著者

MIYAKAWA Tomoki MIURA Hiroaki

出版者

Meteorological Society of Japan

雑誌

気象集誌. 第2輯 (ISSN:00261165)

巻号頁・発行日

pp.2019-034, (Released:2019-02-11)

被引用文献数

7

---

The properties of tropical convection are evaluated using one-month long simulation datasets produced by the non-hydrostatic icosahedral atmospheric model (NICAM) using 3.5-, 7-, and 14-km horizontal meshes with identical cloud-microphysics configurations. The simulations are targeted on the 2nd Madden-Julian oscillation (MJO) event observed in the CINDY2011/DYNAMO field campaign. An increase of high cloud fraction at 200 hPa level and a reduction of surface precipitation occur as the horizontal resolution increases, corresponding to the reduction of precipitation efficiency due to the shorter residence time inside stronger updrafts that occur at the higher resolution. The increase of high cloud fraction is followed by the warming of the troposphere, which results in an increase in the column water vapor and an elevation of the freezing level. The total water condensation is decreased at higher resolutions, which is likely due to a balance with the decreased outgoing longwave radiation (OLR). The reproduced MJOs, which accounted for a large portion of the tropical convections, were similar in the 3.5-km and 14-km simulations in terms of eastward propagation speeds and structures, including the characteristic westward tilt of the moisture anomaly with height. However, the amplitude of the anomalous MJO circulation was considerably smaller in the 3.5-km simulation. The robust resolution dependence and the interpretations presented in this study underline the necessity for a resolution-aware cloud-microphysics optimization method that will have value in the coming era of global cloud-resolving simulations.

著者

SHIBUYA Ryosuke NAKANO Masuo KODAMA Chihiro NASUNO Tomoe KIKUCHI Kazuyoshi SATOH Masaki MIURA Hiroaki MIYAKAWA Tomoki

出版者

Meteorological Society of Japan

雑誌

気象集誌. 第2輯 (ISSN:00261165)

巻号頁・発行日

pp.2021-046, (Released:2021-04-08)

被引用文献数

5

---

In this study, we assessed the prediction skill of the Boreal Summer Intra-Seasonal Oscillation (BSISO) mode of one-month simulations using a global non-hydrostatic atmospheric model (NICAM) with explicit cloud microphysics and with a grid spacing of 14 km. The simulations were run as a series of hindcast experiments every day of August during 2000-2014; a total of 465 simulations were run with a 13950-day integration. On using forecast skill scores for statistical measurements, it was found that the model showed an overall BSISO prediction skill of approximately 24 days. The prediction skill tended to be slightly higher (∼ 2 days) when BSISO events began in the initial phases 7 to 1, which corresponded to the re-initiation phase of the BSISO, where a major convective center over the Philippine Sea decayed and a new convective envelope began aggregating over the western Indian Ocean. The phase speed and the evolution of the amplitude of the BSISO were well simulated by the model with a clear northwestward-southeastward tilted outgoing longwave radiation (OLR) structure over the Maritime continent and the western Pacific. However, the propagation speed was slower during phases 6-7, and the amplitude of the BSISO largely decayed during phases 8-1, which was likely to have been associated with the stagnant behavior of the convective cells over the Philippines. This stagnation of the propagation over the Philippines may be largely attributed to the small background southerlies bias in the model over the Philippines based on regression coefficient analysis using the moist static energy. The bias in the large-scale circulation was likely to have been associated with the bias in the moisture field and the associated background monsoonal circulation. We concluded that the model physics controlling the background fields are important factors for improving the BSISO prediction skill.

著者

INOUE Toshiro RAJENDRAN Kavirajan SATOH Masaki MIURA Hiroaki

出版者

Meteorological Society of Japan

雑誌

気象集誌. 第2輯 (ISSN:00261165)

巻号頁・発行日

pp.2021-066, (Released:2021-07-02)

被引用文献数

4

---

The dual peak semidiurnal variation in surface rainfall rate over the tropics, simulated by a 3.5-km mesh Non-hydrostatic Icosahedral Atmospheric Model (NICAM) for 26-31 December 2006, is analyzed and compared with data from the 17-year winter precipitation climatology of Tropical Rainfall Measuring Mission (TRMM) TMI (TRMM Microwave Imager), Precipitation Radar (PR), and the same 6-day data of Global Satellite Mapping of Precipitation, as well as infrared data from geostationary satellites. We focus on land areas including southern Africa and the Amazon. Over these land areas, the NICAM simulation captures the primary peak in the afternoon and the secondary peak in the early morning, at similar times to those captured by TRMM data. In the PR observation, the primary peak of rainfall is mainly due to convective rain, whereas the secondary peak is due to stratiform rain. In the NICAM simulation, if a simple method is used for classification of convective/stratiform rain, convective rain is dominant all day long and the rainfall rate is generally higher than in the PR observation. However, an analysis of deep convection (DC) areas indicates consistency between the observation and NICAM; the primary peak of rainfall rate occurs at the mature stage of the number of DC areas, while the secondary peak occurs when the mean size of DC areas is almost at its highest point. In the NICAM simulation, however, the relative magnitudes of the two peaks are not represented well, and the contribution of the stratiform rain is underestimated.  The present study indicates that a high-resolution global nonhydrostatic model like NICAM has the potential to overcome the limitations of coarse-resolution general circulation models by reproducing evolution of deep convection, though there is room for improvement.