• Atkeson, T., , H. Greening, , and N. Poor, 2007: Bay region atmospheric chemistry experiment. Atmos. Environ., 41 , 41634164.

  • Betts, A. K., , and M. J. Miller, 1986: A new convective adjustment scheme. Part II: Single column tests using GATE wave, BOMEX, and arctic air-mass data sets. Quart. J. Roy. Meteor. Soc., 112 , 693709.

    • Search Google Scholar
    • Export Citation
  • Chen, F., , and J. Dudhia, 2001: Coupling an advanced land surface/hydrology model with the Penn State–NCAR MM5 modeling system. Part I: Model description and implementation. Mon. Wea. Rev., 129 , 569585.

    • Search Google Scholar
    • Export Citation
  • Conzemius, R. J., , and E. Fedorovich, 2006a: Dynamics of sheared convective boundary layer entrainment. Part I: Meteorological background and large-eddy simulations. J. Atmos. Sci., 63 , 11511178.

    • Search Google Scholar
    • Export Citation
  • Conzemius, R. J., , and E. Fedorovich, 2006b: Dynamics of sheared convective boundary layer entrainment. Part II: Evaluation of bulk model predictions of entrainment flux. J. Atmos. Sci., 63 , 11791199.

    • Search Google Scholar
    • Export Citation
  • Dudhia, J., 1989: Numerical study of convection observed during the winter monsoon experiment using a mesoscale two-dimensional model. J. Atmos. Sci., 46 , 30773107.

    • Search Google Scholar
    • Export Citation
  • Dudhia, J., , D. Gill, , K. Manning, , W. Wang, , C. Bruyere, , S. Kelly, , and K. Lackey, 2005: PSU/NCAR Mesoscale Modeling System tutorial class notes and user’s guide: MM5 Modeling System version 3. NCAR Mesoscale and Microscale Meteorology Division, 402 pp.

    • Search Google Scholar
    • Export Citation
  • El-Askary, H., , R. Gautam, , R. P. Singh, , and M. Kafatos, 2006: Dust storms detection over the Indo-Gangetic basin using multi sensor data. Adv. Space Res., 37 , 728733.

    • Search Google Scholar
    • Export Citation
  • Fang, J., , and R. Wu, 2005: The influence of the geostrophic wind advection approximation on a well-mixed layer. Bound.-Layer Meteor., 114 , 3152.

    • Search Google Scholar
    • Export Citation
  • Gopalakrishnan, S. G., , and R. Avissar, 2000: An LES study of the impacts of land surface heterogeneity on dispersion in the convective boundary layer. J. Atmos. Sci., 57 , 352371.

    • Search Google Scholar
    • Export Citation
  • Gopalakrishnan, S. G., , M. Sharan, , R. T. McNider, , and M. P. Singh, 1998: A study of turbulent and radiative processes in the stable boundary layer under weak wind conditions. J. Atmos. Sci., 55 , 954960.

    • Search Google Scholar
    • Export Citation
  • Goyal, P., , and Sidhartha, 2002: Effect of winds on SO2 and SPM concentrations in Delhi. Atmos. Environ., 36 , 29252930.

  • Grell, G. A., , S. E. Peckham, , R. Schmitz, , S. A. McKeen, , G. Frost, , W. C. Skamarock, , and B. Eder, 2005: Fully coupled “online” chemistry within the WRF model. Atmos. Environ., 39 , 69576975.

    • Search Google Scholar
    • Export Citation
  • Gupta, I., , and R. Kumar, 2006: Trends of particulate in four cities in India. Atmos. Environ., 40 , 25522566.

  • Hong, S-Y., , and H-L. Pan, 1996: Non-local boundary layer vertical diffusion in a medium-range forecast model. Mon. Wea. Rev., 124 , 23222339.

    • Search Google Scholar
    • Export Citation
  • Janjic, Z. I., 1990: The step-mountain coordinates: Physical package. Mon. Wea. Rev., 118 , 14291443.

  • Janjic, Z. I., 1994: The step-mountain eta coordinate model: Further developments of the convection, viscous sublayer, and turbulence closure schemes. Mon. Wea. Rev., 122 , 927945.

    • Search Google Scholar
    • Export Citation
  • Jiang, F., , T. Wang, , T. T. Wang, , M. Xie, , and H. Zhao, 2008: Numerical modeling of a continuous photochemical pollution episode in Hong Kong using WRF-chem. Atmos. Environ., 42 , 87178727.

    • Search Google Scholar
    • Export Citation
  • Kar, S. C., , and N. Ramanathan, 1989: A boundary layer model for the Andaman Islands. Proc. Natl. Acad. Sci. USA, 55A , 871885.

  • Kusaka, H., , F. Chen, , M. Tewari, , and H. Hirakuchi, 2005: Impact of using the urban canopy model on the simulation of the heat island. Extended Abstracts, WRF/MM5 Users’ Workshop, Boulder, CO, NCAR.

    • Search Google Scholar
    • Export Citation
  • Laprise, R., 1992: The Euler equations of motion with hydrostatic pressure as independent variable. Mon. Wea. Rev., 120 , 197207.

  • McNider, R. T., , and R. A. Pielke, 1981: Diurnal boundary-layer development over sloping terrain. J. Atmos. Sci., 38 , 21982212.

  • McNider, R. T., , and R. A. Pielke, 1984: Numerical simulation of slope and mountain flows. J. Climate Appl. Meteor., 23 , 14411453.

  • McNider, R. T., , M. D. Moran, , and R. A. Pielke, 1988: Influence of diurnal and inertial boundary layer oscillations on long-range dispersion. Atmos. Environ., 22 , 24452462.

    • Search Google Scholar
    • Export Citation
  • Mellor, G. L., , and T. Yamada, 1982: Development of a turbulence closure model for geophysical fluid problems. Rev. Geophys. Space Phys., 20 , 851875.

    • Search Google Scholar
    • Export Citation
  • Mitchell, K., 2005: The community Noah land-surface model. User’s Guide Public Release Version 2.7.1, Environmental Modeling Center, NOAA/NCEP, 26 pp.

    • Search Google Scholar
    • Export Citation
  • Mlawer, E. J., , S. J. Taubman, , P. D. Brown, , M. J. Iacono, , and S. A. Clough, 1997: Radiative transfer for heterogeneous atmosphere: RRTM, a validated correlated-k model for the long-wave. J. Geophys. Res., 102 , (D14). 1666316682.

    • Search Google Scholar
    • Export Citation
  • Ooyama, K. V., 1990: A thermodynamic foundation for modeling the moist atmosphere. J. Atmos. Sci., 47 , 25802593.

  • Pagowski, M., , J. Hacker, , and J. W. Bao, 2005: Behavior of WRF BL schemes and land surface models in 1d simulations during BAMEX. Extended Abstracts, WRF/MM5 Users’ Workshop, Boulder, CO, NCAR.

    • Search Google Scholar
    • Export Citation
  • Pandey, J. S., , R. Kumar, , and S. Devotta, 2005: Health risks of NO2, SPM and SO2 in Delhi (India). Atmos. Environ., 39 , 68686874.

  • Rama Krishna, T. V. B. P. S., , M. Sharan, , S. G. Gopalakrishnan, , and Aditi, 2003: Mean structure of the nocturnal boundary layer under strong and weak wind conditions: EPRI case study. J. Appl. Meteor., 42 , 952969.

