Search Results

You are looking at 1 - 10 of 11 items for

  • Author or Editor: Ching-Sen Chen x
  • Refine by Access: All Content x
Clear All Modify Search
Ching-Sen Chen

Abstract

A two-dimensional numerical cloud model was used to investigate a squall line that occurred over the Taiwan Strait on 16 May 1987 during TAMEX (Taiwan Area Mesoscale Experiment). This squall line illustrated multicellular behavior as revealed by many Doppler analyses. The simulated squall line produced 13 new cells successively in a 6-h simulation, thus maintaining itself in a long-lasting state.

The simulation result recaptured many of the interesting features also observed. The air in the front-to-rear inflow was ingested into the squall-line system primarily from the low levels in front of the system and accelerated upward and backward over a surface pool of cold air in the system-relative frame. Beneath this front-to-rear flow, a rear-to-front current developed between 2 and 4 km in altitude. It descended into the cold pool near the surface. Part of this descending air moved forward toward the gust-front region and part of it moved backward. The relative pressure maximum coincided with the cold pool near the surface and was overlaid by the relative pressure minimum in the midlevels where the convective region corresponded to the decrease of pressure deviation. When the aged cells moved back relative to the system, the decreasing pressure deviation moved back as well.

The vertical pressure-gradient force associated with the dynamic part contributed to the formation of a new cell below the level of free convection, while the sum of the vertical pressure-gradient force associated with buoyancy and the buoyancy itself dominated the development of a cell once this cell had ascended above the level of free convection.

The vertical transport of horizontal momentum normal to the squall line was found to he countergradient in a layer between 2.5 and 6.5 km, which could not be explained in terms of diffusion. In addition, this vertical flux of horizontal momentum normal to the squall line was independent of the mean vertical shear below 8.25 km, except at the 1.75-km height where the vertical shear changed signs very rapidly.

An experiment was made to investigate the role of the ice-phase microphysics on the formation and structure of the simulated squall line. The comparison between our simulation results and other model studies will be presented here.

Full access
Ching-Sen Chen and Yi-Leng Chen

Abstract

The rainfall regimes of Taiwan are investigated using the 18-yr European Centre for Medium-Range Weather Forecasts (ECMWF) data (1980–97), the available 38-yr daily rainfall data from 25 conventional surface stations around Taiwan (1961–98), and the 5-yr hourly rainfall data (1994–98) from 249 high-spatial-resolution Automatic Rainfall and Meteorological Telemetry System (ARMTS) stations.

Rainfall over the island is usually generated either by transient disturbances embedded in the prevailing monsoon flow or local rainshowers related to terrain or local winds. With the change in the direction of the prevailing winds between the warm and cold seasons as well as a variety of transient subsynoptic disturbances occurring in different seasons (e.g., winter monsoon cold surges, springtime cold fronts, mei-yu fronts in the early summer, typhoons in summer months, and cold fronts in fall) and the presence of the Central Mountain Range, the regional rainfall climate over the island shows large spatial and temporal variabilities. Nevertheless, despite the presence of these transient disturbances, it is shown that the horizontal distributions of climatological rainfall patterns for different rainfall regimes are strongly dependent on the direction of the low-level prevailing flow with much higher rainfall on the windward side. Furthermore, the seasonal variations in rainfall amount and type (light precipitation versus convective precipitation) are also dependent on the thermodynamic stratification and the availability of moisture.

Full access
Robert B. Wilhelmson and Ching-Sen Chen

Abstract

A three-dimensional numerical simulation is presented in which five new cells (clouds) develop in succession over a 4 h period. The cells that develop have common characteristics including a lifetime of ∼40 min. They form at 30 min intervals in a convergent region along the cold outflow boundary that is established by previous cells. The propagation of the cloud system is several meters per second faster than individual cell movement due to the discrete development of new cells on the right flank of previous ones.

Analysis of the simulation indicates that the most pronounced changes near the surface are due to downdraft development. In contrast, there is only a weak signature of new updraft development near the surface in the convergence field. Each new updraft develops along the cold outflow boundary which moves slowly away from the previous cell. The air participating in the generation of a new cell appears to have originated 1–1.5 km below cloud base (2.6 km). As the cell grows and rain begins falling out of it, the updraft loses its roots in the subcloud layer and a new cell begins to form.

The sounding used to initialize the model was taken on a day that successive development occurred. Comparison of the observed and modeled cell behavior indicates some broad similarities, but also many differences. For example, new cell development was observed every 15 min. An explanation for this difference and the wide range of frequencies for new cell formation observed in other storms will require further study.

