Advanced Search
Article Contents

Performance of Convective Parameterization Schemes in Asia Using RegCM: Simulations in Three Typical Regions for the Period 1998-2002


doi: 10.1007/s00376-014-4158-4

  • This study discusses the sensitivity of convective parameterization schemes (CPSs) in the Regional Climate Model (version 4.3) (RegCM4.3) over East/South Asia. The simulations using different CPSs in RegCM are compared to discover a suitable scheme for this region, as the performance of different schemes is greatly influenced by region and seasonality. Over Southeast China and the Bay of Bengal, the Grell scheme exhibits the lowest RMSEs of summer precipitation compared to observed data. Moreover, the Emanuel over land and Grell over ocean (ELGO) scheme enhances the simulation, in comparison with any single CPS (Grell/Emanuel) over Western Ghats, Sri Lanka, and Southeast India. Over the Huang-Huai-Hai Plain (3H) and Tibetan Plateau (TP) regions of China, the Tiedtke scheme simulates the more reasonable summer precipitation with high correlation coefficient and comparable amplitude. Especially, it reproduces a minimum convective precipitation bias of 8 mm d-1 and the lowest RMSEs throughout the year over East/South Asia. Furthermore, for seasonal variation of precipitation, the Tiedtke scheme results are closer to the observed data over the 3H and TP regions. However, none of the CPSs is able to simulate the seasonal variation over North Pakistan (NP). In comparison with previous research, the results of this study support the Grell scheme over South Asia. However, the Tiedtke scheme shows superiority for the 3H, TP and NP regions. The thicker PBL, less surface latent heat flux, the unique ability of deep convection and the entrainment process in the Tiedtke scheme are responsible for reducing the wet bias.
  • 加载中
  • Bao Y., 2013: Simulations of summer monsoon climate over East Asia with a Regional Climate Model (RegCM) using Tiedtke convective parameterization scheme (CPS). Atmos. Res., 134, 35- 44.
    Bao Y., 2013: Simulations of summer monsoon climate over East Asia with a Regional Climate Model (RegCM) using Tiedtke convective parameterization scheme (CPS). Atmos. Res., 134, 35- 44.
    Basit A., S. R. Shoaib, N. Irfan, and R. Avila, 2012: Simulation of monsoon precipitation over South-Asia using RegCM3. ISRN Meteor. , 12, 754902, doi:10.5402/2012/754902.
    Cao J., X. N. Zhang, Y. L. You, and R. W. Yang, 2007: Applicability of cumulus convective parameter schemes in RegCM3 to the rainfall over the Longitudinal Range-Gorge Region. Chinese Science Bulletin, 52( 2), 115- 121.
    Cha D. H., D. K. Lee, and S. Y. Hong, 2008: Impact of boundary layer processes on seasonal simulation of the East Asian summer monsoon using a regional climate model. Meteorology and. Atmospheric Physics, 100( 1-4), 53- 72.
    Chow K. C., J. C. Chan, 2010: A dual-scheme approach of cumulus parameterization for simulating the Asian summer monsoon. Meteorological Applications, 17( 3), 287- 297.
    Dan L., J. Ji, Z. Xie, F. Chen, G. Wen, and J. E. Richey, 2012: Hydrological projections of climate change scenarios over the 3H region of China: A VIC model assessment. J. Geophys. Res, 117(D11). doi: 10.1029/2011JD017131.
    Dash S. K., M. S. Shekhar, and G. P. Singh, 2006: Simulation of Indian summer monsoon circulation and rainfall using RegCM3. Theor. Appl. Climatol., 86( 1-4), 161- 172.
    Davis N., J. Bowden, F. Semazzi, L. Xie, and B. Önol, 2009: Customization of RegCM3 regional climate model for eastern Africa and a tropical Indian Ocean domain. J. Climate, 22, 3595- 3616.
    Dee, D.P., Coauthors, 2011: The ERA-Interim reanalysis: Configuration and performance of the data assimilation system. Quart. J. Roy. Meteor. Soc., 137, 553- 597.
    Dickinson R. E., R. M. Errico, F. Giorgi, and G. T. Bates, 1989: A regional climate model for the western United States. Climatic Change, 15( 3), 383- 422.
    Elguindi N., X. Bi, F. Giorgi, B. Nagarajan, J. Pal, F. Solmon, and G. Giuliani, 2013: Regional Climate Model RegCM User Manual Version 4.4. The Abdus Salam International Centre for Theoretical Physics, Strada Costiera, Trieste, Italy October 21, 2013, 54 pp.
    Emanuel K. A., M. Živkovic-Rothman, 1999: Development and evaluation of a convection scheme for use in climate models. J. Atmos. Sci., 56( 11), 1766- 1782.
    Fu C., S. Wang, Z. Xiong, W. J. Gutowski, D. K. Lee, J. L. McGregor, and M. S. Suh, 2005: Regional climate model intercomparison project for Asia. Bull. Amer. Meteor. Soc., 86( 2), 257- 266.
    Gianotti R. L., D. Zhang, and E. A. Eltahir, 2012: Assessment of the regional climate model version 3 over the maritime continent using different cumulus parameterization and land surface schemes. J. Climate, 25( 2), 638- 656.
    Giorgi F., G. T. Bates, 1989: The climatological skill of a regional model over complex terrain. Mon. Wea. Rev., 117( 11), 2325- 2347.
    Giorgi F., M. R. Marinucci, and G. T. Bates, 1993: Development of a second generation regional climate model (RegCM2). I. Boundary layer and radiative transfer processes. Mon. Wea. Rev., 121, 2794- 2813.
    Giorg F., Coauthors, 2012: RegCM4: Model description and preliminary tests over multiple CORDEX domains. Climate Res., 52, 7- 29.
    Gochis D. J., W. J. Shuttleworth, and Z.-L. Yang, 2002: Sensitivity of the modeled North American monsoon regional climate to convective parameterization. Mon. Wea. Rev., 130, 1282- 1298.
    Grell G. A., 1993: Prognostic evaluation of assumptions used by cumulus parameterizations. Mon. Wea. Rev., 121( 3), 764- 787.
    Grell G. A., J. Dudhia, and D. R. Stauffer, 1994: A description of the fifth-generation Penn State/NCAR mesoscale model (MM5). NCAR Technical Note NCAR/TN-398+STR, doi: 10.5065/D60Z716B.
    Holtslag A. A. M., E. I. F. De Bruijn, and H. L. Pan, 1990: A high resolution air mass transformation model for short-range weather forecasting. Mon. Wea. Rev., 118( 8), 1561- 1575.
    Hourdin, F., Coauthors, 2006: The LMDZ4 general circulation model: Climate performance and sensitivity to parameterized physics with emphasis on tropical convection. Climate Dyn., 27( 7-8), 787- 813.
    Huang W. R., J. C. L. Chan, and A. Y. M. Au-Yeung, 2013: Regional climate simulations of summer diurnal rainfall variations over East Asia and Southeast China. Climate Dyn., 40( 7-8), 1625- 1642.
    Huffman G. J., D. T. Bolvin, E. J. Nelkin, D. B. Wolff, R. F. Adler, G. Gu, and E. F. Stocker, 2007: The TRMM Multi-satellite Precipitation Analysis (TMPA): Quasi-global, multiyear, combined-sensor precipitation estimates at fine scales. Journal of Hydrometeorology, 8( 1), 38- 55.