    • Search Google Scholar
    • Export Citation
  • Rogers, E., , T. Black, , B. Ferrier, , Y. Lin, , D. Parrish, , and G. DiMego, 2001: Changes to the NCEP Meso Eta Analysis and Forecast System: Increase in resolution, new cloud microphysics, modified precipitation assimilation, modified 3DVAR analysis. NWS Tech. Procedures Bull., 24 pp. [Available online at http://www.emc.ncep.noaa.gov/mmb/mmbpll/eta12tpb/ and from National Weather Service, Office of Meteorology, 1325 East-West Highway, Silver Spring, MD 20910].

    • Search Google Scholar
    • Export Citation
  • Sharan, M., , R. T. McNider, , S. G. Gopalakrishnan, , and M. P. Singh, 1995: Bhopal gas leak: A numerical simulation of episodic dispersion. Atmos. Environ., 29 , 20612070.

    • Search Google Scholar
    • Export Citation
  • Sharan, M., , S. G. Gopalakrishnan, , R. T. McNider, , and M. P. Singh, 1996: Bhopal gas leak: A numerical investigation of the prevailing meteorological conditions. J. Appl. Meteor., 35 , 16371657.

    • Search Google Scholar
    • Export Citation
  • Sharan, M., , S. G. Gopalakrishnan, , R. T. McNider, , and M. P. Singh, 2000: Bhopal gas leak: A numerical investigation on the possible influence of urban effects on the prevailing meteorological conditions. Atmos. Environ., 34 , 539552.

    • Search Google Scholar
    • Export Citation
  • Singh, M. P., , R. T. McNider, , and J. T. Lin, 1993: An analytical study of diurnal wind-structure variation in the boundary layer and the low-level nocturnal jet. Bound.-Layer Meteor., 63 , 397423.

    • Search Google Scholar
    • Export Citation
  • Skamarock, W. C., , J. B. Klemp, , J. Dudhia, , D. O. Gill, , D. M. Barker, , W. Wang, , and J. G. Powers, 2005: A description of the advanced research WRF version 2. NCAR Tech. Note NCAR/TN-468+STR, 100 pp.

    • Search Google Scholar
    • Export Citation
  • Smirnova, T. G., , J. M. Brown, , and S. G. Benjamin, 1997: Performance of different soil model configurations in simulating ground surface temperature and surface fluxes. Mon. Wea. Rev., 125 , 18701884.

    • Search Google Scholar
    • Export Citation
  • Smirnova, T. G., , J. M. Brown, , S. G. Benjamin, , and D. Kim, 2000: Parameterization of cold-season processes in the MAPS land-surface scheme. J. Geophys. Res., 105 , (D3). 40774086.

    • Search Google Scholar
    • Export Citation
  • Stull, R., 1993: Review of transilient turbulence theory and non-local mixing. Bound.-Layer Meteor., 62 , 2196.

  • Stull, R., , and A. G. M. Driedonks, 1987: Applications of the transilient turbulence parameterization to atmospheric boundary-layer simulations. Bound.-Layer Meteor., 40 , 209239.

    • Search Google Scholar
    • Export Citation
  • Takemi, T., 2006: Impacts of moisture profile on the evolution and organization of midlatitude squall lines under various shear conditions. Atmos. Res., 82 , 3754.

    • Search Google Scholar
    • Export Citation
  • Wu, D., , X. Tie, , C. Li, , Z. Ying, , A. K. Lau, , J. Huang, , X. Deng, , and X. Bi, 2005: An extremely low visibility event over the Guangzhou region: A case study. Atmos. Environ., 39 , 65686577.

    • Search Google Scholar
    • Export Citation
  • Yadav, S., , M. S. Chauhan, , and A. Sharma, 2007: Characterisation of bio-aerosols during dust storm period in N-NW India. Atmos. Environ., 41 , 60636073.

    • Search Google Scholar
    • Export Citation
  • Zhang, Q., , C. Zhao, , X. Tie, , Q. Wei, , M. Huang, , G. Li, , Z. Ying, , and C. Li, 2006: Characterizations of aerosols over the Beijing region: A case study of aircraft measurements. Atmos. Environ., 40 , 45134527.

    • Search Google Scholar
    • Export Citation
  • Zilitinkevich, S. S., 1995: Non-local turbulent transport: Pollution dispersion aspects of coherent structure of convective flows. Air Pollution Theory and Simulation, H. Power, N. Moussiopoulos, and C. A. Brebbia, Eds., Vol. 1, Air Pollution III, Computational Mechanics Publications, 53–60.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    The selected domain over the northern Indian region. Here the shaded area with asterisks represents Himalayan region, the shaded area with crosses represents Indo-Gangetic plain, and the shaded area with plus signs represent the Thar Desert. 1: Gujarat, 2: Madhya Pradesh, 3: Uttar Pradesh, 4: Uttaranchal, 5: Himachal Pradesh, 6: Jammu and Kashmir, 7: Punjab, 8: Haryana, and 9: Rajasthan; AH: Ahmedabad, DE: Delhi, and JO: Jodhpur.

  • View in gallery

    Regional land surface characteristics: (a) topography, where the contours signify height in meters, (b) soil type, where the numerical values signify 19 USGS soil categories, (c) vegetation fraction during the summer case, (d) vegetation fraction during the winter case, (e) albedo during the summer case, and (f) albedo during the winter case.

  • View in gallery

    Diurnal variation of surface temperature during summer case over three northern Indian cities: (a) Delhi, (b) Ahmedabad, and (c) Jodhpur for the control simulation (thin solid line), idealized simulation (thin dashed line), YSU+Noah simulation (thick dashed line), MYJ+RUC simulation (thick solid line), and observations from IMD (filled circles).

  • View in gallery

    Diurnal variation of surface temperature during winter case over three northern Indian cities: (a) Delhi, (b) Ahmedabad, and (c) Jodhpur for the control simulation (thin solid line), idealized simulation (thin dashed line), and observations from IMD (filled circles).

  • View in gallery

    Diurnal variation of surface wind during summer case over three northern Indian cities: (a) wind speed over Delhi, (b) wind direction over Delhi, (c) wind speed over Ahmedabad, (d) wind direction over Ahmedabad, (e) wind speed over Jodhpur, and (f) wind direction over Jodhpur for the control simulation (thin solid line), idealized simulation (thin dashed line), YSU+Noah simulation (thick dashed line), MYJ+RUC simulation (thick solid line), and observations from IMD (filled circles).

  • View in gallery

    Diurnal variation of surface wind during winter case over three northern Indian cities: (a) wind speed over Delhi, (b) wind direction over Delhi, (c) wind speed over Ahmedabad, (d) wind direction over Ahmedabad, (e) wind speed over Jodhpur, and (f) wind direction over Jodhpur for the control simulation (thin solid line), idealized simulation (thin dashed line), and observations from IMD (filled circles).