Full access
Ching-Sen Chen and Harold D. Orville

Abstract

The effect of seeding with carbon black dust in the tropical atmosphere is simulated in a two-dimensional time-dependent cloud model which covers a region 6.4 km × 3.2 km in the x and z directions with 200 m grid intervals. Initial conditions are taken from the mean “hurricane season” soundings for the West Indies area (Jordan, 1958). An equation for the continuity of carbon black dust is added to the model. Scavenging of the carbon black dust by precipitation and cloud water content and diffusion of carbon dust to the ocean surface are included in the model. Radiative heating of the atmosphere by the carbon black is simulated. The heating rate is set proportional to the concentration of carbon black dust at a grid point. Air motion is simulated by horizontal density differences caused by horizontal gradients of the heating rate. The circulations lead to cooling and moistening in updrafts and warming and drying in downdrafts in a conditionally stable atmosphere. In the meantime, the carbon black is redistributed.

Three cases are used to test the concept of carbon black dust seeding. The first is a case having a horizontally homogeneous concentration of carbon black, namely, an “Even” case. This case shows that the atmosphere is heated by carbon particles but that no air motion occurs. Changing the initial distribution of carbon black dust to a “rectangular” pattern with gradations of carbon black concentration results in a second case called the “Layer-A” case. In this case, the vertical velocity decreases very rapidly with time after it passes its maximum value of 27 cm s−1 at 10 min. Eighteen percent of the original carbon black dust is lost in 100 min due primarily to diffusion to the ocean surface. Changing the dew-point curve in the mean sounding and using the initial carbon black dust pattern of the layer-A case results in a “Layer-B” case, the third case. A small cloud forms in this case. Comparison of the total amount of carbon black in the Layer-A and Layer-B case shows that small amounts of cloud and rain are not an efficient mechanism to deplete the carbon particles; these water contents do, however, decrease the solar energy absorbed by carbon black particles and hence weaken air motion. The maximum value of the vertical velocity increases by 15 cm s−1 if solar insolation can pass through cloud and rain without any loss.

Results of this numerical study are not encouraging for the direct formation of cloud lines by the spread of carbon black dust in the tropical atmosphere, unless the atmosphere is much more humid than normal. No conclusions concerning mesoscale effects of the solar heating and indirect formation of cloud lines are possible within the framework of this cloud-scale model.

Full access
Ching-Sen Chen, Wen-Sheen Chen, and Zensing Deng

Abstract

The field program TAMEX (Taiwan Area Mesoscale Experiment) was held during May and June 1987. One of its objectives was to study the cited of terrain on precipitation systems. On 7 June 1987 a band of radar echo, orientated from north to south, developed during the afternoon along the western slope and mountainous area of Taiwan island. Before this system moved eastward toward the Pacific Ocean in the late afternoon, it dumped more than 100 mm of precipitation at a few stations in only a few hours. The analysis of radar data from CAA radar revealed that the precipitation occurred over western-sloped terrain and a mountain plateau in the early afternoon. The system was wider than 60 km in the east-west direction, and the echo top was higher than 10 km. The maximum reflectivity was over 50 dBZ along the steep slope and near the mountain peak. The precipitation system over the mountain area extended eastward with the passage of time; meanwhile, new echoes continually formed along the western-sloped area and moved eastward. They intensified as they moved toward the mountain peak merging with the precipitation system. Through this mechanism the precipitation system could maintain itself for several hours and produce a large amount of rainfall.

A two-dimensional numerical cloud model with a terrain-following coordinate system, similar to the one developed by Durran and Klemp, was used to investigate the topographic effect on the precipitation system. A smoother terrain feature was used for the lower boundary, with a 30-km-wide mountain plateau (of less than 1 km in height) and sloped terrain on the western and eastern sides. Surface heating and boundary-layer moisture supply were parameterized in the model. Simulation results indicated that during the early simulation a cell formed near the foothills of the west slope and moved eastward. As it climbed up the sloping terrain it intensified. Its speed decreased and its high intensity was maintained over the slope and the mountain plateau. At the same time, a new cell formed west of the older cell and moved eastward. Finally this new cell merged into the western side of the older one near the mountain peak to form one precipitation system and moved eastward slowly. Thus, the intensity of the merged system was enhanced over the mountain plateau. While this system maintained its high intensity and moved eastward, new cells continually formed along the western slope and moved eastward to merge into the western side of the precipitation system over the mountainous area. The intensity of the precipitation system was enhanced for a few hours over the mountain itself and became a long-lasting system. Toward the end of the simulation, this long-lasting system had moved near the eastern slope and had still maintained its intensity. At the same time, the low-level temperature decreased over the mountainous area as a result of precipitation evaporation. When new cells, forming over the western slope, moved toward the mountain plateau, they entered their decaying stage 45 min after their occurrence. They did not merge into the existing system on the eastern part of the mountain; therefore, the precipitation over the mountain plateau became weaker.