    Im E. S., J. B. Ahn, A. R. Remedio, and W. T. Kwon, 2008a: Sensitivity of the regional climate of East/Southeast Asia to convective parameterizations in the RegCM3 modelling system. Part 1: Focus on the Korean peninsula. International Journal of Climatology, 28( 14), 1861- 1877.
    Im E. S., W. J. Gutowski, and F. Giorgi, 2008b: Consistent changes in twenty-first century daily precipitation from regional climate simulations for Korea using two convection parameterizations. Geophys. Res. Lett., 35, doi: 10.1029/2008 GL034126.
    Jones R. G., J. M. Murphy, and M. Noguer, 1995: Simulation of climate change over Europe using a nested regional climate model. Part I: Assessment of control climate, including sensitivity to location of lateral boundaries. Quart. J. Roy. Meteor. Soc., 121, 1413- 1449.
    Kaspar F., U. Cubasch, 2008: Simulation of East African precipitation patterns with the regional climate model CLM. Meteorologische Zeitschrift, 17( 4), 511- 517.
    Kiehl J. T., J. J. Hack, G. B. Bonan, B. A. Boville, D. L. Williamson, and P. J. Rasch, 1998: The national center for atmospheric research community climate model: CCM3. J. Climate, 11( 6), 1131- 1149.
    Lau K. M., H. Weng, 2001: Coherent modes of global SST and summer rainfall over China: An assessment of the regional impacts of the 1997-98 El Nino. J. Climate, 14( 6), 1294- 1308.
    Lee D. Y., C. Y. Tam, and C. K. Park, 2008: Effects of multicumulus convective ensemble on East Asian summer monsoon rainfall simulation. J. Geophys. Res., 113(D24), doi: 10.1029/ 2008JD009847.
    Liang X. Z., K. E. Kunkel, and A. N. Samel, 2001: Development of a regional climate model for US Midwest applications. Part I: Sensitivity to buffer zone treatment. J. Climate, 14( 23), 4363- 4378.
    Liang X. Z., M. Xu, K. E. Kunkel, G. A. Grell, and J. S. Kain, 2007: Regional climate model simulation of U.S.-Mexico summer precipitation using the optimal ensemble of two cumulus parameterizations. J. Climate, 20( 20), 5201- 5207.
    Lin R., T. Zhou, and Y. Qian, 2014: Evaluation of global monsoon precipitation changes based on five reanalysis datasets. J. Climate, 27( 3), 1271- 1289.
    Liu Y., Y. Ding, 2007: Sensitivity study of the South China Sea summer monsoon in 1998 to different cumulus parameterization schemes. Adv. Atmos. Sci., 24, 360-376, doi: 10.1007/s00376-007-0360-y.
    Liu Y., Y. Ding, 2007: Sensitivity study of the South China Sea summer monsoon in 1998 to different cumulus parameterization schemes. Adv. Atmos. Sci., 24, 360-376, doi: 10.1007/s00376-007-0360-y.
    Liu H., D. L. Zhang, and B. Wang, 2010: Impact of horizontal resolution on the regional climate simulations of the summer 1998 extreme rainfall along the Yangtze River Basin. J. Geophys. Res., 115(D12), doi: 10.1029/2009JD012746.
    Mamgain A., L. Mariotti, E. Coppola, F. Giorgi, and S. K. Dash, 2013: Sensitivity of RegCM4.3 two convection schemes on Indian summer monsoon for the South Asia CORDEX domain. EGU General Assembly Conference Abstracts, Vol. 15, 4812.
    Martínez-Castro, D., A. Vichot-Llano, A. Bezanilla-Morlot, A. Centella-Artola, 2013: Sensitivity study on convection schemes in regCM4.3, in the simulation of air temperature and precipitation fields in Central America and the Caribbean. International Conference on Regional Climate CORDEX, 4-7 November 2013, Brussels, Belgium.
    Mitchell T. D., P. D. Jones, 2005: An improved method of constructing a database of monthly climate observations and associated high-resolution grids. International Journal of Climatology, 25( 6), 693- 712.
    Octaviani M., K. Manomaiphiboon, 2011: Performance of regional climate model RegCM3 over Thailand. Climate Research, 47( 3), 171- 186.
    Pal J. S., E. E. Small, and E. A. Eltahir, 2000: Simulation of regional-scale water and energy budgets: Representation of subgrid cloud and precipitation processes within RegCM. J. Geophys. Res., 105( D24), 29 579- 29 594.
    Pal J. S., F. Giorgi, X. Bi, N. Elguindi, F. Solmon, S. A. Rauscher, and A. L. Steiner, 2007: Regional climate modeling for the developing world: The ICTP RegCM3 and RegCNET. Bull. Amer. Meteor. Soc., 88( 9), 1395- 1409.
    Rahman M. M., M. N. Islam, A. U. Ahmed, and R. Afroz, 2007a: Comparison of RegCM3 simulated meteorological parameters in Bangladesh: Part I——Preliminary result for rainfall. Sri Lankan Journal of Physics, 8, 1- 9.
    Rahman M. M., M. N. Islam, A. U. Ahmed, and R. Afroz, 2007b: Comparison of RegCM3 simulated meteorological parameters in Bangladesh: Part II——Preliminary result for temperature. Sri Lankan Journal of Physics, 8, 11- 19.
    Reynolds R. W., 1988: A real-time global sea surface temperature analysis. J. Climate, 1, 75- 86.
    Reynolds R. W., D. C. Marsico, 1993: An improved real-time global sea surface temperature analysis. J. Climate, 6, 114- 119.
    Segele Z. T., L. M. Leslie, and P. J. Lamb, 2009: Evaluation and adaptation of a regional climate model for the Horn of Africa: Rainfall climatology and interannual variability. International Journal of Climatology, 29( 1), 47- 65.
    Simmons A., C. Uppala, D. Dee, and S. Kobayashi, 2007: ERA-Interim: New ECMWF reanalysis products from 1989 onwards. ECMWF Newsletter, No. 110, 25- 35.
    Singh A. P., R. P. Singh, P. V. S. Raju, and R. Bhatla, 2011: The impact of three different cumulus parameterization schemes on the Indian summer monsoon circulation. The International Journal of Ocean and Climate Systems, 2, 27- 44.
    Singh G. P., J. H. Oh, J. Y. Kim, and O. Y. Kim, 2006: Sensitivity of summer monsoon precipitation over East Asia to convective parameterization schemes in RegCM3. Sola, 2, 29- 32.
    Sinha P., U. C. Mohanty, S. C. Kar, S. K. Dash, and S. Kumari, 2013: Sensitivity of the GCM driven summer monsoon simulations to cumulus parameterization schemes in nested RegCM3. Theor. Appl. Climatol., 112( 1-2), 285- 306.
    Sinha P., U. C. Mohanty, S. C. Kar, and S. Kumari, 2014: Role of the Himalayan orography in simulation of the Indian summer monsoon using RegCM3. Pure and Applied Geophysics, 171, 1385-1407, doi: 10.1007/s00024-013-0675-9.
    Tiedtke M., 1993: Representation of clouds in large-scale models. Mon. Wea. Rev., 121, 3040- 3061.
    Tiedtke M., 1989: A comprehensive mass flux scheme for cumulus parameterization in large-scale models. Mon. Wea. Rev, 117( 8), 1779- 1800.
    Tchotchou L. A. D., M. Kamga, 2010: Sensitivity of the simulated African monsoon of summers 1993 and 1999 to convective parameterization schemes in RegCM3. Theor. Appl. Climatol., 100, 207- 220.
    Uppala S., D. Dee, S. Kobayashi, P. Berrisford, and A. Simmons, 2008: Towards a climate data assimilation system: Status update of ERA-Interim. European Centre for Medium-Range Weather Forecasts (ECMWF) Newsletter, No. 115, 12- 18.