  • View in gallery

    Vertical potential temperature profiles over three northern Indian stations during summer case: (a) Delhi at 0000 UTC 21 May, (b) Delhi at 1200 UTC 21 May, (c) Ahmedabad at 1200 UTC 20 May, (d) Ahmedabad at 0000 UTC 21 May, (e) Jodhpur at 1200 UTC 20 May, and (f) Jodhpur at 0000 UTC 21 May for the control simulation (thin solid line), idealized simulation (thin dashed line), YSU+Noah simulation (thick dashed line), MYJ+RUC simulation (thick solid line), and observations from Wyoming Weather Web data archive (filled circles).

  • View in gallery

    Vertical profiles of potential temperature over three northern Indian stations during winter case: (a) Delhi at 1200 UTC 9 Dec, (b) Delhi at 0000 UTC 10 Dec, (c) Ahmedabad at 1200 UTC 9 Dec, (d) Ahmedabad at 0000 UTC 10 Dec, (e) Jodhpur at 1200 UTC 9 Dec, and (f) Jodhpur at 0000 UTC 10 Dec for the control simulation (thin solid line), idealized simulation (thin dashed line), and observations from Wyoming Weather Web data archive (filled circles).

  • View in gallery

    Vertical profiles of wind speed over three northern Indian stations during summer case: (a) Delhi at 0000 UTC 21 May, (b) Delhi at 1200 UTC 21 May, (c) Ahmedabad at 1200 UTC 20 May, (d) Ahmedabad at 0000 UTC 21 May, (e) Jodhpur at 1200 UTC 20 May, and (f) Jodhpur at 0000 UTC 21 May for the control simulation (thin solid line), idealized simulation (thin dashed line), YSU+Noah simulation (thick dashed line), MYJ+RUC simulation (thick solid line), and observations from Wyoming Weather Web data archive (filled circles).

  • View in gallery

    As in Fig. 9 but for wind direction. The angles within the range 360°–450° lie in the first quadrant.

  • View in gallery

    As in Fig. 8 but for wind speed.

  • View in gallery

    As in Fig. 8 but for wind direction. The angles within the range 360°–450° lie in the first quadrant.

  • View in gallery

    Synoptic situations over Indian region from Indian daily weather map at 0300 UTC (a) 20 May 2005 and (b) 9 Dec 2004, and (c) symbols used in (a) and (b).

  • View in gallery

    Surface flow fields during control simulation in summer case at (a) 0600 UTC 20 May, (b) 1200 UTC 20 May, (c) 0600 UTC 21 May, and (d) 1200 UTC 21 May.

  • View in gallery

    Difference of surface control flow fields with respect to analysis during the summer case at (a) 0600 UTC 20 May, (b) 1200 UTC 20 May, (c) 0600 UTC 21 May, and (d) 1200 UTC 21 May.

  • View in gallery

    Surface flow fields during summer case in YSU+Noah simulation at (a) 0600 UTC 20 May and (b) 0600 UTC 21 May, and in MYJ+RUC simulation at (c) 0600 UTC 20 May and (d) 0600 UTC 21 May.

  • View in gallery

    Surface flow fields during the control simulation in winter case at (a) 0600 UTC 9 Dec, (b) 1200 UTC 9 Dec, (c) 1800 UTC 9 Dec, and (d) 0000 UTC 10 Dec.

  • View in gallery

    Difference of the surface control flow fields with respect to analysis during winter case at (a) 0600 UTC 9 Dec, (b) 1200 UTC 9 Dec, (c) 1800 UTC 9 Dec, and (d) 0000 UTC 10 Dec.

  • View in gallery

    Surface flow fields during the summer case in the idealized simulation at (a) 0600 UTC 21 May and (b) 1200 UTC 21 May, and the corresponding difference in surface flow fields with respect to the control simulation at (c) 0600 UTC 21 May and (d) 1200 UTC 21 May.

  • View in gallery

    Near-surface forward trajectories during summer case from (a) Jodhpur starting at 0000 UTC, (b) Jodhpur starting at 1200 UTC, (c) OA (27°N, 72°E) starting at 0000 UTC, (d) OA (27°N, 72°E) starting at 1200 UTC, (e) OB (28°N, 72.5°E) starting at 0000 UTC, (f) OB (28°N, 72.5°E) starting at 1200 UTC, (g) OC (29°N, 75°E) starting at 0000 UTC, and (h) OC (29°N, 75°E) starting at 1200 UTC simulation. Here, the filled square indicates origin of the dust and the filled diamond represents Delhi. The solid line represents the control simulation and dotted line represents idealized simulation.

  • View in gallery

    Near-surface forward trajectories during 26–28 May 2005 case from different points (same as that of Fig. 20) over the Thar region starting at (a) 0000 and (b) 1200 UTC 26 May.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 25 25 8
PDF Downloads 22 22 7

Study of Regional-Scale Boundary Layer Characteristics over Northern India with a Special Reference to the Role of the Thar Desert in Regional-Scale Transport

View More View Less
  • 1 Centre for Atmospheric Sciences, Indian Institute of Technology Delhi, New Delhi, India
  • | 2 Hurricane Research Division, NOAA/AOML/OAR/DOC, Miami, Florida
© Get Permissions
Full access

Abstract

Extensive contrasts of land surface heterogeneities have a pivotal role in modulating boundary layer processes and consequently, the regional-scale dispersion of air pollutants. The Weather Research and Forecasting (WRF) modeling system has been used to analyze the regional-scale boundary layer features over northern India. Two cases, 9–11 December 2004 and 20–22 May 2005, representing the winter and summer season, respectively, are chosen for the simulations. The model results have been compared with the observations from the India Meteorological Department (IMD) and Wyoming Weather Web data archive over three cities: Delhi, Ahmedabad, and Jodhpur. The simulations show that the thermal stratifications and the associated wind pattern are very well supported by land surface characteristics over the region. The results signify that the underlying land surface along with the prevailing hemispheric-scale meteorological processes (synoptic conditions) is the driver of the simulated patterns. The study implies that thermally driven regional circulations play a major role in the transport of particulate matter from the Thar Desert to Delhi and its neighboring regions during summer.

Corresponding author address: Prof. Maithili Sharan, Centre for Atmospheric Sciences, Indian Institute of Technology Delhi, Hauz Khas, New Delhi – 110 016, India. Email: mathilis@cas.iitd.ernet.in

Abstract

Extensive contrasts of land surface heterogeneities have a pivotal role in modulating boundary layer processes and consequently, the regional-scale dispersion of air pollutants. The Weather Research and Forecasting (WRF) modeling system has been used to analyze the regional-scale boundary layer features over northern India. Two cases, 9–11 December 2004 and 20–22 May 2005, representing the winter and summer season, respectively, are chosen for the simulations. The model results have been compared with the observations from the India Meteorological Department (IMD) and Wyoming Weather Web data archive over three cities: Delhi, Ahmedabad, and Jodhpur. The simulations show that the thermal stratifications and the associated wind pattern are very well supported by land surface characteristics over the region. The results signify that the underlying land surface along with the prevailing hemispheric-scale meteorological processes (synoptic conditions) is the driver of the simulated patterns. The study implies that thermally driven regional circulations play a major role in the transport of particulate matter from the Thar Desert to Delhi and its neighboring regions during summer.

