[Postscript]

Numerical Studies of Storms with Heavy Rainfall/Snowfall in Meteorological Research Institute, Japan Meteorological Agency


Masanori Yoshizaki
Forecast Research Department
Meteorological Research Institute, Japan Meteorological Agency
1-1, Nagamine, Tsukuba, Ibaraki, Japan 305



1. Introduction

In Japan, studies on the formation and development mechanisms of heavy rainfall/snowfall and their forecast are important and urgent subjects in mesoscale meteorology. Heavy rainfall/snowfall are brought by meso-alpha or meso-beta scale disturbances, such as cloud clusters along the Baiu front in early summer, rain systems associated with typhoons and cyclones, snow storms in the winter season, and so on.

These disturbances are embedded in large-scale disturbances and the hierarchical structures are commonly observed. The large-scale disturbances are well forecasted by operational regional prediction models. However, forecast of disturbances smaller than meso-alpha scale, especially with precipitation, are still difficult.

Our group (Forecast Research Department, Meteorological Research Institute (MRI), Japan Meteorological Agency (JMA)) studies mesoscale diturbances and weather phenomena theoretically, analytically, numerically and statistically. Here I would like to introduce our recent works related to numerical studies of storms with heavy rainfall/snowfall. Our main activities are development of numerical techniques and models (hydrostatic and non-hydrostatic versions), and validation/comparison of numerical simulations of observed/analysed mesoscale events.

2. Researches by hydrostatic regional models

The Japan Spectral Model (JSM) is a limited-area hydrostatic model with a sigma-vertical coordinate, which had been in operation at JMA until February 1996. It is now being used as a research model mainly to investigate the mechanisms of mesoscale disturbances by reducing the horizontal resolution to 10-20 km, which was 30 km in operational model.

Extratropical cyclones, which pass along the southern coast of the Japan Islands in the winter season, sometimes bring strong wind and heavy snowfall around Tokyo. Takano(1996) applied the JSM to the rapidly developing cyclone on 12 February 1994. The model simulated development of the cyclone successfully. He showed that snowfall was intensified by the frontogenesis on the eastern and northern sides of the cyclone.

The correction of positional errors of precipitation between the JSM forecast fields and the observations is an important problem for assimilation. Aonashi and Shibata (1996) developed an assimilation method, which modifies the relative humidity (RH) and divergence of background data provided by the JSM, and evaluated the impact of assimilating precipitable water content (PWC) and rain flag data derived from SSM/I data. This method reduces the positional errors of the precipitation forecasts by the JSM. The assimilation of PWC is found to be effective in eliminating model precipitation in the observed rain-free areas, while that of rain flag data is found to be effective in generating model precipitation in the observed precipitation areas.

3. Researches by non-hydrostatic cloud models (NHMs)

A three-dimensional anelastic non-hydrostatic model has been developed by Ikawa and Saito (1991), Saito (1994), and Kato (1995). It can be nested in a one-way manner within an outer coarse-mesh model. The warm rain scheme is generally used as the precipitation parameterization. This model has been applied to several heavy rainfall cases.

Kato(1996) studied a torrential rainfall case, which occurred over the southern Kyushu Island on 1 August 1993. A quasi-stationary rainband was observed for a long time along the Baiu front, where highly humid air was advected from the southwest and the temperature contrast was large. Under such an environmental condition, a quasi-stationary rainband aligned along the front was successfully simulated. First, shallow convective rolls parallel to the middle westerly winds with a space interval of about 25 km appear (Fig. 1), and then the low pressure is induced in the lower layer as they are developed into severe convective systems, forming a convergence line. The low-level jet (LLJ) around the rainband was also simulated. The LLJ was produced by the lower-level low pressure induced by the active convection in the southern front of the rainband.



Fig1
Figure 1: Vertical velocity field at t = 90 min simulated by the NHM with horizontal resolution of 2 km (Kato,1996). The arrows describe horizontal winds near the surface.



A similar event on 6 August 1993 was studied by Saito (1996) by using a newly developed semi-implicit fully compressible version of the NHM.

Seko et al. (1996) studied another quasi-stationary rainband, which persisted for more than 10 hours over the Kanto area when Typhoon 9426 (Orchid) approached the Japan Islands. The back building features of the rainband, such as carrot-shaped pattern, and multicellular structure, were successfully simulated (Fig. 2). They also tried several sensitivity experiments and showed that the horizontal temperature gradient and the vertical wind shear play important roles in organizing the quasi-stationary rainband.



Fig2
Figure 2: (a) Fields of 1 hour precipitation intensity (mm/hr) observed by the Doppler radar at Tsukuba on 29 September 1994 (Seko et al.,1996). JST is Japan standard time (JST = UTC + 9 hours). A large arrow indicates the carrot-shaped rainband and long arrows show the movement of cells in the band. Dotted areas indicate the heavy rain areas more than 35 dBZ. (b) Fields of 1 hour precipitation intensity between 2 h 36 m and 3 h 24 m simulated by the NHM with horizontal resolution of 2 km. Solid lines indicate the movement of cells. The contour interval is 10 mm/hr.