Several sensitivity tests have been made to study the effect of varying the magnitude of surface heating, the boundary-layer moisture supply, the height of the terrain, and the temperature, moisture, and wind profiles on the simulation result. The result indicated that low-level and midlevel moisture were important for the formation of new cells over the western slope and a long-lasting system over the mountain area, respectively. The initial wind speed of 7 m s−1 below 4 km and calm wind above 4 km was used in the model; then a long-lasting precipitation system over the mountainous area appeared. If the wind speed was reduced to 3.5 m s−1, only new cells formed over the western slope. If the maximum height of the terrain was decreased from 1 to 0.5 km, then only new cells formed over the slope area. Hence, sensitivity tests indicated that the combination of the adequate thermodynamic structure, the westerly wind pattern, and the correct size of the mountain could help form both the new cells over the sloped terrain and a long-lasting system over mountain areas as in northern Taiwan on 7 June 1987 during TAMEX. The surface heating effect played the role of creating the upslope wind and augmentation of this precipitation system.

Full access
Shu-Chih Yang, Shu-Hua Chen, Shu-Ya Chen, Ching-Yuang Huang, and Ching-Sen Chen

Abstract

Global positioning system (GPS) radio occultation (RO) data have been broadly used in global and regional numerical weather predictions. Assimilation with the bending angle often performs better than refractivity, which is inverted from the bending angle under spherical assumption and is sometimes associated with negative biases at the lower troposphere; however, the bending angle operator also requires a higher model top as used in global models. This study furnishes the feasibility of bending-angle assimilation in the prediction of heavy precipitation systems with a regional model. The local RO operators for simulating bending angle and refractivity are implemented in the Weather Research and Forecasting (WRF)–local ensemble transform Kalman filter (LETKF) framework. The impacts of assimilating RO data from the Constellation Observing System for Meteorology Ionosphere and Climate (COSMIC) using both operators are evaluated on the prediction of a heavy precipitation episode during Southwest Monsoon Experiment intensive observing period 8 (SoWMEX-IOP8) in 2008. Results show that both the refractivity and bending angle provide a favorable condition for generating this heavy rainfall event. In comparison with the refractivity data, the advantage of assimilating the bending angle is identified in the midtroposphere for deepening of the moist layer that leads to a rainfall forecast closer to the observations.

Full access
Jen-Hsin Teng, Ching-Sen Chen, Tai-Chi Chen Wang, and Yi-Leng Chen

Abstract

A north–south-oriented, multicellular squall line during the Taiwan Area Mesoscale Experiment’s (TAMEX) intensive observation period 2 was studied, using dual-Doppler radar data, as it moved over the island of Taiwan. Over the open ocean, it moved eastward at a constant speed (∼15.5 m s−1).

As the squall line encountered Taiwan Island’s complex mountainous areas, the leading edge moved slower in the mountain ridge areas than over the valleys. As a result, its orientation became approximately parallel to the terrain contours. In the upper levels, the eastward movement of the squall line aloft was less affected by the terrain compared to the low levels. As a result, the westward tilt of the rising motion at the front became less significant. Part of the system-relative rear-to-front flow entered the rising branch of the storm’s circulation as a result of orographic lifting, as illustrated by the trajectory analysis, and was enhanced by latent heat release. The system-relative rear-to-front flow did not appear to descend to the lowest levels. Furthermore, it is unlikely that the low-level cold pool behind the leading edge would move upslope. When the squall line moved over the higher terrain areas, the midlevel system-relative westerly flow from the rear entered the storm’s updraft. The vertical motion pattern was dominated by orographic lifting and sinking. Low-level system-relative front-to-rear inflow was absent. Weak echo maxima were found, mainly associated with orographic lifting ahead of or near the mountain peaks. These changes occurred within a short time (∼40 min) after the squall line encountered the mountainous terrain.

The environmental low-level wind ahead of the squall line exhibited a weak southerly flow west of the mountain ridge areas and a weak southwesterly flow in the valley areas because of island blocking. As a result, the low-level system-relative front-to-rear inflow was weaker in the valley areas than that in ridge areas. Nevertheless, the low-level moisture inflow from the front decreased rapidly in both the former and the latter areas when the squall line encountered mountainous terrain. This was attributed to a rapid decrease in the depth of the low-level inflow layer and less moisture availability in the higher terrain areas. The drier midlevel air entered the storm from the west with the echo tops decreasing as the squall line moved farther over the mountainous terrain.