Corresponding author address: Prof. Maithili Sharan, Centre for Atmospheric Sciences, Indian Institute of Technology Delhi, Hauz Khas, New Delhi – 110 016, India. Email: mathilis@cas.iitd.ernet.in

1. Introduction

Land surface can be heterogeneous over a wide range of spatial scales because of variability in vegetation, terrain, soil texture and wetness, surface roughness, and urbanization. Several studies have indicated that extensive contrasts of land surface created by surface heterogeneities at scales of tens of kilometers have a strong influence on the structure and behavior of the planetary boundary layer (PBL) or atmospheric boundary layer (ABL) and subsequently the dispersion of air pollutants (Gopalakrishnan and Avissar 2000). Factors such as turbulence, radiation, advection, and entrainment are significant in determining the structure and behavior of the PBL (Stull 1993; Gopalakrishnan et al. 1998; Conzemius and Fedorovich 2006a,b) over fairly homogeneous terrain. The evolution of the PBL is also influenced by the wind pattern aloft (Singh et al. 1993; Rama Krishna et al. 2003; Fang and Wu 2005). However, the structure of the PBL over heterogeneous terrain becomes more complex (McNider and Pielke 1981, 1984; Stull and Driedonks 1987; McNider et al. 1988; Sharan et al. 1996, 2000) as it additionally depends upon the topographical features. For instance, the studies of Kar and Ramanathan (1989) show the effect of land surface processes on the growth of PBL. Since the degree of complexity increases in the analysis of PBL processes over a heterogeneous surface, the studies for examining the regional-scale boundary layer features are limited.

The regional weather and climate of north India are strongly influenced by the Himalayas and the Thar Desert. The Himalayas act as a barrier to the cold north winds from central Asia, so that northern India is warm or only mildly cool during winter and hot during summer. December and January are typically the coldest months with mean temperatures of 10°–15°C. Highs in Delhi, India, range from 16° to 25°C and the nighttime temperatures range between 2° and 8°C. The hottest months for the northern Indian region are May and June with mean temperatures over 32°C and maximum temperatures over 40°C. Often during the summer, “hazelike” conditions are observed over the Indo-Gangetic basin (El-Askary et al. 2006), including the city of Delhi, and an unusually high concentration of particulate matter (PM) is encountered, especially during the daytime. Dust plays a vital role in regulating the health conditions of human beings (Yadav et al. 2007).

Since there are few observations available for flow fields over the northern Indian region, it is indeed difficult to illustrate the transport and regional-scale dispersion of dust simply based on observations. Consequently, to reconstruct the dispersion scenario over the Delhi region, an alternative approach is to use a high-resolution mesoscale model, as the model can simulate the physics even if there are not enough observations to generate high-resolution initial conditions. There have been some modeling efforts to evaluate the PM contents during such episodes (Pandey et al. 2005; Gupta and Kumar 2006) though none of them have provided a physical reasoning for the occurrence of such episodes. Goyal and Sidhartha (2002) have pointed out that the strong winds flowing from the Thar region are largely responsible for the increase of PM concentration over the Delhi region.

The local meteorological conditions over the Delhi region are usually influenced by the regional land surface and the flow fields over northern India. Thus, it is necessary to use a model with advanced land surface, physics, and dynamical schemes in the study of the PBL processes over the northern Indian region. Recently, a multi-institutional effort has been made to develop a mesoscale model, namely the Weather Research and Forecasting (WRF) modeling system, by incorporating advanced dynamics, physics, and numerical schemes (Skamarock et al. 2005) like that of the fifth-generation Pennsylvania State University–National Center for Atmospheric Research (NCAR) Mesoscale Model (MM5; Dudhia et al. 2005). The WRF modeling system has been used in several studies based on local and mesoscale events (Kusaka et al. 2005; Pagowski et al. 2005; Wu et al. 2005; Takemi 2006). Zhang et al. (2006) have coupled the WRF modeling system with a tracer transport model to study the aerosol size distributions within and above the ABL by analyzing the aircraft measurements over the Beijing region in China. Recently, WRF-Chem, the combination of the WRF model with chemistry modules (Grell et al. 2005), has been used to study a photochemical pollution episode over the Hong Kong region in China where the concentrations of pollutants such as ozone, nitrogen dioxide, and PM are considerably high (Jiang et al. 2008).

The current study is an attempt to analyze the regional-scale boundary layer features over the heterogeneous terrain of northern India and to examine the possible role of the Thar Desert in transporting dust to the Delhi region using the WRF modeling system. The numerical model and experimental design are described in section 2. The results obtained from the numerical simulations are discussed in section 3. Conclusions drawn from this study are given in section 4.

2. Numerical model and experimental design

a. Numerical model

The present study uses the Advanced Research WRF (ARW) modeling system, version 2.1, to analyze some of the regional-scale boundary layer features over the northern Indian region. Though the detailed description of the model is available in Skamarock et al. (2005), some of the model features have been described here for the sake of completeness.

The ARW modeling system is basically dependent upon a set of governing equations (Skamarock et al. 2005) written in flux form on an Eulerian solver for conservation of mass, momentum, energy, and scalars such as moisture (Ooyama 1990). These equations use mass-based vertical coordinate (Laprise 1992) η = (pdhpdht)/μd, where μd is the mass of dry air, pdh is the hydrostatic pressure of the dry atmosphere, and pdht is the corresponding value at the top; η varies from 1 at the surface to 0 at the upper boundary of the model domain. The prognostic variables in the equations are μd, velocity components u (zonal wind component), υ (meridional wind component), and w (vertical wind component), potential temperature (θ), and geopotential (φ). The nonconserved variables such as temperature (T), pressure (p), and density (ρ) are diagnosed from the conserved prognostic variables. The governing equations are solved numerically using the physically relevant initial and boundary conditions. The present study uses the initial and boundary conditions from the National Centers for Environmental Prediction (NCEP) final analysis data (available online at http://dss.ucar.edu/datasets/ds083.2/data/).

The simulation of a weather event by a model requires a minimum set of physics options, namely radiation, surface layer, boundary layer, land surface, convection, and microphysics schemes. The relevant physical parameterizations considered in the current work (Table 1) are briefly described here.

The Noah land surface model (LSM) has been adopted for the land surface parameterization (Mitchell 2005). This is a four-layer soil temperature and moisture model with canopy moisture and snow cover prediction (Chen and Dudhia 2001). It takes into account vegetation categories, monthly vegetation fraction, and soil texture. An experiment is also carried out with Rapid Update Cycle (RUC) LSM (Smirnova et al. 1997, 2000). In contrast to Noah LSM, RUC LSM uses six subsoil layers.

The schemes used in the present study for incorporating the effects of longwave and shortwave radiations are (i) Rapid Radiative Transfer Model (RRTM) for longwave radiation (Mlawer et al. 1997) and (ii) shortwave radiation parameterization from Dudhia (1989). These schemes are responsible for providing the atmospheric heating due to radiative flux divergence and surface downward longwave and shortwave radiations for the ground heat budget.

The present work adopts the Mellor–Yamada–Janjic (MYJ) scheme (Janjic 1990) as well as the Yonsei University (YSU) scheme (Hong and Pan 1996) for turbulence parameterization in the PBL. The MYJ PBL scheme represents a nonsingular implementation of turbulence closure (Mellor and Yamada 1982) throughout the full range of atmospheric turbulent regimes. YSU scheme is based upon local gradients of wind and potential temperature to deal with the turbulent regime of the whole atmospheric column. The YSU scheme also has a countergradient term to represent the nonlocal unstable boundary layer fluxes. In addition, it considers an explicit treatment of entrainment. The surface layer schemes used along with these PBL schemes are based upon the Monin–Obukhov approach. However, the surface layer treatment with MYJ PBL includes the parameterization of the viscous sublayer over land and water (Janjic 1994) with variable roughness height for temperature and humidity (Zilitinkevich 1995).

