-land breeze, caused by the thermal differences between land and sea, and the mountain-valley breeze, caused by the thermal differences between mountains and plains or basins, cause deep convection to propagate from sea to land in the afternoon and from land to sea after midnight, and the convection over mountains propagates from mountains to plains after midnight. The different diurnal variations of deep convection over different underlying surfaces show that not only there are general mountainous, marine and multi-peak deep convection, but also there is longer-duration deep convection over coastal areas and other deep convection triggered and maintained by larger-scale weather systems in South China during summer.
Due to the higher temporal and spatial resolution and the better integrality of long-term satellite infrared (IR) Brightness Temperature (TBB) data, a climatology of deep convection during summer over South China and the adjacent waters is presented in this paper based on the 1-hourly infrared IR TBB data during June-August of 1996-2007 (except 2004). The results show that the geographic distribution of deep convection denoted by TBB ≤ -52°C over South China and the adjacent waters are basically consistent with previous statistical results based on surface thunderstorm observations and low-orbit satellite lightning observations. The monthly, ten-day, five-day and diurnal variations of deep convection in this region are focused on in this paper. There are 5 active deep-convection areas in June-August. The monthly variations of the deep convection are closely associated with the large-scale atmospheric circulations. The deep convection over the land areas of South China is more active in June while that over the South China Sea is more active in July and August. The development of deep convection is prominently intermittent and its period is about 3 to 5 five-day periods. However, the deep convection over the coastal areas in South China remains more active during summer and has no apparent intermittence. The ten-day and five-day variations of deep convection show that there are different variations of deep convection over different areas in South China and the adjacent waters. The tendency of deep convection over the land areas of South China is negatively correlated with that over the South China Sea. The diurnal variations of deep convection show that the sea-land breeze, caused by the thermal differences between land and sea, and the mountain-valley breeze, caused by the thermal differences between mountains and plains or basins, cause deep convection to propagate from sea to land in the afternoon and from land to sea after midnight, and the convection over mountains propagates from mountains to plains after midnight. The different diurnal variations of deep convection over different underlying surfaces show that not only there are general mountainous, marine and multi-peak deep convection, but also there is longer-duration deep convection over coastal areas and other deep convection triggered and maintained by larger-scale weather systems in South China during summer.
2013, 19(1): 16-27.
Abstract:
By using NCEP GODAS monthly sea surface height (SSH) and 160-station monthly precipitation data in China, the seasonal and interannual characteristics of SSH are analyzed over the tropical Pacific, and correlations between SSH and summer rainfall are discussed. The results are shown as follows: (1) The tropical Pacific SSH takes on a “V” pattern in the climatic field with an eastward opening, and it is higher in the western part (in the northwestern part) than in the eastern part (in the southwestern part). The high-value areas are more stable in the northwest, and the value range (greater than 0.8 m) is larger in spring and summer than in autumn and winter. The high-value area in the southwestern part is the largest (smallest) and more northerly (southerly) in spring (summer). SSH is higher in spring and autumn than in summer and winter over the equatorial zone. (2) The interannual anomalies of the SSH are the strongest over the tropical western and southwestern Pacific and are stronger in winter and spring than in summer and autumn. The interannual anomalies are also strong over the equatorial middle and eastern Pacific. The distribution ranges are larger and the intensities are stronger in the autumn and winter. There is a close relationship between the SSH interannual anomalies and ENSO events in autumn, winter and spring. (3) When ENSO events take place in winter, according to the simultaneous relationship among the tropic Pacific SSH, 850 hPa wind fields and the summer precipitation of China, it can be predicted that the precipitation will be significantly more than normal over the south of the Yangtze River, especially over Dongting Lake and Poyang Lake region, eastern Qinghai-Tibet Plateau, Yangtze-Huai River Valley, eastern part of Inner Mongolia and less than normal over the area of Great Band of Yellow River, North China and South China in successive summers.