    Vannitsem S., F. Chomé, 2005: One-way nested regional climate simulations and domain size. J. Climate, 18( 1), 229- 233.
    Wang W., N. L. Seaman, 1997: A comparison study of convective parameterization schemes in a mesoscale model. Mon. Wea. Rev., 125( 2), 252- 278.
    Wang B., R. Wu, 1997: Peculiar temporal structure of the South China Sea summer monsoon. Adv. Atmos. Sci., 14(2), 177-194, doi: 10.1007/s00376-997-0018-9.
    Wang X., Q. Bao, K. Liu, G. Wu, and Y. Liu, 2011: Features of rainfall and latent heating structure simulated by two convective parameterization schemes. Science China Earth Sciences, 54( 11), 1779- 1788.
    Wu W., A. H. Lynch, and A. Rivers, 2005: Estimating the uncertainty in a regional climate model related to initial and lateral boundary conditions. J. Climate, 18( 7), 917- 933.
    Xue Y., R. Vasic, Z. Janjic, F. Mesinger, and K. E. Mitchell, 2007: Assessment of dynamic downscaling of the continental US regional climate using the Eta/SSiB regional climate model. J. Climate, 20( 16), 4172- 4193.
    Yatagai A., K. Kamiguchi, O. Arakawa, A. Hamada, N. Yasutomi, and A. Kitoh, 2012: APHRODITE: Constructing a long-term daily gridded precipitation dataset for Asia based on a dense network of rain gauges. Bull. Amer. Meteor. Soc., 93( 9), 1401- 1415.
    Yhang Y. B., S. Y. Hong, 2008: Improved physical processes in a regional climate model and their impact on the simulated summer monsoon circulations over East Asia. J. Climate, 21( 5), 963- 979.
    Zanis P., C. Douvis, I. Kapsomenakis, I. Kioutsioukis, D. Melas, and J. S. Pal, 2009: A sensitivity study of the Regional Climate Model (RegCM3) to the convective scheme with emphasis in central eastern and southeastern Europe. Theor. Appl. Climatol., 97( 3-4), 327- 337.
    Zhong Z., 2006: A possible cause of a regional climate model's failure in simulating the East Asian summer monsoon. Geophys. Res. Lett. , 33, L24707, doi:10.1029/2006GL027654.
  • [1] Peihua Qin, Zhenghui Xie, Rui Han, Buchun Liu, 2023: Evaluation and projection of population exposure to temperature extremes over the Beijing-Tianjin-Hebei region using a high-resolution regional climate model RegCM4 ensemble, ADVANCES IN ATMOSPHERIC SCIENCES.  doi: 10.1007/s00376-023-3123-5
    [2] Jie WU, Xuejie GAO, Yingmo ZHU, Ying SHI, Filippo GIORGI, 2022: Projection of the Future Changes in Tropical Cyclone Activity Affecting East Asia over the Western North Pacific Based on Multi-RegCM4 Simulations, ADVANCES IN ATMOSPHERIC SCIENCES, 39, 284-303.  doi: 10.1007/s00376-021-0286-9
    [3] Xuejie GAO, Ying SHI, Zhenyu HAN, Meili WANG, Jia WU, Dongfeng ZHANG, Ying XU, Filippo GIORGI, 2017: Performance of RegCM4 over Major River Basins in China, ADVANCES IN ATMOSPHERIC SCIENCES, 34, 441-455.  doi: 10.1007/s00376-016-6179-7
    [4] ZENG Heqing, JIA Gensuo, 2013: Impacts of Snow Cover on Vegetation Phenology in the Arctic from Satellite Data, ADVANCES IN ATMOSPHERIC SCIENCES, 30, 1421-1432.  doi: 10.1007/s00376-012-2173-x
    [5] Xinlin YANG, Jianhua SUN, 2018: Organizational Modes of Severe Wind-producing Convective Systems over North China, ADVANCES IN ATMOSPHERIC SCIENCES, 35, 540-549.  doi: 10.1007/s00376-017-7114-2
    [6] Xue Feng, Zeng Qingcun, 1999: Diagnostic Study on Seasonality and Interannual Variability of Wind Field, ADVANCES IN ATMOSPHERIC SCIENCES, 16, 537-543.  doi: 10.1007/s00376-999-0029-9
    [7] GUAN Zhaoyong, LU Chuhan, MEI Shilong, CONG Jing, 2010: Seasonality of Interannual Inter-hemispheric Oscillations over the Past Five Decades, ADVANCES IN ATMOSPHERIC SCIENCES, 27, 1043-1050.  doi: 10.1007/s00376-009-9126-z
    [8] BUHE Cholaw, Ulrich CUBASCH, LIN Yonghui, JI Liren, 2003: The Change of North China Climate in Transient Simulations Using the IPCC SRES A2 and B2 Scenarios with a Coupled Atmosphere-Ocean General Circulation Model, ADVANCES IN ATMOSPHERIC SCIENCES, 20, 755-766.  doi: 10.1007/BF02915400
    [9] Hu Zengzhen, Tsuyoshi Nitta, 1997: Seasonality of the Interaction between Convection over the Western Pacific and General Circulation in the Northern Hemisphere, ADVANCES IN ATMOSPHERIC SCIENCES, 14, 541-553.  doi: 10.1007/s00376-997-0072-3
    [10] Guoping SHI, Xinfa QIU, Yan ZENG, 2018: New Method for Estimating Daily Global Solar Radiation over Sloped Topography in China, ADVANCES IN ATMOSPHERIC SCIENCES, 35, 285-295.  doi: 10.1007/s00376-017-6243-y
    [11] Qingchang QIN, Xueshun SHEN, Chungang CHEN, Feng XIAO, Yongjiu DAI, Xingliang LI, 2019: A 3D Nonhydrostatic Compressible Atmospheric Dynamic Core by Multi-moment Constrained Finite Volume Method, ADVANCES IN ATMOSPHERIC SCIENCES, 36, 1129-1142.  doi: 10.1007/s00376-019-9002-4
    [12] Mengyu DENG, Riyu LU, Chaofan LI, 2022: Contrasts between the Interannual Variations of Extreme Rainfall over Western and Eastern Sichuan in Mid-summer, ADVANCES IN ATMOSPHERIC SCIENCES, 39, 999-1011.  doi: 10.1007/s00376-021-1219-3
    [13] Xue Feng, Bi Xunqiang, Lin Yihua, 2001: Modelling the Global Monsoon System by IAP 9L AGCM, ADVANCES IN ATMOSPHERIC SCIENCES, 18, 404-412.  doi: 10.1007/BF02919319
    [14] Ho Nam CHEUNG, ZHOU Wen, Hing Yim MOK, Man Chi WU, Yaping SHAO, 2013: Revisiting the Climatology of Atmospheric Blocking in the Northern Hemisphere, ADVANCES IN ATMOSPHERIC SCIENCES, 30, 397-410.  doi: 10.1007/s00376-012-2006-y
    [15] SONG Lianchun, A. J. CANNON, P. H. WHITFIELD, 2007: Changes in Seasonal Patterns of Temperature and Precipitation in China During 1971--2000, ADVANCES IN ATMOSPHERIC SCIENCES, 24, 459-473.  doi: 10.1007/s00376-007-0459-1
    [16] LIU Guimei, WANG Hui, SUN Song, HAN Boping, 2003: Numerical Study on the Velocity Structure around Tidal Fronts in the Yellow Sea, ADVANCES IN ATMOSPHERIC SCIENCES, 20, 453-460.  doi: 10.1007/BF02690803
    [17] Fang Juan, Wu Rongsheng, 2001: Topographic Effect on Geostrophic Adjustment and Frontogenesis, ADVANCES IN ATMOSPHERIC SCIENCES, 18, 524-538.  doi: 10.1007/s00376-001-0042-0
    [18] LI Lijuan, WANG Bin, Yuqing WANG, Hui WAN, 2007: Improvements in Climate Simulation with Modifications to the Tiedtke Convective Parameterization in the Grid-Point Atmospheric Model of IAP LASG (GAMIL), ADVANCES IN ATMOSPHERIC SCIENCES, 24, 323-335.  doi: 10.1007/s00376-007-0323-3
    [19] Xianghui KONG, Aihui WANG, Xunqiang BI, Dan WANG, 2019: Assessment of Temperature Extremes in China Using RegCM4 and WRF, ADVANCES IN ATMOSPHERIC SCIENCES, 36, 363-377.  doi: 10.1007/s00376-018-8144-0
    [20] Zhang Yu, Lu Shihua, 2002: Development and Validation of a Simple Frozen Soil Parameterization Scheme Used for Climate Model, ADVANCES IN ATMOSPHERIC SCIENCES, 19, 513-527.  doi: 10.1007/s00376-002-0083-z