The schemes used for microphysics and cumulus convection while simulating both summer and winter cases are the Ferrier scheme (Rogers et al. 2001) and Betts–Miller–Janjic scheme (Betts and Miller 1986; Janjic 1994), respectively.

b. Model domain and experiment design

To analyze the regional-scale boundary layer characteristics over northern India, the selected domain (Fig. 1) includes the western parts of the Thar Desert, a part of the Indian Ocean, and the Himalayan region as well. The number of grid points in the domain is 126 (both in x and y directions). The latitude–longitude of the central point is specified as 27°N, 76°E to generate the grids within the domain while using the Mercator map projection. The resolution of the domain is 15 km. The vertical resolution includes 31 levels from the surface in terrain-following vertical coordinates (η) and the top pressure is 250 hPa (the layers above it are not expected to influence the boundary layer features).

Six-hourly NCEP–NCAR final analysis data with 1° × 1° resolution are used in the model for initial and lateral boundary conditions. Topographical features (with a resolution of 30 s), vegetation, soil (with a resolution of 30 s), soil temperature (with a resolution of 1°), albedo, and land use (with a resolution of 30 s) are from the U.S. Geological Survey (USGS) global data. The soil is sandy and dry in the western part of Rajasthan (Fig. 1) and the soil moisture content is very low in the region (Fig. 2b). In the southern side of the domain, black soil having relatively higher moisture content is present (Fig. 2b). The Indo-Gangetic plain (Fig. 1) has alluvial soil (silt type) with comparatively higher moisture content than the western Rajasthan (Fig. 2b). The extent of vegetation is less over the Thar Desert region as compared with the Indo-Gangetic plain during both winter and summer (Figs. 2c,d). More dryness prevails over the Thar Desert region as compared with the Indo-Gangetic plain during summer relative to winter because of differences in land surface characteristics such as albedo (Figs. 2e,f) and vegetation fraction (Figs. 2c,d).

Two case studies representing the winter (9–11 December 2004) and summer (20–22 May 2005) seasons are primarily considered for the simulations. The winter days are accompanied by prevailing low to moderate winds over the region. However, the summer days are accompanied by relatively strong and variable winds over the region. The criteria for selecting these days for the simulation are (i) clear sky and fair weather conditions with negligible precipitation, (ii) high temperatures (≥∼40°C) over all of the selected cities during the summer case, and (iii) low-to-moderate temperature conditions (0° to 25°C) during the winter case associated with low to moderate winds at least over Delhi and its surrounding region. In each case, the model is initialized at 0000 UTC (coordinated universal time) or 0530 Indian standard time (IST) and the simulations are carried out for 48 h each.

In each case study, the simulation with the parameterization schemes MYJ PBL and Noah LSM is referred to as the control simulation. In addition, the following sensitivity experiments for the summer case are carried out: (i) considering the parameterizations YSU PBL and Noah LSM, referred to as YSU+Noah, (ii) taking the parameterizations MYJ PBL and RUC LSM, referred to as MYJ+RUC, and (iii) an idealized simulation considering the domain with barren land having a uniform background albedo of 0.25 and roughness length of 1 cm. The idealized simulation is carried out for both summer and winter cases.

3. Results and discussion

The simulated surface temperature and wind over Delhi, Ahmedabad, and Jodhpur are compared with the available observations from the India Meteorological Department (IMD). The vertical profiles of potential temperature and wind over these three cities are also compared with the observations from the Wyoming Weather Web data archive. The near-surface flow fields are analyzed to study the regional-scale features over the northern Indian region.

In this section, the results from simulations with 15-km resolution are given since the results from the simulations with 12-km resolution do not show significant improvement in the regional-scale boundary layer characteristics even though an attempt was made in this direction. If the simulations are carried out at further higher resolutions (e.g., at 3 km), the study is expected to show a more detailed description of mesoscale circulation features as compared with that seen in the simulations at 15- or 12-km resolutions. However, the more detailed features shown by fine-resolution models are not necessarily better unless one can verify them. Therefore, the results presented here should be viewed with regard to other operational models having approximately the same resolution as taken in the present study.

a. Surface temperature and wind

The simulated surface temperatures (Figs. 3 and 4) and wind (Figs. 5 and 6) over the cities Delhi, Ahmedabad, and Jodhpur are compared with those observed in the summer and winter cases. The surface temperatures computed from the model are comparable to those observed in both the cases (Table 2; Figs. 3, 4). However, the surface temperature over Delhi is slightly overpredicted by the control simulation during the nocturnal conditions in the winter case (Fig. 4a). The winter is usually associated with relatively low wind and strong stability. Thus, the warm bias shown by the model can be due to the strong stability and low-wind conditions (Sharan et al. 1995) prevailing over Delhi during the winter season. On the other hand, during the summer case, the computed surface temperatures from the control simulation are closer to the observations for all three cities.

The surface temperatures computed in the idealized simulation show almost similar diurnal variation as compared to those in the control simulation in both summer and winter cases (Table 2; Figs. 3, 4). However, the predicted surface temperatures from the idealized simulation vary slightly in magnitude from those of the control simulation most of the time because of the difference in underlying land surface. The MYJ+RUC sensitivity underpredicts the surface temperature as compared to the observations while the computed surface temperature from the YSU+Noah simulation behaves similar to that of the control simulation during the summer case.

The computed surface wind deviates from the observations in both summer (Fig. 5) and winter (Fig. 6) cases. The observations as well as simulations indicate that there is a large variation in the magnitude and direction of winds over Delhi during the summer case (Fig. 5a). For instance, the model computes fairly large values of wind speed over Delhi in the early morning hours as well as at night, though the observations show low wind speed (<2.0 m s−1) at these hours (Fig. 5a). However, Ahmedabad and Jodhpur show relatively less variation in wind speed and direction (Fig. 5). On the other hand, in the winter case, the wind speed is relatively smaller in magnitude and there is not much variation in wind direction (Fig. 6). In both cases, the model is not able to reproduce the observed wind speed and direction properly though the simulated values of wind speed and direction are relatively closer to the observations over Ahmedabad. Even though the parameterizations are changed while carrying out the sensitivity studies, the results did not improve in the summer case. Since the idealized simulation considers a barren land with smaller roughness length (i.e., 1 cm), it is expected that there will be relatively smaller frictional drag resulting in larger values of wind speed at the surface for both the summer and winter cases. Please note that the roughness length in control simulation ranges from 0.01 to 80 cm. Thus, the model with the idealized simulation predicts relatively larger values of surface winds.