By using NCEP GODAS monthly sea surface height (SSH) and 160-station monthly precipitation data in China, the seasonal and interannual characteristics of SSH are analyzed over the tropical Pacific, and correlations between SSH and summer rainfall are discussed. The results are shown as follows: (1) The tropical Pacific SSH takes on a “V” pattern in the climatic field with an eastward opening, and it is higher in the western part (in the northwestern part) than in the eastern part (in the southwestern part). The high-value areas are more stable in the northwest, and the value range (greater than 0.8 m) is larger in spring and summer than in autumn and winter. The high-value area in the southwestern part is the largest (smallest) and more northerly (southerly) in spring (summer). SSH is higher in spring and autumn than in summer and winter over the equatorial zone. (2) The interannual anomalies of the SSH are the strongest over the tropical western and southwestern Pacific and are stronger in winter and spring than in summer and autumn. The interannual anomalies are also strong over the equatorial middle and eastern Pacific. The distribution ranges are larger and the intensities are stronger in the autumn and winter. There is a close relationship between the SSH interannual anomalies and ENSO events in autumn, winter and spring. (3) When ENSO events take place in winter, according to the simultaneous relationship among the tropic Pacific SSH, 850 hPa wind fields and the summer precipitation of China, it can be predicted that the precipitation will be significantly more than normal over the south of the Yangtze River, especially over Dongting Lake and Poyang Lake region, eastern Qinghai-Tibet Plateau, Yangtze-Huai River Valley, eastern part of Inner Mongolia and less than normal over the area of Great Band of Yellow River, North China and South China in successive summers.
2013, 19(1): 28-38.
Abstract:
Progresses in the research for season division and seasonal changes are reviewed systematically in this paper, which particularly introduces the results of a national natural science foundation project, “the nonlinear identification of seasonal changes and its responses to global warming”. The project developed two objective and quantitative methods for season division: the non-linear similarity measure (NSM) method and the regional multi-element optimal dissection (RMOD) method, which differ from traditional ones for season division. Besides, the project further investigated seasonal changes and their responses to global warming, analyzed evolution characteristics of the starting date and length of seasons in history, revealed their relationships with extreme events and precipitation patterns in rainy seasons, and thereby built a research system for seasonal changes in China under the global warming background. Finally, the authors make an outlook on the research for season division and seasonal changes and put forward several issues associated with season and climate changes that need to be further explored.
Progresses in the research for season division and seasonal changes are reviewed systematically in this paper, which particularly introduces the results of a national natural science foundation project, “the nonlinear identification of seasonal changes and its responses to global warming”. The project developed two objective and quantitative methods for season division: the non-linear similarity measure (NSM) method and the regional multi-element optimal dissection (RMOD) method, which differ from traditional ones for season division. Besides, the project further investigated seasonal changes and their responses to global warming, analyzed evolution characteristics of the starting date and length of seasons in history, revealed their relationships with extreme events and precipitation patterns in rainy seasons, and thereby built a research system for seasonal changes in China under the global warming background. Finally, the authors make an outlook on the research for season division and seasonal changes and put forward several issues associated with season and climate changes that need to be further explored.
2013, 19(1): 39-48.
Abstract:
Interannual variability of landfalling tropical cyclones (TCs) in China during 1960-2010 is investigated. By using the method of partial least squares regression (PLS-regression), canonical ENSO and ENSO Modoki are identified to be the factors that contribute to the interannual variability of landfalling TCs. El Niño Modoki years are associated with a greater-than-average frequency of landfalling TCs in China, but reversed in canonical El Niño years. Significant difference in genesis locations of landfalling TCs in China for the two kinds of El Niño phases occurs dominantly in the northern tropical western North Pacific (WNP). The patterns of low-level circulation anomalies and outgoing longwave radiation (OLR) anomalies associated with landfalling TC genesis with different types of El Niño phases are examined. During canonical El Niño years, a broad zonal band of positive OLR anomalies dominates the tropical WNP, while the circulation anomalies exhibit a meridionally symmetrical dipole pattern with an anticyclonic anomaly in the subtropics and a cyclonic anomaly near the tropics. In El Niño Modoki years, a vast region of negative OLR anomalies, roughly to the south of 25°N with a strong large-scale cyclonic anomaly over the tropical WNP, provides a more favorable condition for landfalling TC genesis compared to its counterpart during canonical El Niño years. For more landfalling TCs formed in the northern tropical WNP in El Niño Modoki years, there are more TCs making landfall on the northern coast of China in El Niño Modoki years than in canonical El Niño years. The number of landfalling TCs is slightly above normal in canonical La Niña years. Enhanced convection is found in the South China Sea (SCS) and the west of the tropical WNP, which results in landfalling TCs forming more westward in canonical La Niña years. During La Niña Modoki years, the landfalling TC frequency are below normal, owing to an unfavorable condition for TC genesis persisting in a broad zonal band from 5°N to 25°N. Since the western North Pacific subtropical high (WNPSH) in La Niña Modoki years is located in the westernmost region, TCs mainly make landfall on the south coast of China.