Get Citation+

Export:  

Share Article

Manuscript History

Manuscript received: 11 July 2014
Manuscript revised: 20 September 2014
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Performance of Convective Parameterization Schemes in Asia Using RegCM: Simulations in Three Typical Regions for the Period 1998-2002

  • 1. START Temperate East Asia Regional Center and Key Laboratory of Regional Climate-Environment for Temperate East Asia, Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing 100029
  • 2. University of Chinese Academy of Sciences, Beijing 100049
  • 3. Global Change Impact Studies Centre, Ministry of Climate Change, Islamabad, Pakistan

Abstract: This study discusses the sensitivity of convective parameterization schemes (CPSs) in the Regional Climate Model (version 4.3) (RegCM4.3) over East/South Asia. The simulations using different CPSs in RegCM are compared to discover a suitable scheme for this region, as the performance of different schemes is greatly influenced by region and seasonality. Over Southeast China and the Bay of Bengal, the Grell scheme exhibits the lowest RMSEs of summer precipitation compared to observed data. Moreover, the Emanuel over land and Grell over ocean (ELGO) scheme enhances the simulation, in comparison with any single CPS (Grell/Emanuel) over Western Ghats, Sri Lanka, and Southeast India. Over the Huang-Huai-Hai Plain (3H) and Tibetan Plateau (TP) regions of China, the Tiedtke scheme simulates the more reasonable summer precipitation with high correlation coefficient and comparable amplitude. Especially, it reproduces a minimum convective precipitation bias of 8 mm d-1 and the lowest RMSEs throughout the year over East/South Asia. Furthermore, for seasonal variation of precipitation, the Tiedtke scheme results are closer to the observed data over the 3H and TP regions. However, none of the CPSs is able to simulate the seasonal variation over North Pakistan (NP). In comparison with previous research, the results of this study support the Grell scheme over South Asia. However, the Tiedtke scheme shows superiority for the 3H, TP and NP regions. The thicker PBL, less surface latent heat flux, the unique ability of deep convection and the entrainment process in the Tiedtke scheme are responsible for reducing the wet bias.

1. Introduction
  • Efforts have been made in recent decades to develop and improve regional climate models (RCMs), but many problems remain; for example, the lateral boundary conditions (Liang et al., 2001; Wu et al., 2005), horizontal resolution, size match of a region (Vannitsem and Chomé, 2005; Xue et al., 2007; Liu et al., 2010), and the parameterization schemes of unique physical processes (Cha et al., 2008; Yhang and Hong, 2008). Moreover, the selection of an appropriate convective parameterization scheme (CPS) in RCMs is a major source of error and has significant impact on regional climate model predictions (Pal et al., 2007). The Regional Climate Model Intercomparison Project (RMIP) for Asia concluded that the simulations of RCMs are highly sensitive to CPS choices (Fu et al., 2005). Generally, no single CPS can perform well universally for all atmospheric systems because, for instance, the processes of convection in the tropics can be significantly different from those in the midlatitudes (Wang and Seaman, 1997; Singh et al., 2006; Chow and Chan, 2010). Many studies have been conducted in different regions to validate the sensitivity of CPSs, e.g., America (Gochis et al., 2002; Liang et al., 2007), Europe (Wang and Seaman, 1997; Zanis et al., 2009), Africa (Davis et al., 2009; Segele et al., 2009; Tchotchou and Kamga, 2010), the Caribbean region (Martínez-Castro et al., 2013), the Maritime Continent (Gianotti et al., 2012), and South and East Asia (Dash et al., 2006; Chow and Chan, 2010; Singh et al., 2011; Sinha et al., 2014).

    (Wang and Seaman, 1997) reported inconclusive results when they compared four CPSs, with no single scheme emerging as the best. The Regional Climate Model (RegCM) with a horizontal resolution of 55 km was used by (Dash et al., 2006) for a four-year simulation (1993-96), and the results showed that the Grell scheme performed better for simulating the summer precipitation over India. Using the Anthes-Kuo scheme in RegCM produced comparable results for area-averaged precipitation over the Caribbean (Martínez-Castro et al., 2013). (Singh et al., 2006) statistically analyzed the climate over East Asia. Their results indicated that the Emanuel and Grell schemes minimize the biases and performs well over this region, especially Korea, but that issues of overestimation still remain. (Liang et al., 2007) simulated the summer precipitation over the U.S. and Mexico using an ensemble of the Grell and Kain-Fritschl cumulus schemes. This approach produced far superior performance and considerable improvement was achieved compared to each individual scheme over the entire study domain. RegCM was customized with the Emanuel scheme for precipitation simulation by (Davis et al., 2009). This model configuration predicted the rainfall over eastern Africa and the tropical Indian Ocean more realistically, but overestimation of precipitation also occurred. (Segele et al., 2009) concluded that the Emanuel scheme performs better when selecting 1984 as a dry year and 1996 as a wet year. (Octaviani and Manomaiphiboon, 2011) demonstrated that the Emanuel scheme performs well, followed by the Anthes-Kuo scheme, when a double-nested 60 and 20 km resolution domain is used. According to (Basit et al., 2012), the Grell scheme captures well the monsoon phenomenon, especially for the mountainous regions of North Pakistan, with the Arakawa-Schubert (AS) and Fritsch-Chappell (FC) closures both performing satisfactorily. (Huang et al., 2013) simulated the diurnal variation of rainfall over Southeast China during 1998-2002, and reported better results when using the Grell scheme; whereas, the simulation of summer mean rainfall over East Asia was more realistically simulated using the Emanuel scheme. The CPSs play an important role in the simulation of summer precipitation in monsoon regions and the performances of CPSs have numerous uncertainties (Lee et al., 2008; Bao, 2013).