The deviation of the results from observations while predicting surface temperature may be due to the (i) difference in number and depth of subsoil layer considered in both the land surface schemes and (ii) specification of bottom soil moisture and temperature. Similarly, the deviation of results from surface wind observations could be because (i) the synoptically considered initial wind fields do not correspond to the locally observed values and thus provide a scope for improvement in the initial conditions, (ii) the land surface features over the domain are not properly represented, and (iii) the model resolution is still too coarse to predict the wind properly.

b. Potential temperature profiles

The vertical potential temperature profiles are relatively closer to observations in both the summer (Fig. 7) and winter (Fig. 8) cases over Delhi, Ahmedabad, and Jodhpur. The profiles at 1200 UTC show the persistence of a mixed layer within the PBL during the summer case (Figs. 7b,c,e) being consistent with the observations at these hours whereas a near-neutral condition is observed at the surface during the evening hours over all the cities in the winter case (Figs. 8a,c,e). The radiation inversion follows a similar trend as in the observations at 0000 UTC over all cities during both summer (Figs. 7a,d,f) and winter (Figs. 8b,d,f) cases.

Over Ahmedabad and Jodhpur, the computed potential temperatures from the control simulation are relatively closer to the observations as compared to that of the idealized simulation in the summer case (Figs. 7c–f). However in all other cases of summer (Figs. 7a,b) and winter (Fig. 8), no significant difference is observed between the results from the idealized and control simulations.

The model with the MYJ+RUC simulation does not improve the results in predicting the vertical potential temperature profiles within the convective boundary layer (Figs. 7b,c,e). However, in relatively stable conditions, the computed values of potential temperatures from the MYJ+RUC simulation are relatively closer to the observations than those from the control simulation (Figs. 7a,f) during the summer case. The computed potential temperatures from the YSU+Noah sensitivity experiment are relatively closer to the observations as compared to those of the control simulation as well as MYJ+Noah simulation in unstable conditions (Figs. 7b,c,e). Thus, the vertical profiles of potential temperature are sensitive to land surface and PBL parameterization schemes.

c. Wind profiles

The simulated vertical wind profiles (Figs. 9 –12) are less satisfactory compared to those for potential temperature though the observations and predictions show similar trends in most of the cases. However, the basic features (e.g., wind maximum) are captured only in a qualitative manner by the model in some cases (Figs. 9a,d, 11d). During strongly convective conditions, the model predicts mixed layer profiles within 1 km from the surface (Figs. 9b,c,e, 10b,c,e). In these cases, the computed wind with the MYJ+RUC simulation is relatively closer to the observations as compared to that of the control simulation and it performs better than the YSU+Noah simulation within the PBL. The computed values of wind speed and direction in the idealized simulation differ from the observed values as well as those in the control simulation during the summer and winter cases within the PBL and the model performs better in the control simulation as compared to the idealized simulation.

The discrepancy in the model prediction may be due to the parameterization used, effect of local/regional land surface features and the given initial conditions, whereas the observed wind profiles are instantaneous. On the other hand, the temperature profiles are relatively steady under a given synoptic condition at a particular time though the temperature (potential temperature) profiles in PBL will always evolve with time and hence the simulated temperatures (potential temperature) are found to be relatively closer to the observations.

d. Circulation features and regional transport

The monsoon is the principal component of the Indian climate system. The southwest monsoon plays a vital role in governing the weather conditions in the region during June–August. However, a highly convective condition persists during the premonsoon months (May–June) over the northern Indian region because of the presence of a low pressure trough over the Indo-Gangetic plain. A typical summertime synoptic situation has been depicted in Fig. 13a from the Indian Daily Weather Report (IDWR) for 20 May 2005. Because of the existence of a low pressure trough over the Indo-Gangetic plain, the wind flows from the relatively high pressure area of the Arabian Sea toward this region passing over the Thar Desert. Thus, there is a possibility for convergence to occur over the Indo-Gangetic plain when it is counteracted by the flows from the Himalayan side. This is further analyzed from the results of the ARW modeling system as well. The mesoscale circulations are not well captured in the fields from the global models. While WRF will spin up these circulations, it takes awhile.

The southwest monsoon withdraws from the northern parts of the country by September so that a high pressure area is established over the region. This becomes most intense during December–January (Fig. 13b; when the winter season exists in this region). In such a situation, a ridgelike condition over the Indo-Gangetic plain is created because of the presence of high pressure. The surface wind and the wind aloft within a kilometer of the surface are usually weak to moderate. As a result, the wind flow creates a divergence at the surface or within 1 km from the surface and often the vertical stratification is marked by stable condition, which makes the environment unfavorable for convective activity. Such a typical winter case has been depicted in Fig. 13b from IDWR for 9 December 2004.

To look into the regional-scale circulation features within the ABL over the northern Indian region, the 10-m flow fields are analyzed here. For instance, Fig. 14 shows the regional circulation patterns at the surface from the control simulation during the summer case in 6-h intervals within the convective boundary layer (CBL) for the first (Figs. 14a,b) and second (Figs. 14c,d) days. The model shows the flow of wind from the Thar Desert toward the Indo-Gangetic plain (Figs. 14a,b) that continues throughout the day and even until the next morning (Fig. 14). The flow is counteracted by an opposing “alongslope” southeasterly flow at the foothills of the Himalayas creating a convergence zone around Delhi and its surrounding region over the Indo-Gangetic plain (Fig. 14a), where the wind becomes weak and variable. The flow along the foothills is better resolved by the model as compared with that of the global analysis (Figs. 15a,b). Consequently, the center of convergence is located over the Delhi region (Fig. 14a). The model with the resolution of 15 km simulates the dynamic convergence (Figs. 14a,b), which is not brought out well by the coarse-resolution global analysis dataset, as evident from the difference fields (Figs. 15a,b). Note that the dynamic convergence occurs initially because of the prevailing synoptic condition (Fig. 13a). Any PM transported after mixing in the CBL of the Thar Desert (Figs. 7e, 9e, 10e) on 20 May may linger over Delhi and its surroundings on the following day (21 May) and may not be transported beyond the line of convergence (Fig. 14a). It should be emphasized that a well-defined circulation sets up around Delhi on 21 May (Fig. 14c), which is captured well by the model but could not be captured by the coarse-resolution global analysis (Fig. 15c).

Early on 21 May, the flow changes its course (i.e., the flow over Thar Desert was either easterly or northwesterly), resulting in the observed culmination of the hazelike episode over Delhi. The results are further supported by the sensitivity experiments considering the YSU PBL scheme and RUC LSM (Fig. 16). Both of the sensitivity experiments such as the YSU+Noah and MYJ+RUC simulations signify the isolation of dynamic convergence over the Indo-Gangetic plain (Figs. 16a,c). However, changing the PBL parameterization from the MYJ to YSU scheme does not improve the isolation of the dynamical convergence and the regional flow fields (Figs. 16a,b) during the summer case. In contrast, the change in land surface parameterization to RUC LSM makes the flow fields stronger as compared with the control simulation over the Indo-Gangetic plain at the foothills of the Himalayas (Figs. 16c,d). Thus, the computed results are sensitive to land surface and boundary layer parameterization schemes within the ABL at the vicinity of the earth’s surface.

The difference fields of the control simulation with respect to global analysis show that the regional flow patterns produced by better-resolved heterogeneous land surface are more illustrative within a CBL that occurs on a typical summer day in northern India (Fig. 15). For the winter case depicted in Fig. 17, there is a persistent low to moderate wind prevailing over the Thar Desert region throughout the day. The difference of control flow fields from the analysis (Fig. 18) signifies that the model indeed has a role in illustrating the flow pattern at the surface even during a relatively stable situation in winter. The difference observed during both summer (Fig. 15) as well as winter (Fig. 18) is due to the role of better-resolved, underlying land surface and model physics. This has been further discussed in the next subsection.