Interannual variability of landfalling tropical cyclones (TCs) in China during 1960-2010 is investigated. By using the method of partial least squares regression (PLS-regression), canonical ENSO and ENSO Modoki are identified to be the factors that contribute to the interannual variability of landfalling TCs. El Niño Modoki years are associated with a greater-than-average frequency of landfalling TCs in China, but reversed in canonical El Niño years. Significant difference in genesis locations of landfalling TCs in China for the two kinds of El Niño phases occurs dominantly in the northern tropical western North Pacific (WNP). The patterns of low-level circulation anomalies and outgoing longwave radiation (OLR) anomalies associated with landfalling TC genesis with different types of El Niño phases are examined. During canonical El Niño years, a broad zonal band of positive OLR anomalies dominates the tropical WNP, while the circulation anomalies exhibit a meridionally symmetrical dipole pattern with an anticyclonic anomaly in the subtropics and a cyclonic anomaly near the tropics. In El Niño Modoki years, a vast region of negative OLR anomalies, roughly to the south of 25°N with a strong large-scale cyclonic anomaly over the tropical WNP, provides a more favorable condition for landfalling TC genesis compared to its counterpart during canonical El Niño years. For more landfalling TCs formed in the northern tropical WNP in El Niño Modoki years, there are more TCs making landfall on the northern coast of China in El Niño Modoki years than in canonical El Niño years. The number of landfalling TCs is slightly above normal in canonical La Niña years. Enhanced convection is found in the South China Sea (SCS) and the west of the tropical WNP, which results in landfalling TCs forming more westward in canonical La Niña years. During La Niña Modoki years, the landfalling TC frequency are below normal, owing to an unfavorable condition for TC genesis persisting in a broad zonal band from 5°N to 25°N. Since the western North Pacific subtropical high (WNPSH) in La Niña Modoki years is located in the westernmost region, TCs mainly make landfall on the south coast of China.
2013, 19(1): 49-58.
Abstract:
This work investigates the distribution of high winds above Beaufort scale 6 in the offshore zones of China using high-resolution satellite measurements. A numerical experiment is carried out in order to find out the effects of Taiwan Island on the formation of strong winds. The analysis indicates that the distribution of high wind occurrence is similar to that of the average wind velocity in winter. High winds tend to be anchored in special topographical regions, such as the Taiwan Strait, the Bashi Channel and the southeast coast of Vietnam. High winds occur much more frequently over the warmer than the colder flank of Kuroshio front as it meanders from Taiwan to Japan. The frequency of high winds decreases drastically in spring. The Taiwan Strait maintains the largest high wind occurrence. Besides, high winds remain frequent in the Bashi Channel, the southeast tip of Taiwan Island and the warmer flank of Kuroshio front. In summer, high winds generally occur infrequently except over a broad region off the southeast coast of Vietnam near 10°N and the frequency there decreases from southwest to northeast. High winds around Taiwan Island present near axisymmetric distribution with larger frequency along southeast-northwest direction and smaller frequency along southwest-northeast direction. The dominant direction of high winds exhibits a counterclockwise circulation surrounding the island. The frequency of high winds increases rapidly in autumn and almost repeats the distribution that appears in winter. The simulation results suggest that the effects of Taiwan Island topography on high winds vary with seasons. In winter, topography is the major cause of high winds in the surrounding oceanic zones. High winds in both Taiwan Strait and the southeast corner of the island disappear and the frequency decreases gradually from south to north when the terrain is removed. However, in summer, high wind frequency derived from two simulations with and without terrain is almost identical. We attribute this phenomenon to the factors which are responsible for the formation of high winds.