    Table 1 provides a list of recent RegCM recent studies on CPSs over East, South, and Southeast Asia. Most of the studies were based on model simulations of 1-5 years (Dash et al., 2006; Singh et al., 2006; Im et al., 2008a; Basit et al., 2012), but with some based on 10 years or more (Cao et al., 2007; Im et al., 2008b). Horizontal resolutions of 50 km or more have been used in several studies (Dash et al., 2006; Rahman et al., 2007a, 2007b; Liu and Ding, 2007; Chow and Chan, 2010; Huang et al., 2013; Bao, 2013). However, few studies have also used a resolution of less than 50 km (Singh et al., 2006; Gianotti et al., 2012) in a nested domain setup (Im et al., 2008a; Octaviani and Manomaiphiboon, 2011). Studies show that CPSs present good results with a smaller model domain over East Asia, but they are unable to model tropical cyclones and extreme events (Zhong, 2006). Convective precipitation derived from different convection parameterizations is a major contributor to the performance of some models in the summer season (Im et al., 2008a). In simulating the Asian summer monsoon, the dual scheme approach enhances model performance (Chow and Chan, 2010). Errors in models persist irrespective of the choice of CPS or land surface scheme and lateral boundary conditions (Gianotti et al., 2012). The performances of CPSs improve with a smaller horizontal resolution; for example, (Sinha et al., 2013) reported results that were better when using 30 km rather than 90 km. Overall, the previous results summarized in Table 1 suggest that the Grell scheme is best for South Asia, the Emanuel scheme for East Asia, while both are suitable for Southeast Asia.

    Figure Table 1.  Overview of relevent studies using CPSs in RegCM over Asia

    In the present study, sensitivity experiments using five different CPSs [(1) Grell (Grell, 1993); (2) Emanuel (Emanuel and Živkovic-Rothman, 1999); (3) Tiedtke (Tiedtke, 1989); (4) Emanuel over land and Grell over ocean (ELGO); and (5) Grell over land and Emanuel over ocean (GLEO)] in RegCM are carried out to discover a suitable scheme for East/South Asia. In some previous studies, the Emanuel scheme has been found to perform better over the ocean, but tends to produce excessive precipitation over land. Besides, the Grell scheme reportedly performs better over land but produces weak precipitation over the tropical ocean (Elguindi et al., 2013). On the other hand, in some regions, the Grell scheme over the ocean shows significantly improved simulation results (Mamgain et al., 2013). For these reasons, the options of mixed convection (i.e. GLEO and ELGO) are added to the model to enhance its performance. After introducing the component schemes (i.e. Grell, Emanuel, Tiedke) in section 2, we then describe the model and methods used in section 3. The results are presented in section 4, before summarizing the key conclusions of the study in section 5.

2. The convective parameterization schemes
  • (Grell, 1993) represents cloud and environment as two steady-state circulations (an updraft and a downdraft that are undiluted, i.e., no entrainment or detrainment occurs along the cloud edges such that the mass flux is constant with height) in the Grell scheme. The scheme can be used when a lifted parcel reaches moist convection level, and the mixing of cloud, air, and the environment occurs only at the top and bottom. This scheme is very simple in nature and focuses on the statistical equilibrium between large-scale processes and convection. The minimum and maximum levels of moist static energy indicate the initiating levels of the downdraft and updraft. The Grell scheme activates when a lifted parcel of air reaches the condensation and moist convection level in the updraft, and is calculated by a saturated parcel. M d denotes the downdraft mass flux and M u represent the updraft mass flux. Hence, the scheme can be shown in the form of an equation:

    $$ M_{\rm d}=\dfrac{\beta I_1}{I_2}M_{\rm u} . $$

    The normalized updraft condensation is denoted by I1, the normalized downdraft evaporation is I2, and β is the fraction of updraft condensation that re-evaporates during the downdraft and depends on the change in the wind direction and speed that typically changes between 0.3 and 0.5. Hence, we have the equation

    $$ P=I_1m_{\rm b}(1-\beta) , $$

    where P is precipitation. The detrainment and mass fluxes at the top and bottom of the cloud regulate the moistening and heating in the Grell scheme with the cooling effect during the downdrafts. The Grell scheme is flexible to apply with different assumptions [i.e. Arakawa-Schubert (AS) and Fritsch-Chappell (FC)].

  • The Emanuel scheme assumes that mixing in clouds is highly inhomogeneous and episodic. Modeling of convective fluxes is based on an idealized condition of sub-cloud-scale downdrafts and updrafts. The process of convection is initiated if the level of neutral buoyancy is greater than the cloud base level. Air is lifted between these two levels and a fraction of the condensed vapor forms precipitation, whereas the residual forms the cloud (Elguindi et al., 2013). It is assumed that the cloud is mixed with the air from the environment permitting to the range of mixtures that descend or ascend to particular levels of neutral buoyancy. The mixing detrainment and entrainment rates are the functions of the vertical gradients of buoyancy in clouds. The fraction of the total cloud base mass flux that combines with the environment at all levels is proportional to the rate of change of buoyancy with altitude. The upward mass flux at the cloud base decreases towards the sub-cloud layer quasi equilibrium. The Emanuel scheme is designed in such a way that it provides several advantages regarding convection options. It consists of a technique that converts the cloud water into precipitation in cumulus clouds. The ice processes are temperature-dependent, allowing the auto-conversion threshold water content to be used. The precipitation is included in hydrostatic, unsaturated, and a single downdraft that carries water and heat. The Emanuel scheme also considers the carrying of passive tracers.

  • Tiedtke scheme was originally developed for application on the global scale. This scheme is dependent on the mass flux and moisture convergence. The triggering mechanism is based on convection, i.e. if the temperature of the parcel exceeds the temperature of the environment by a fixed temperature threshold, it creates conditional instability. It has shallow and deep convection, determining the cloud base mass flux from the PBL equilibrium and mass flux closure from CAPE, respectively. In deep convection, the reduction of CAPE is the integrated effect of the convective heating, and the middle level exists if there is large-scale ascent with mass flux closure. In the Tiedtke parameterization, updraft is sensitive to entrained air from the free troposphere (Hourdin et al., 2006). Thus, the convection can be reduced by the dry, free troposphere. The convective cloud cover and cloud water content sources are described as functions of the detrainment of mass from the specific content of cloud water in the updrafts, convective updrafts, and the density of cloudy air. An updraft air parcel is assumed to detrain into existing cloud as well as into cloud-free air, making sure the real limits of zero cloud cover and full cloud cover. Zero cloud cover is considered in which updraft air detrains only into clear air and full cloud cover is considered where all updraft air detrains into existing clouds. The mass for detrainment is acquired from the cumulus parameterization for the updraft mass flux (Tiedtke, 1993). Cumulus clouds can occur if a deep layer of conditional instability and large-scale moisture convergence exist. Within the lower half of the troposphere, an increase in vertical mass flux is connected to the moisture convergence in the column of the atmosphere. Hence, deep convection occurs as the undiluted parcel of air rises adiabatically with the dry adiabatic lapse rate that positively regulates the buoyancy until it touches the LCL and becomes saturated. The precipitation will occur due to turbulent eddies and stronger large-scale moisture convergence (Tiedtke, 1989).

3. Methods
  • The model used in this study is version 4.3 of RegCM. The first version of RegCM was developed in the late 1980s (Giorgi and Bates, 1989; Dickinson et al., 1989). Since then, it has been continuously upgraded to RegCM2 (Giorgi et al., 1993), RegCM3 (Pal et al., 2000, 2007) and RegCM4 (Giorgi et al., 2012). Earth System Physics (ESP), Abdul Salam International Center for Theoretical Physics (ICTP), Italy, maintains the latest version, i.e., RegCM4. The model's latest versions are more user-friendly, flexible, and portable, having important updates made to the source code. The dynamic core of RegCM4 is similar to RegCM3, which was based on the hydrostatic version of the National Center for Atmospheric Research (NCAR)/Penn State mesoscale model (MM5) (Grell et al., 1994). The PBL computations are parameterized by the scheme of (Holtslag et al., 1990); the land surface model is the Biosphere-Atmosphere Transfer Scheme (BATS) (Dickinson et al., 1989); the radiation scheme is the modified NCAR Community Climate Model version 3 (CCM3) (Kiehl et al., 1998); and the resolvable-scale precipitation is signified by the system of subgrid explicit moisture scheme (SUBEX) (Pal et al., 2000).