A comparison study of the NHM with the JSM was made by Kato (1996). Generally the JSM well simulates mesoscale precipitation systems with the horizontal scale of larger than 100 km. However, the JSM sometimes predicted falsely moving systems, while the observed ones were nearly stationary. Kato showed that the moist convective adjustment (MCA) in the JSM is a main cause among several candidates for discrepancies, such as coarse resolution (30 km), initialization scheme and model's precipitation scheme. The stabilizing effect of the MCA cools the air in the lower adjusted grids and produces compensating subsidence in the lower layer. This subsidence induces the decrease of moisture near the surface, and breaks the supply of highly humid air needed for the maintenance of the rainband.

4. Future works

As have shown above, our activities are multilateral and on the way to the final ends to forecast the storm-scale prediction of heavy rainfall/snowfall. Although we got satisfactory numerical results for several cases, futher studies are still needed. Applications of numerical models to analysed events are few, and more case studies must be done. Positional errors of precipitation areas were often seen in the nested NHMs, and therefore improvements of the NHMs as well as the parent models are required. Studies of initialization and retrieval methods such as Aonashi and Shibata (1996) and Yoshizaki and Seko (1994) should be also promoted.

To increase case studies, we participated or are going to participate in special observations. In "Tsukuba Area Precipitation Studies", we observed mesoscale precipitation phenomena over the Kanto area for more than three years by ourselves, using upper-air soundings and several remote measurement systems such as Doppler radars, with cooperation of other research institutes and universities. The GPS (Global Positioning System) meteorology will start, in which operational data of PWC with the horizontal resolution of about 20 km are applicable to meso-analyses and initialization of mesoscale prediction models. The "East China Sea - Baiu Project" (ECS-Baiu) is planned to observe cloud clusters along the Baiu front in 1998 and 1999, in which JMA observation vessels, air-borne Doppler radar, and remote-sensing measuring systems on surface will be operated over and around the East China Sea.

In this article, I have introduced numerical works in our group alone. The MRI is actively promoting widespread observational and numerical studies of other mesoscale disturbances such as local fronts, local winds (e.g., Saito 1994), cloud physics and so on.

Acknowledgement

I would like to thank Dr. K. Saito, Mr. K. Aonashi, Mr. I. Takano, Mr. H. Seko and Mr. T. Kato for helpful comments.

References

Aonashi, K. and A. Shibata, 1996: The impact of assimilating SSM/I precipitable water and rain flag data on humidity analysis and short-term precipitation forecasts. J. Meteor. Soc. Japan, 74, 77-99.

Ikawa, M. and K. Saito, 1991: Description of a nonhydrostatic model developed at the Forecast Research Department of the MRI. Tech. Rep. of the MRI, 28, 238 pp.

Kato, T., 1995: A box-Lagrangian rain drop-scheme. J. Meteor. Soc. Japan, 73, 241-245.

Kato, T., 1996: Numerical simulation of line-shaped torrential rainband occurred in the southern Kyushu, Japan on 1 August 1993. Preprints, 7th Conf. on Mesoscale Processes, Reading, UK, Amer. Meteor. Soc., 437-438.

Saito, K., 1994: A numerical study of the local downslope wind "Yamaji-kaze" in Japan. Part 3: Numerical simulation of the 27th September 1991 windstorm with a non-hydrostatic multi-nested model. J. Meteor. Soc. Japan, 72, 301-329.

Saito, K., 1996: Semi-implicit fully compressible version of the MRI mesoscale nonhydrostatic model. -Forecast experiment of the 6 August 1993 Kagoshima torrential rain- (Submitted to J. Meteor. Soc. Japan.)

Seko H., T. Kato, K. Saito, M. Yoshizaki, K. Kusunoki, H. Okamura and Members of Tsukuba Area Precipitation Studies, 1996: Analytical and numerical studies of a quasi-stationary precipitation band observed over the Kanto area associated with Typhoon 9426 (Orchid). (Submitted to J. Meteor. Soc. Japan.)

Takano, I., 1996: A case study of an extratropical cyclone over the Kuroshio current - analysis and simulation. Proc. Autumm Meeting of Japan Meteorological Society (1996), B203.

Yoshizaki, M. and H. Seko, 1994: A retrieval of thermodynamic and microphysical variables by using wind data in simulated multi-cellular convective storms. J. Meteor. Soc. Japan, 72, 31-42.


Dr. Masanori Yoshizaki
Forecast Research Department
Meteorological Research Institute, Japan Meteorological Agency
1-1, Nagamine, Tsukuba, Ibaraki, Japan 305
TEL: +81-298-53-8631
FAX: +81-298-53-8649
E-mail: myoshiza@mri-jma.go.jp