This work investigates the distribution of high winds above Beaufort scale 6 in the offshore zones of China using high-resolution satellite measurements. A numerical experiment is carried out in order to find out the effects of Taiwan Island on the formation of strong winds. The analysis indicates that the distribution of high wind occurrence is similar to that of the average wind velocity in winter. High winds tend to be anchored in special topographical regions, such as the Taiwan Strait, the Bashi Channel and the southeast coast of Vietnam. High winds occur much more frequently over the warmer than the colder flank of Kuroshio front as it meanders from Taiwan to Japan. The frequency of high winds decreases drastically in spring. The Taiwan Strait maintains the largest high wind occurrence. Besides, high winds remain frequent in the Bashi Channel, the southeast tip of Taiwan Island and the warmer flank of Kuroshio front. In summer, high winds generally occur infrequently except over a broad region off the southeast coast of Vietnam near 10°N and the frequency there decreases from southwest to northeast. High winds around Taiwan Island present near axisymmetric distribution with larger frequency along southeast-northwest direction and smaller frequency along southwest-northeast direction. The dominant direction of high winds exhibits a counterclockwise circulation surrounding the island. The frequency of high winds increases rapidly in autumn and almost repeats the distribution that appears in winter. The simulation results suggest that the effects of Taiwan Island topography on high winds vary with seasons. In winter, topography is the major cause of high winds in the surrounding oceanic zones. High winds in both Taiwan Strait and the southeast corner of the island disappear and the frequency decreases gradually from south to north when the terrain is removed. However, in summer, high wind frequency derived from two simulations with and without terrain is almost identical. We attribute this phenomenon to the factors which are responsible for the formation of high winds.
2013, 19(1): 59-66.
Abstract:
Addressing the difficulties of scattered and sparse observational data in ocean science, a new interpolation technique based on information diffusion is proposed in this paper. Based on a fuzzy mapping idea, sparse data samples are diffused and mapped into corresponding fuzzy sets in the form of probability in an interpolation ellipse model. To avoid the shortcoming of normal diffusion function on the asymmetric structure, a kind of asymmetric information diffusion function is developed and a corresponding algorithm-ellipse model for diffusion of asymmetric information is established. Through interpolation experiments and contrast analysis of the sea surface temperature data with ARGO data, the rationality and validity of the ellipse model are assessed.
Addressing the difficulties of scattered and sparse observational data in ocean science, a new interpolation technique based on information diffusion is proposed in this paper. Based on a fuzzy mapping idea, sparse data samples are diffused and mapped into corresponding fuzzy sets in the form of probability in an interpolation ellipse model. To avoid the shortcoming of normal diffusion function on the asymmetric structure, a kind of asymmetric information diffusion function is developed and a corresponding algorithm-ellipse model for diffusion of asymmetric information is established. Through interpolation experiments and contrast analysis of the sea surface temperature data with ARGO data, the rationality and validity of the ellipse model are assessed.
2013, 19(1): 67-79.
Abstract:
A quantitative diagnosis is carried out for the upward branch of a local meridional circulation over southern China (SC) during the abnormal snowstorms with severe freezing rain from 10 January to 3 February 2008. The diagnostic study shows that the upward branch is mainly associated with the zonal advection of westerly momentum and meridional temperature advection instead of the latent heating (which is commonly the dominant factor in many other storm cases). The corresponding weather analyses indicate that (1) the zonal advection of westerly momentum represents the effect of the upper-level divergence on the anticyclone-shear side in the entrance of a 200 hPa westerly jet with a westward deviation from its climatological location over southwestern Japan; (2) the meridional temperature advection represents the interaction between the mid-lower layer (850 to 400 hPa) warm advection over SC (ahead of temperature and pressure troughs with the latter trough deeper than the former in the Bay of Bengal) and cold advection over north China (steered by an underlying flow at 500 hPa); (3) the relatively weak vapor transport (compared to that of spring, summer and autumn) from the Bay of Bengal and the South China Sea to SC and the existence of a temperature inversion layer in the lower troposphere over SC diminish the effect of latent heating. With the significant increase of vapor transport after 24 January, the role of latent heating is upgraded to become the third positive contributor to the upward branch over SC.
A quantitative diagnosis is carried out for the upward branch of a local meridional circulation over southern China (SC) during the abnormal snowstorms with severe freezing rain from 10 January to 3 February 2008. The diagnostic study shows that the upward branch is mainly associated with the zonal advection of westerly momentum and meridional temperature advection instead of the latent heating (which is commonly the dominant factor in many other storm cases). The corresponding weather analyses indicate that (1) the zonal advection of westerly momentum represents the effect of the upper-level divergence on the anticyclone-shear side in the entrance of a 200 hPa westerly jet with a westward deviation from its climatological location over southwestern Japan; (2) the meridional temperature advection represents the interaction between the mid-lower layer (850 to 400 hPa) warm advection over SC (ahead of temperature and pressure troughs with the latter trough deeper than the former in the Bay of Bengal) and cold advection over north China (steered by an underlying flow at 500 hPa); (3) the relatively weak vapor transport (compared to that of spring, summer and autumn) from the Bay of Bengal and the South China Sea to SC and the existence of a temperature inversion layer in the lower troposphere over SC diminish the effect of latent heating. With the significant increase of vapor transport after 24 January, the role of latent heating is upgraded to become the third positive contributor to the upward branch over SC.