  • The model domain covers Southeast Asia over the area (4°-56°N, 49°-138°E). The climatology of this region is mainly affected by the monsoon season. Over most parts of the Southeast Asian region, rainfall occurs during the months of June to September, also known as the summer monsoon, depending on the seasonal direction of the prevailing surface wind. High rainfall associated with regional orography occurs parallel to India, the west coast of the Indochina peninsula, South China, and the South China Sea. There is a wide zone of around 20°N extending towards the northeast from the north part of the Bay of Bengal (BoB) that receives substantial rainfall. Southeast Asian summer monsoon rainfall is linked with the variability of rainfall in this monsoon zone (Wang and Wu, 1997). The rainfall over monsoon zone is associated with the synoptic-scale convective systems produced over the warm oceans surrounded by the subcontinent that transport moisture to the northeast. Partly overlapping disturbances are also generated over this zone due to active spells of the monsoon. The summer monsoon even influences the global climate through energy exchange and hydrological processes, as well as the regional climate (Lau and Weng, 2001).

    Three sub-regions are selected from the model domain for detailed analysis (Fig. 1): North Pakistan [NP: (34.5°-37°N, 71°-78°E)]; Tibetan Plateau [TP: (30°-37°N, 80°-95°E)]; and Huang-Huai-Hai Plain [3H: (31°-41°N, 112°-121°E)]. The selected sub-regions consist of two mountain areas (NP and TP) and one irrigated plain (3H) region (Dan et al., 2012) in China. 3H is occupied by 425 million people and provides 50% of China's grain, where the crop production is greatly dependent on irrigation from groundwater and surface runoff. NP and TP consist of the Himalaya, Hindu Kush, and Karakoram, with the world's third-largest ice store. The domain contains 336 grid points in the latitudinal direction and 280 grid points in the longitudinal direction, with a central point of (32°N, 94°E) and 18 vertical sigma levels. This domain is quite large for the development of internal model mesoscale circulations and regional forcing (Jones et al., 1995; Singh et al., 2006).

    RCMs are usually run at high resolution when the output of RCMs is used for impact studies. In this study, RegCM4 is run at 20 km at its maximum finer resolution because the hydrostatic engine of RegCM does not allow a resolution finer than 20 km (Elguindi et al., 2013). Five simulations with five different CPSs are carried out for the period 1997-2002. The first year, 1997, is used as model spin-up, and so the analysis is considered from 1998 to 2002. The lateral and initial boundary conditions are taken from ERA-interim data (Simmons et al., 2007; Uppala et al., 2008), available since 1979 to the present day with 1.5° horizontal resolution. The ERA-interim dataset is a realistic state-of-the-art and widely used model solution. It has good coverage of upper-air measurements over the whole globe and is regarded as relatively more reliable than other reanalysis datasets (Lin et al., 2014), especially over the Northern Hemisphere. It also assimilates more advanced and observational data (Dee et al., 2011).

    Figure 1.  Model domain and topography (units: m). Rectangles show the three sub-regions of our study: North Pakistan (NP); Tibetan Plateau (TP); Huang–Huai–Hai Plain (3H).

    Figure 2.  1998-2002 mean of precipitation (mm d-1) and wind (m s-1) at 850 hPa: annual (left panels), MAMJJA (middle panels), SONDJF (right panels); in (a-c) ERA-Interm's wind and TRMM's precipitation, (d-f) the Tiedtke scheme, (g-i) the Grell scheme, (j-l) the Emanuel scheme, (m-o) the GLEO scheme, and (p-r) the ELGO scheme. Vectors indicate wind, colors represent precipitation.

    The SST data are acquired by optimum interpolation (OI) processing, i.e., the weighted monthly mean of the given set of data on the basis of the observed Reynolds SST field where the variance of the estimate is kept to the minimum. The OI SST analysis is produced weekly on a 1° grid. The analysis uses in situ and satellite SSTs, plus SSTs simulated by sea-ice cover. Before the analysis is computed, the satellite data are adjusted for biases using the method of (Reynolds, 1988) and (Reynolds and Marsico, 1993). Topographic and vegetation cover datasets was taken from 30-arc-second (GTOPO30), aggregated to 10 arc minutes and global land cover characterization (GLCC) respectively. To validate the model results, three observational datasets are used for precipitation and temperature including: monthly-mean and 3-hourly datasets available from 0.25°×0.25° products from the Tropical Rainfall Measurement Mission (TRMM 3B42, Huffman et al., 2007), over the range 50°N-50°S since January 1998 to the present day. TRMM 3A25 data with distinct convective and stratiform rainfall are used to validate the convective precipitation. The monthly mean temperature and precipitation data for the period 1901-2000 over the global land area at regular intervals of 0.5° are from the Climate Research Unit (CRU) TS2.1 dataset (Mitchell and Jones, 2005). The gridded precipitation data for the period 1957-2007 are from the Asian Precipitation Highly-Resolved Observational Data Integration Towards Evaluation of water resources (APHRODITE) dataset (Yatagai et al., 2012), provided by the Meteorological Research Institute of the Japan Meteorological Agency (MRI/JMA) and the Research Institute for Humanity and Nature (RIHN) at 0.25° resolution.

    Figure 3.  1998-2002 mean difference of precipitation and wind at 850 hPa between RegCM and TRMM (shaded, mm d-1) and ERA-Interim (vectors, m s-1), separately, in MAM (left panels), JJA (middle-left panels), SON (middle-right panels) and DJF (right panels) with (a-d) the Tiedtke scheme, (e-h) the Grell scheme, (i-l) the Emanuel scheme, (m-p) the GLEO scheme, and (q-t) the ELGO scheme.

4. Results
  • The effects of different CPSs on average temperature, precipitation, and winds at 850 hPa during a five-year period (1998-2002) are presented in this section. First, we describe the simulation results over the whole of the East/South Asia domain for the four seasons defined as: spring (March-May; MAM); summer (June-August; JJA); autumn (September-November; SON); and winter (December-February; DJF). And second, we analyze the statistical results, including correlation coefficient, normalized standard deviation, and RMSE values, over the three sub-regions (3H, NP, and TP).

  • The seasonal and annual means of precipitation and wind are examined to understand the moisture flow from one region to another (Fig. 2). It can be seen that RegCM4.3 is able to properly reproduce the circulation for the East/South Asian monsoon. During MAM-JJA, atmospheric pressure decreases over the Asian landmass due to intense surface heating, and at the same time surface pressure remains relatively high over the cool sea to the south. Moist air from the equator moves northward under this seasonally developed pressure gradient carrying water vapor from the BoB and Arabian Sea. South/southwestward wind patterns develop due to the Coriolis force and prevail over South/Southeast Asia during the summer monsoon from May to September. During SON-DJF, north/northeast winds move from northern high latitudes affected by the Siberian high, carrying cold air from northern land to most parts of East/South Asia and causing winter precipitation. The annual means of precipitation and circulation are more affected by MAM-JJA, whereas SON-DJF has less impact. The annual mean and MAM-JJA circulation patterns are more similar in Fig. 2, which clearly indicates that MAM-JJA is dominant in terms of the annual mean of circulation and precipitation.