Full access
Ching-Sen Chen, Yi-Leng Chen, Che-Ling Liu, Pay-Liam Lin, and Wan-Chin Chen

Abstract

The seasonal variations of heavy rainfall days over Taiwan are analyzed using 6-yr (1997–2002) hourly rainfall data from about 360 rainfall stations, including high-spatial-resolution Automatic Rainfall and Meteorological Telemetry System stations and 25 conventional stations. The seasonal variations and spatial variations of nontyphoon and typhoon heavy rainfall occurrences (i.e., the number of rainfall stations with rainfall rate >15 mm h−1 and daily accumulation >50 mm) are also analyzed. From mid-May to early October, with abundant moisture, potential instability, and the presence of mountainous terrain, nontyphoon heavy rainfall days are frequent (>60%), but only a few stations recorded extremely heavy rainfall (>130 mm day−1) during the passage of synoptic disturbances or the drifting of mesoscale convective systems inland. During the mei-yu season, especially in early June, these events are more widespread than in other seasons. The orographic effects are important in determining the spatial distribution of heavy rainfall occurrences with a pronounced afternoon maximum, especially during the summer months under the southwesterly monsoon flow. After the summer–autumn transition, heavy rainfall days are most frequent over northeastern Taiwan under the northeasterly monsoon flow. Extremely heavy rainfall events (>130 mm day−1) are infrequent during the winter months because of stable atmospheric stratification with a low moisture content. Typhoon heavy rainfall events start in early May and become more frequent in late summer and early autumn. During the analysis period, heavy rainfall occurrences are widespread and dominated by extremely heavy rainfall events (>130 mm day−1) on the windward slopes of the storm circulations. The spatial distribution of typhoon heavy rainfall occurrences depends on the typhoon track with very little diurnal variation.

Full access
Chuan-Chi Tu, Yi-Leng Chen, Ching-Sen Chen, Pay-Liam Lin, and Po-Hsiung Lin

Abstract

Two contrasting localized heavy rainfall events during Taiwan’s early summer rainy season with the daily rainfall maximum along the windward mountain range and coast were studied and compared using a combination of observations and numerical simulations. Both events occurred under favorable large-scale settings including the existence of a moisture tongue from the tropics. For the 31 May case, heavy rainfall occurred in the afternoon hours over the southwestern windward slopes after a shallow surface front passed central Taiwan. The orographic lifting of the prevailing warm, moist, west-southwesterly flow aloft, combined with a sea breeze–upslope flow at the surface provided the localized lifting needed for the development of heavy precipitation. On 16 June before sunrise, pronounced orographic blocking of the warm, moist, south-southwesterly flow occurred because of the presence of relatively cold air at low levels as a result of nocturnal and rain evaporative cooling. As a result, convective systems intensified as they moved toward the southwestern coast. During the daytime, the cold pool remained over southwestern Taiwan without the development of onshore/upslope flow. Furthermore, with a south-southwesterly flow aloft parallel to terrain contours, orographic lifting aloft was absent and preexisting rain cells offshore diminished after they moved inland. Over northern Taiwan on the lee side, a sea breeze/onshore flow developed in the afternoon hours, resulting in heavy thundershowers. These results demonstrate the importance of diurnal and local effects on determining the location and timing of the occurrences of localized heavy precipitation during the early summer rainy season over Taiwan.

Full access
Han-Ching Chen, Fei-Fei Jin, Sen Zhao, Andrew T. Wittenberg, and Shaocheng Xie

Abstract

This study examines historical simulations of ENSO in the E3SM-1-0, CESM2, and GFDL-CM4 climate models, provided by three leading U.S. modeling centers as part of the Coupled Model Intercomparison Project phase 6 (CMIP6). These new models have made substantial progress in simulating ENSO’s key features, including: amplitude; timescale; spatial patterns; phase-locking; spring persistence barrier; and recharge oscillator dynamics. However, some important features of ENSO are still a challenge to simulate. In the central and eastern equatorial Pacific, the models’ weaker-than-observed subsurface zonal current anomalies and zonal temperature gradient anomalies serve to weaken the nonlinear zonal advection of subsurface temperatures, leading to insufficient warm/cold asymmetry of ENSO’s sea surface temperature anomalies (SSTA). In the western equatorial Pacific, the models’ excessive simulated zonal SST gradients amplify their zonal temperature advection, causing their SSTA to extend farther west than observed. The models underestimate both ENSO’s positive dynamic feedbacks (due to insufficient zonal wind stress responses to SSTA) and its thermodynamic damping (due to insufficient convective cloud shading of eastern Pacific SSTA during warm events); compensation between these biases leads to realistic linear growth rates for ENSO, but for somewhat unrealistic reasons. The models also exhibit stronger-than-observed feedbacks onto eastern equatorial Pacific SSTAs from thermocline depth anomalies, which accelerates the transitions between events and shortens the simulated ENSO period relative to observations. Implications for diagnosing and simulating ENSO in climate models are discussed.

Restricted access