2013, 19(1): 80-86.
Abstract:
Using the regional air-sea coupled climate model RegCM3-POM, a series of numerical experiments are performed to simulate the summer climate in 1997 and 1998 with different coupling time steps. The results show that the coupled model has good performance on the simulation of the summer sea surface temperature (SST) in 1997 and 1998, and the simulation results of CPL1 (with the coupling time step at 1 hour) are similar to those of CPL6 (with the coupling time step at 6 hours). The coupled model can well simulate SST differences between 1997 and 1998. As for the simulation of the drought in 1997 and the flood in 1998, the results of CPL6 are more accurate. The coupled model can well simulate the drought in 1997 over North China, and compared with the results of the atmosphere model RegCM3, the simulation ability of the coupled model is improved. The coupling model has better ability in the simulation of the circulation in the middle and low levels, and the water vapor transportation in the coupling model is reasonable in both 1997 and 1998. RegCM3 (an uncoupled model) cannot correctly simulate the transportation path differences between 1997 and 1998, but the coupled model can simulate the differences well.
Using the regional air-sea coupled climate model RegCM3-POM, a series of numerical experiments are performed to simulate the summer climate in 1997 and 1998 with different coupling time steps. The results show that the coupled model has good performance on the simulation of the summer sea surface temperature (SST) in 1997 and 1998, and the simulation results of CPL1 (with the coupling time step at 1 hour) are similar to those of CPL6 (with the coupling time step at 6 hours). The coupled model can well simulate SST differences between 1997 and 1998. As for the simulation of the drought in 1997 and the flood in 1998, the results of CPL6 are more accurate. The coupled model can well simulate the drought in 1997 over North China, and compared with the results of the atmosphere model RegCM3, the simulation ability of the coupled model is improved. The coupling model has better ability in the simulation of the circulation in the middle and low levels, and the water vapor transportation in the coupling model is reasonable in both 1997 and 1998. RegCM3 (an uncoupled model) cannot correctly simulate the transportation path differences between 1997 and 1998, but the coupled model can simulate the differences well.
2013, 19(1): 87-96.
Abstract:
In 2005, significant rainfall reinforcement and severe disaster was induced by tropical cyclone (TC) Talim after it made landfall on the east of China. Observational analyses show that it has relationship with cold air intrusion. For investigating the impact of cold air intensity, we make use of Weather Research and Forecasting (WRF) model, the synthesizer of NCEP/NCAR reanalysis data and Japan regional spectral model data, to carry out numerical experiments. Results show that rainfall reinforcement occurs in all experiments. Different intensity of cold air can modify the rainfall distribution and intensity significantly. In the rainfall center, the increment maximum of rainfall is twice as large as that of the minimum. Moderate cold air intrusion may result in the strongest rainfall reinforcement. Different cold air intensity can lead to different motion of low-level convergence lines and fronts. There is a good relationship between the rainfall region and the eastern part of the front. On one hand, strong cold air weakens the TC intensity by its intrusion into the TC center and results in weak convergence and a convergent zone and a rain band shifted southward. On the other hand, weak cold air reduces the convergence and moves the convergent zone and rain band northward. Moderate cold air intrusion maintains strong low-level convergence and high-level divergence, keeping strong upward motion over certain regions. Consequently, the rain band begins to stagnate and rainfall reinforces abruptly therein.
In 2005, significant rainfall reinforcement and severe disaster was induced by tropical cyclone (TC) Talim after it made landfall on the east of China. Observational analyses show that it has relationship with cold air intrusion. For investigating the impact of cold air intensity, we make use of Weather Research and Forecasting (WRF) model, the synthesizer of NCEP/NCAR reanalysis data and Japan regional spectral model data, to carry out numerical experiments. Results show that rainfall reinforcement occurs in all experiments. Different intensity of cold air can modify the rainfall distribution and intensity significantly. In the rainfall center, the increment maximum of rainfall is twice as large as that of the minimum. Moderate cold air intrusion may result in the strongest rainfall reinforcement. Different cold air intensity can lead to different motion of low-level convergence lines and fronts. There is a good relationship between the rainfall region and the eastern part of the front. On one hand, strong cold air weakens the TC intensity by its intrusion into the TC center and results in weak convergence and a convergent zone and a rain band shifted southward. On the other hand, weak cold air reduces the convergence and moves the convergent zone and rain band northward. Moderate cold air intrusion maintains strong low-level convergence and high-level divergence, keeping strong upward motion over certain regions. Consequently, the rain band begins to stagnate and rainfall reinforces abruptly therein.