    Figure 4.  1998-2002 mean of surface latent heat flux (W m-2) of MAM (left panels), JJA (middle-left panels), SON (middle-right panels), and DJF (right panels) in (a-d) the Tiedtke scheme, (e-h) the Grell scheme, (i-l) the Emanuel scheme, (m-p) the GLEO scheme, and (q-t) the ELGO scheme.

    The results of RegCM4.3-simulated precipitation and winds are compared with TRMM and ERA-Interim data (Fig. 3) separately. It is clear that the Grell simulations produce more rainfall over the Arabian Sea and BoB throughout the year, whereas other schemes show a dry bias. In summer, the Emanuel and Tiedtke schemes underestimate the precipitation over the Indian Peninsula, whereas the Grell scheme overestimates precipitation by 6 mm d-1 over south India due to the local cyclone for air convergence. Furthermore, over southwest China, the Indochina peninsula, and Northeast China, the Emanuel scheme produces more precipitation and the Tiedtke scheme produces less precipitation over South China, probably due to less surface latent heat flux (Fig. 4) and entrained air from the troposphere. In winter, all the schemes produce less biased results of precipitation as compared to summer due to less convective activity. Table 2 lists the seasonal RMSEs of precipitation for the five schemes over the whole domain. The Grell scheme shows a small RMSE value for precipitation of 3.41 mm d-1 for the summer season (JJA), whereas the Emanuel scheme shows a high RMSE of 5.20 mm d-1, followed by GLEO (4.57 mm d-1). During MAM, the Tiedtke scheme shows the lowest RMSE values in comparison to the other CPSs. In winter, the Emanuel and Tiedtke schemes have low RMSE values for precipitation (∼1.7 mm d-1). RMSE values are higher during summer precipitation than in other seasons. Table 3 lists the precipitation RMSE values on a seasonal basis over the southeast China (SEC), BoB, the Western Ghats (WG), and Sri Lanka/southeast India (SLSEI). For SEC, the Grell scheme performs effectively compared to the other schemes during JJA, SON, and DJF, while the Tiedtke scheme performs better for MAM. Over the BoB, the Grell scheme shows effective results for JJA and SON, while GLEO and the Tiedtke scheme improve their outcomes for MAM and DJF, respectively. Over WG, the Grell scheme shows enhanced performance during MAM, SON, and DJF. However, the ELGO scheme produces considerable results over WG and SLSEI compared to each standalone scheme (i.e. Grell and Emanuel) during JJA, indicating that in regions that contain both land and ocean, the combination of the Grell scheme over the ocean and the Emanuel scheme over land produces suitable results.

    The amount of convective precipitation is very sensitive and directly influenced by the choice of convection scheme (Wang and Seaman, 1997). Figure 5 illustrates the differences of seasonal mean convective precipitation between RegCM4.3 and TRMM 3A25 data. It is clear that in JJA the Tiedtke scheme produces the least biased convective precipitation. However, it overestimates rainfall by 4-6 mm d-1 over southern China and adjoining areas, and underestimates precipitation over some parts of India and NP. Figure 5 also indicates that the Grell, Emanuel, GLEO and ELGO schemes overestimate precipitation by 4-12 mm d-1 over the Indian peninsula, SEC, and adjoining areas. The Emanuel scheme shows a precipitation bias of 4-6 mm d-1 over India and the BoB, while over SEC and the Indochina peninsula it generates more convective precipitation and the bias exceeds 12 mm d-1. The ELGO scheme reduces the wet bias over India and the Indochina peninsula, and produces the second-smallest wet bias over Southwest China. The GLEO scheme produces a small dry bias over the BoB, close to that of the Emanuel scheme, but over India and the Indochina peninsula it shows a large wet bias. In comparison with TRMM data, the Tiedtke scheme reproduces well-distributed seasonal variability and intensity of convective precipitation, with the smaller bias probably due to less convection in the scheme. The Tiedtke scheme also produces minimum RMSE values for all the seasons with TRMM (Table 4).

    Figure 5.  As in Fig. 3, but for convective precipitation (mm d-1).

    The seasonal mean temperatures at the height of 2 m in the five schemes along with CRU data are shown in Fig. 6. Overall, the simulated surface air temperatures with the five CPSs agree closely with CRU data throughout the year. The results show that RegCM4.3 is able to capture the mean state and seasonal cycle of the surface air temperature over East/South Asia.

    Figure 6.  1998-2002 mean temperature at 2 m (°C) of MAM (left panels), JJA (middle-left panels), SON (middle-right panels) and DJF (right panels) in (a-d) CRU data, (e-h) the Tiedtke scheme, (i-l) the Grell scheme, (m-p) the Emanuel scheme, (q-t) the GLEO scheme, and (u-x) the ELGO scheme.

    In order to study the inter-model differences in temperature at the height of 2 m, the differences between RegCM4.3 when using the five CPSs and CRU data are shown in Fig. 7. Also shown are the differences between ERA-Interim and CRU data (Figs. 7u-x). In general, surface air temperature bias patterns are similar in RegCM4.3 with the different CPSs. All the schemes result in a similar cold bias throughout the year over the west of the TP and NP. The possible reasons for this cold bias are: (1) the difference in the representation of elevation in the model and actual station values; (2) the rapid variation in the topography of mountainous regions, e.g., the problem of accommodating and processing orographic lifting is very significant in the NP; and (3) the actual height of the Himalaya is more than prescribed in RegCM (Sinha et al., 2014). The Tiedtke scheme shows a relatively smaller cold bias among the five CPSs in summer. For surface air temperature during summer, the Tiedtke scheme even shows a warm bias (more heating) because of more radiation in the surrounding regions and a high rate of evaporation and convection, which later transports moisture towards the northeast. Increased heating makes the atmosphere warmer over most parts of the domain when compared to other schemes that show cold bias. The Tiedtke scheme results show a thick PBL over India and most parts of China (not shown) with less precipitation. Since convection is primarily initiated within the PBL region that responds to surface turbulent heat fluxes driven by incoming solar radiation after relatively short time scales, convective activity affects large-scale atmospheric dynamics by vertical transport of heat and moisture and adiabatic heating. It is also influenced by the interaction of cumulus clouds with radiation. The resilient contact of cumulus clouds with longwave and shortwave radiation causes atmospheric motion, moisture convergence and dispersion over the surface as well as in the atmosphere. This moisture is carried to the northeast, leaving most parts of India and China relatively drier. In winter, all the schemes produce a cold bias over SEC and the Indochina peninsula, with a center over Thailand, west of the TP, and Southwest China. The temperature bias patterns in RegCM are similar in different CPSs. The Grell, Emanuel, ELGO and GLEO schemes exhibit RMSE values between 2.09 and 2.47°C during summer (JJA) (Table 2), whereas the Tiedtke scheme shows the highest RMSE value of 3.11°C and produces worse results than the other schemes. In winter, the Tiedtke and Grell schemes show smaller RMSEs. It is also clear that the RMSE values for temperature are larger in winter than in other seasons.

    Figure 7.  As in Fig. 3, but for temperature at 2 m (°C) along with the difference between ERA-Interim and CRU data in last bottom panels (u, v, w, x).