In the coastal areas, wind is observed to be flowing from the ocean side during the summer case (Fig. 14) whereas the opposite type of flow is observed during the winter case (Fig. 17). The results also show variable winds in other places within the domain. A very careful examination indeed shows small pockets of local effects. Nevertheless, the prevailing relatively stronger stability on a winter day (Fig. 8) inhibits the stronger local-scale flow and vertical motions. Consequently, land surface effects may not have a major influence on the regional-scale transport in this case.

Sharan et al. (1996) have shown that at about 5 km above ground level, the prevailing pattern is westerlies north of about 28° latitude, and easterlies, south of 20°N latitude, which are highly variable in the 8° zone during winter. The model-simulated northwesterly flow across the Indo-Gangetic plain, for instance, is consistent with the westerlies aloft.

e. Role of underlying land surface in regional-scale flow development

The differential land surface features (Fig. 2), especially albedo and soil texture, have significant influence on regional-scale circulation. The control simulation isolates the regional flow fields better as compared to the global analysis both in summer (Fig. 14) and winter (Fig. 15) even though the simulated diurnal surface wind speed and direction do not agree very well with observations (Figs. 5, 6). This could be due to the parameterization schemes, the resolution of the model, and/or underlying land surface features as mentioned earlier. Gopalakrishnan and Avissar (2000) have performed some large eddy simulations of the CBL and found that the horizontal pressure gradient created by idealized land surface heterogeneities on the order of about 10 km has a remarkable influence on the CBL and the subsequent mesoscale transport of passive pollutants. The current findings exemplify those consequences. To look into the impact of the underlying land surface, an additional idealized simulation is carried out. In the idealized simulation with barren land, the flow is expected to be stronger than that of the control simulation because of relatively less frictional drag. For instance, the comparison of Figs. 14c,d and 19a,b implies that the flow fields in the summer case on 21 May are relatively stronger in the idealized simulation as compared to the control simulation over the Indo-Gangetic plain (Figs. 19c,d) while creating a regional circulation over Delhi and its neighboring region at 0600 UTC and the northwesterly flow at 1200 UTC. To reconstruct the dispersion scenario over Delhi and its neighboring regions, the type of flow patterns assumes significance and consequently, the underlying land surface characteristics (especially roughness) influence the flow features. Thus, the transport of dust from the Thar Desert to Delhi and its neighboring regions during summer may be regulated by the underlying land surface characteristics.

f. Trajectory of the flow during regional-scale transport

The possible role of the Thar region in the transport of dust particles toward Delhi in the CBL is further evident from the trajectories drawn from different places of the Thar Desert for the summer case (Fig. 20). These trajectories are drawn using horizontal wind components at the 10-m level. A rectangular region 25°–30°N, 72°–80°E including the Thar Desert and Delhi is considered so the trajectories can be visualized clearly. The trajectories from OA (27°N, 72°E) in both the control and idealized simulations extend south of the Delhi region (Fig. 20c). Similarities are observed in the case of the trajectory from Jodhpur in both the control and idealized simulations. The trajectory drawn from OB (28°N, 72.5°E) over the Thar Desert in the idealized simulation just reaches Delhi (28.32°N, 77.2°E), whereas the corresponding trajectory in the control simulation extends northeast of Delhi (Fig. 20e). The trajectories from OC (29°N, 75°E) extend east of Delhi (Fig. 20g). The trajectories starting at 1200 UTC 20 May from different places in the Thar Desert such as Jodhpur (Fig. 20b), OA (Fig. 20d), and OB (Fig. 20f) imply that the dust particles would not be able to be transported close to the vicinity of Delhi if the transport started at this hour. However, the dust can still reach east of Delhi even if the transport started at 1200 UTC from OC (Fig. 20h). Thus, the forward trajectories shown from the various points in the Thar region reveal that the particulates tend to transport from the Thar region toward the vicinity of Delhi if the transportation starts at 0000 UTC 20 May. However, the comparison also implies that the transport of dust toward Delhi and the areas on its northern side is possible from the northern areas of the Thar Desert in a particular synoptic condition (Fig. 13a). The comparison of trajectories from the control and idealized simulations shows that there is some signature of the impact of underlying land surface on the transport of Thar dust to form hazelike conditions over Delhi and its neighboring regions. This supports our conclusion regarding the role of underlying land surface in modulating the flow patterns, as illustrated from the difference flow fields of idealized and control simulations (Figs. 19c,d).

To further support the possible role of the Thar region in the transport of dust particles toward Delhi and its neighboring region in the convective conditions of summer, an additional simulation is carried out for 48 h by using the same set of physics options as that of the control simulation and initializing the model at 0000 UTC 26 May 2005. The characteristic features obtained in this case are almost similar to those reported for the 20–22 May 2005 case. The forward trajectories shown from the various points in the Thar region (Fig. 21) reveal that the particulates tend to transport from the Thar region toward the vicinity of Delhi. This result supports our conclusions about the role of Thar dust in the formation of hazelike conditions over Delhi and its neighboring regions.

4. Concluding remarks

An analysis of regional-scale flow and PBL characteristics over the northern Indian region is carried out using the ARW modeling system. This model uses 6-hourly NCEP–NCAR final analysis data with 1° × 1° resolution as input and USGS data for the information regarding the underlying land surface. The model is configured for a domain with a resolution of 15 km over the northern Indian region. The simulations are carried out for 48 h by initializing the model at 0000 UTC (0530 h IST) in each case for winter (9–11 December 2004) and summer (20–22 and 26–28 May 2005). The computed temperature and wind profiles are compared with the observations from IMD and the Wyoming Weather Web data archive. The results from the present study signify that the large-scale topography and underlying land surface along with hemispheric-scale meteorological processes or synoptic conditions are the drivers for the observed (and simulated) patterns. Nevertheless, the simulated results of regional-scale flows and the behavior of diurnal variations of surface wind speed over the three cities might be improved with better initial and boundary conditions. The significant conclusions drawn from this study are given as follows:

  • (i) The simulated surface temperatures over the cities of Delhi, Ahmedabad, and Jodhpur are comparable to the observations in both summer as well as winter cases with the exception of Delhi during winter case. The overprediction of surface temperature in the winter case over Delhi implies that the model shows warm bias and this situation can be due to the prevailing stable and low-wind conditions, which are not well captured by the model. The model is able to capture the clear sky potential temperature stratification within the PBL, both in summer and winter cases, in a reasonable manner over the northern Indian cities of Delhi, Ahmedabad, and Jodhpur. Such a feature is expected as the PBL schemes are well tuned for clear sky weather conditions under which PBL potential temperature structures are largely controlled by local land surface heating/cooling forces.
  • (ii) There is a large variation in wind speed and direction over Delhi as compared to other cities during the summer case and none of the simulations could predict the surface wind over Delhi and Jodhpur properly. However, a relatively better agreement is seen for Ahmedabad. Similarly, the basic features in the vertical profiles of wind during both cases have been captured only in a qualitative manner and the model-produced winds aloft do not agree well with the observations.
  • (iii) The prevailing stronger stability and relatively low winds in the winter case do not support the local/regional-scale flow significantly so as to be responsible for the regional-scale transport. Thus, a distinct regional-scale feature favorable for regional-scale transport over the northern Indian region resulting from the differential regional land surface variations could not be isolated in the winter case. However, because of the strongly convective condition prevailing over northern Indian during summer, the local/regional-scale flow and vertical motions of air masses are very well supported. Consequently, the distinct regional-scale circulation resulting initially from the synoptic conditions is further influenced by differential land surface over the region so as to be responsible for increasing the PM concentration over Delhi and its neighboring region during summer. This condition arises as the PM dispelled from the surface after mixing in the CBL over the Thar Desert is transported toward the Indo-Gangetic plain by the regional thermal flow field created over the northern Indian region, resulting in a hazelike condition over Delhi, which diminishes as the flow changes its course with time. Thus, the thermally driven regional wind systems play a major role in the transport of PM from the Thar region to Delhi region.