2013, 19(1): 97-103.
Abstract:
Using daily rainfall data of 11 observatory stations over Shanghai for the period 1960-2007, the spatial differences of rainfall over the Shanghai region during periods with slow and rapid urbanization respectively are investigated based on spatial standard deviation of rainfall and its relative variables. Results show that spatial differences increase with the acceleration of urbanization. Spatial distributions of annual rainfall and rainstorm frequency exhibit distinct urban ‘rain-island’ features during the rapid period of urbanization (1960-1983) while it is opposite in the case of slow urbanization (1984-2007). Changes in the spatial distribution of annual rainfall trends also take place during different periods. Specifically, the variation of annual rainfall exhibits consistent trends over the Shanghai region in the slow urbanization periods. However, inconsistent spatial distribution of variations has taken place over the central districts and suburbs of Shanghai during the rapid urbanization stage. Since the speeding-up of urbanization, the annual rainfall amount over central districts of Shanghai tends to increase while that in the suburbs shows a decreasing trend. In addition, as far as different seasons are concerned, the speed of urbanization exerts insignificant influences on the spatial distribution of rainfall during winter and spring. On the contrary, the rainfall during summer and autumn (especially summer) is featured with an island effect during the rapid urbanization period.
Using daily rainfall data of 11 observatory stations over Shanghai for the period 1960-2007, the spatial differences of rainfall over the Shanghai region during periods with slow and rapid urbanization respectively are investigated based on spatial standard deviation of rainfall and its relative variables. Results show that spatial differences increase with the acceleration of urbanization. Spatial distributions of annual rainfall and rainstorm frequency exhibit distinct urban ‘rain-island’ features during the rapid period of urbanization (1960-1983) while it is opposite in the case of slow urbanization (1984-2007). Changes in the spatial distribution of annual rainfall trends also take place during different periods. Specifically, the variation of annual rainfall exhibits consistent trends over the Shanghai region in the slow urbanization periods. However, inconsistent spatial distribution of variations has taken place over the central districts and suburbs of Shanghai during the rapid urbanization stage. Since the speeding-up of urbanization, the annual rainfall amount over central districts of Shanghai tends to increase while that in the suburbs shows a decreasing trend. In addition, as far as different seasons are concerned, the speed of urbanization exerts insignificant influences on the spatial distribution of rainfall during winter and spring. On the contrary, the rainfall during summer and autumn (especially summer) is featured with an island effect during the rapid urbanization period.
2013, 19(1): 104-108.
Abstract:
The LS-SVM (Least squares support vector machine) method is presented to set up a model to forecast the occurrence of thunderstorms in the Nanjing area by combining NCEP FNL Operational Global Analysis data on 1.0°×1.0° grids and cloud-to-ground lightning data observed with a lightning location system in Jiangsu province during 2007-2008. A dataset with 642 samples, including 195 thunderstorm samples and 447 non-thunderstorm samples, are randomly divided into two groups, one (having 386 samples) for modeling and the rest for independent verification. The predictors are atmospheric instability parameters which can be obtained from the NCEP data and the predictand is the occurrence of thunderstorms observed by the lightning location system. Preliminary applications to the independent samples for a 6-hour forecast of thunderstorm events show that the prediction correction rate of this model is 78.26%, false alarm rate is 21.74%, and forecasting technical score is 0.61, all better than those from either linear regression or artificial neural network.
The LS-SVM (Least squares support vector machine) method is presented to set up a model to forecast the occurrence of thunderstorms in the Nanjing area by combining NCEP FNL Operational Global Analysis data on 1.0°×1.0° grids and cloud-to-ground lightning data observed with a lightning location system in Jiangsu province during 2007-2008. A dataset with 642 samples, including 195 thunderstorm samples and 447 non-thunderstorm samples, are randomly divided into two groups, one (having 386 samples) for modeling and the rest for independent verification. The predictors are atmospheric instability parameters which can be obtained from the NCEP data and the predictand is the occurrence of thunderstorms observed by the lightning location system. Preliminary applications to the independent samples for a 6-hour forecast of thunderstorm events show that the prediction correction rate of this model is 78.26%, false alarm rate is 21.74%, and forecasting technical score is 0.61, all better than those from either linear regression or artificial neural network.
粤公网安备 4401069904700002号