  • The results of seasonal variation of precipitation over the three sub-regions (3H, TP and NP) (Fig. 1) are presented in Fig. 8. TRMM and APHRODITE data show more precipitation in the region of 3H but less in TP and NP. The model results in Fig. 8 for JJA also show similar results. Over the 3H region, precipitation reaches the maximum, i.e., around 4mm d-1 in July and the minimum in December. The Grell and GLEO schemes show that precipitation in summer exceeds 6 mm d-1, while in the Emanuel and ELGO scheme results its amplitude reaches double that of the APHRODITE data. The seasonal variation of the Tiedtke scheme produces simulation results closer to TRMM and APHRODITE data. The results over the TP are similar to the 3H region, with all CPSs showing more precipitation throughout the year except the Tiedtke scheme, which shows a dry bias in summer. Small variations in precipitation are shown according to APHRODITE and TRMM data for the NP. APHRODITE produces maximum precipitation in April, whereas TRMM produces maximum precipitation in July. The seasonal variation of simulated precipitation shows larger differences in winter than summer if we compare with TRMM and APHRODITE. None of the CPSs generate results that reflect the observed amplitude variation (high and low) of rainfall in February and August. Overall, in the three sub-regions, the simulation results of the Grell (Emanuel) and GLEO (ELGO) schemes are quite similar, which share the same CPS over land, and are therefore influenced directly by the CPS over land rather than over the ocean.

    Figure 8.  1998-2002 mean of monthly precipitation (mm d-1) over the three sub-regions (a, 3H; b, TP; c, NP) of the Tiedtke, Grell, Emanuel, GLEO, ELGO, APHRODITE and TRMM.

    Vertical profiles of summer temperature and water vapor mixing ratio are compared with ERA-interim data (Fig. 9). Over the 3H region, the Tiedtke scheme shows warmer temperatures near the surface up to 500 hPa associated with the maximum temperature at the height of 2 m for the same region (Fig. 7). Above 500 hPa, the temperature differences become very small. Over the TP and NP, the warmer temperature in the Tiedtke scheme reaches up to the height of 200 hPa, and the differences between RegCM4.3 and ERA-Interim show warmer temperatures at this level compared to the lower troposphere. The warm difference (RegCM minus ERA-Interim) decreases as the altitude increases up to 600 hPa, and even turns to cold differences in the middle troposphere by the Grell and GLEO schemes above 600 hPa. This temperature decrease is due to frequent convective activities in the Grell scheme. Furthermore, the differences in temperature increase and attain maximum values of around 12°C at 150 hPa. The Grell and GLEO schemes show more unstable atmospheric conditions when the altitude increases up to 600 hPa. In the Tiedtke scheme the slow convective process decreases the ascending motion of the air before the condensation level and creates a more stable structure with less precipitation, possibly due to less surface latent heat flux and entrainment. This reduces the wet bias of convective precipitation in the Tiedtke scheme (Fig. 5). The Tiedtke scheme retains more water vapor in the troposphere but shows a reduced precipitation difference (Fig. 3b). The underestimation of precipitation in the Tiedtke scheme was also reported by (Kaspar and Cubasch, 2008), (Wang et al., 2011), and (Martínez-Castro et al., 2013). The atmosphere is drier in the Tiedtke scheme at 400-700 hPa, as shown in Fig. 9e, and at 850 hPa in Fig. 9f. The Emanuel scheme is more active in transporting humid air from low levels to high levels, generating more precipitation.

    Figure 9.  1998-2002 mean area-averaged vertical profiles of the difference of temperature (upper panels, °C) and water vapor mixing ratio (lower panels, g kg-1) between RegCM and ERA-interim in JJA for the five schemes.

    Figures 10a and b show the correlation coefficient and normalized standard deviation results for summer surface air temperature and precipitation, respectively, with the five CPSs over the 3H, TP and NP sub-regions. For surface air temperature, the simulations of the different CPSs are closer to each other and the correlation ranges from 0.72 to 0.97 compared with CRU data. Over the 3H, the Emanuel and ELGO schemes show strong correlation and standard deviation of temperature which is approximately 1, while over the NP and TP, the Grell scheme shows strong correlation and a normalized standard deviation ratio of closer to 1. For precipitation, the correlation is weaker than for temperature over the three sub-regions, and the correlation and standard deviation show large differences in different schemes. Over the 3H region, the correlation of the Grell and Tiedtke schemes with TRMM data is about 0.6, which is double that for the Emanuel scheme (0.3). Over the TP, the Tiedtke scheme produces strong positive correlation (0.8) and standard deviation closer to 1 in simulating summer precipitation. Over the NP, all the CPSs show very weak correlation, i.e., less than 0.3, and low standard deviation for summer precipitation.

    Figure 10.  Taylor diagram for (a) mean temperature at 2 m and (b) mean precipitation during JJA for the period 1998-2002 over the three sub-regions (1, 3H; 2, TP; 3, NP) of the Tiedtke, Grell, Emanuel, GLEO and ELGO schemes.

    RMSE values are calculated for all schemes over the three sub-regions for precipitation and temperature (Table 5). The Tiedtke scheme performs better in simulating precipitation over the 3H region all year round, especially in MAM and JJA. For temperature in the 3H region, ELGO produces better results for JJA, SON, and DJF. Over the NP, the Tiedtke scheme produces good approximations of precipitation except for JJA. The ELGO scheme produces better results for JJA compared to the Tiedtke scheme. Over the TP, the Tiedtke scheme produces appropriate results throughout the year for precipitation and temperature.

5.Conclusion
  • RegCM4.3 successfully simulates the characteristics of East/South Asia summer monsoon circulation at 850 hPa. The performance of all CPSs is strongly dependent on spatial and seasonal variation of a region. Over SEC and the BoB, the Grell scheme performs more effectively than the other schemes. However, all the schemes tend to show a dry bias over the Arabian Sea and BoB throughout the year, except the Grell scheme which produces surplus rainfall. Over WG, the Grell scheme shows better performance throughout the year except in JJA, when the ELGO scheme produces reasonable results. The Tiedtke scheme improves the simulation of convective precipitation in both spatial distribution and magnitude for all the seasons with minimum RMSE values. In the case of surface air temperature, all the CPSs produce almost the same cold bias (0-10°C) over the west of the TP and NP throughout the year.

    Generally, the Tiedtke scheme shows close agreement with TRMM and APHRODITE data for precipitation over the three sub-regions. The annual cycle of precipitation shows that the model simulations with the different schemes are close to the observed data over the 3H and TP regions, except over NP, where no scheme produces the seasonal variation because the actual height of the Himalaya is more than prescribed in RegCM. The simulations are processed with 20 km horizontal resolution, and hence the problem of a leeward sudden change in orography exists. The model is unable to process the issues of orography, and thus some balancing parameters in the model are needed to overcome the simulation problems in this region. In the vertical profiles of temperature and water vapor mixing ratio, the atmosphere in the Tiedtke scheme is warmer and drier over the TP and NP compared to the other CPSs, which generates a precipitation difference. The Tiedtke scheme produces the dry conditions over the 3H region, and hence less precipitation due to less surface latent heat flux. This result is supported by the Tiedtke scheme's mechanism of moisture transportation and precipitation towards the northeast. For precipitation, the CPSs differ distinctly from one another. Among all the schemes, the Tiedtke scheme shows a strong correlation and standard deviation of precipitation and temperature. However, the correlation and standard deviation is very weak in NP in the results of all the schemes.

    Overall, the Tiedtke scheme yields better results for precipitation and temperature as compared to observed data, with the lowest RMSE values over the three sub-regions. Previous studies have recommended the Grell scheme for South Asia, the Emanuel scheme for East Asia, and both schemes (Grell and Emanuel) with equal weighting for Southeast Asia. This study agrees with previous results that the Grell scheme is best for South Asia; however, over most parts of China and the three sub-regions, our results are quite different from previous studies and suggest the Tiedtke scheme to be superior. We believe that the reason for this finding is the unique characteristics of deep convection and entrainment in the Tiedtke scheme, which reduces the wet bias of precipitation.

Reference

Catalog

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return