Dust storm and haze events are common for the northern Indian cities during the summer season. The current study is a preliminary step in this subject. In this study, the better-resolved land surface and the type of parameterization schemes seem to have a role in regulating the mesoscale flow patterns. However, the regional-scale features are largely regulated by synoptic conditions at a resolution of 15 km as discussed above. If the study is extended to the meso-γ-scale or microscale level, the effect of better-resolved land surface and parameterization schemes may be more prevalent. However, further studies are needed to increase the understanding of the role of the regional-scale transport resulting from heterogeneous land surfaces. Although the model gives satisfactory results in the current framework, some of the shortcomings like resolution of the model and lack of extensive observational data cannot be avoided. Thus, a field campaign like that of the Bay Region Atmospheric Chemistry Experiment (Atkeson et al. 2007) needs to be designed to have an extensive observational network of (i) concentration measurements of PM, (ii) surface layer and upper-air observations, and (iii) turbulence measurements. Nevertheless, future studies will take care of some of these aspects to gain further insight into the regional-scale boundary layer characteristics over the northern Indian region.

Acknowledgments

We gratefully acknowledge the anonymous reviewers for their valuable comments and suggestions to improve the paper in all directions. The authors thank John Michalakes and the WRF help team (NCAR), Paul D. Smith (GNU make project), Dr. Mukul Tewari (NCAR), Dr. J. Dudhia (NCAR), Dr. Melissa P. Free (NOAA/ARL), Dr. S. C. Kar (NCMRWF), and Dr. Kanti Prasad (IIT Delhi) for their valuable suggestions and help. We also thank IMD (New Delhi), Dr. Larry Oolman, and the Wyoming Weather Web for their observational data support.

REFERENCES

  • Atkeson, T., , H. Greening, , and N. Poor, 2007: Bay region atmospheric chemistry experiment. Atmos. Environ., 41 , 41634164.

  • Betts, A. K., , and M. J. Miller, 1986: A new convective adjustment scheme. Part II: Single column tests using GATE wave, BOMEX, and arctic air-mass data sets. Quart. J. Roy. Meteor. Soc., 112 , 693709.

    • Search Google Scholar
    • Export Citation
  • Chen, F., , and J. Dudhia, 2001: Coupling an advanced land surface/hydrology model with the Penn State–NCAR MM5 modeling system. Part I: Model description and implementation. Mon. Wea. Rev., 129 , 569585.

    • Search Google Scholar
    • Export Citation
  • Conzemius, R. J., , and E. Fedorovich, 2006a: Dynamics of sheared convective boundary layer entrainment. Part I: Meteorological background and large-eddy simulations. J. Atmos. Sci., 63 , 11511178.

    • Search Google Scholar
    • Export Citation
  • Conzemius, R. J., , and E. Fedorovich, 2006b: Dynamics of sheared convective boundary layer entrainment. Part II: Evaluation of bulk model predictions of entrainment flux. J. Atmos. Sci., 63 , 11791199.

    • Search Google Scholar
    • Export Citation
  • Dudhia, J., 1989: Numerical study of convection observed during the winter monsoon experiment using a mesoscale two-dimensional model. J. Atmos. Sci., 46 , 30773107.

    • Search Google Scholar
    • Export Citation
  • Dudhia, J., , D. Gill, , K. Manning, , W. Wang, , C. Bruyere, , S. Kelly, , and K. Lackey, 2005: PSU/NCAR Mesoscale Modeling System tutorial class notes and user’s guide: MM5 Modeling System version 3. NCAR Mesoscale and Microscale Meteorology Division, 402 pp.

    • Search Google Scholar
    • Export Citation
  • El-Askary, H., , R. Gautam, , R. P. Singh, , and M. Kafatos, 2006: Dust storms detection over the Indo-Gangetic basin using multi sensor data. Adv. Space Res., 37 , 728733.

    • Search Google Scholar
    • Export Citation
  • Fang, J., , and R. Wu, 2005: The influence of the geostrophic wind advection approximation on a well-mixed layer. Bound.-Layer Meteor., 114 , 3152.

    • Search Google Scholar
    • Export Citation
  • Gopalakrishnan, S. G., , and R. Avissar, 2000: An LES study of the impacts of land surface heterogeneity on dispersion in the convective boundary layer. J. Atmos. Sci., 57 , 352371.

    • Search Google Scholar
    • Export Citation
  • Gopalakrishnan, S. G., , M. Sharan, , R. T. McNider, , and M. P. Singh, 1998: A study of turbulent and radiative processes in the stable boundary layer under weak wind conditions. J. Atmos. Sci., 55 , 954960.

    • Search Google Scholar
    • Export Citation
  • Goyal, P., , and Sidhartha, 2002: Effect of winds on SO2 and SPM concentrations in Delhi. Atmos. Environ., 36 , 29252930.

  • Grell, G. A., , S. E. Peckham, , R. Schmitz, , S. A. McKeen, , G. Frost, , W. C. Skamarock, , and B. Eder, 2005: Fully coupled “online” chemistry within the WRF model. Atmos. Environ., 39 , 69576975.

    • Search Google Scholar
    • Export Citation
  • Gupta, I., , and R. Kumar, 2006: Trends of particulate in four cities in India. Atmos. Environ., 40 , 25522566.

  • Hong, S-Y., , and H-L. Pan, 1996: Non-local boundary layer vertical diffusion in a medium-range forecast model. Mon. Wea. Rev., 124 , 23222339.

    • Search Google Scholar
    • Export Citation
  • Janjic, Z. I., 1990: The step-mountain coordinates: Physical package. Mon. Wea. Rev., 118 , 14291443.

  • Janjic, Z. I., 1994: The step-mountain eta coordinate model: Further developments of the convection, viscous sublayer, and turbulence closure schemes. Mon. Wea. Rev., 122 , 927945.

    • Search Google Scholar
    • Export Citation
  • Jiang, F., , T. Wang, , T. T. Wang, , M. Xie, , and H. Zhao, 2008: Numerical modeling of a continuous photochemical pollution episode in Hong Kong using WRF-chem. Atmos. Environ., 42 , 87178727.

    • Search Google Scholar
    • Export Citation
  • Kar, S. C., , and N. Ramanathan, 1989: A boundary layer model for the Andaman Islands. Proc. Natl. Acad. Sci. USA, 55A , 871885.

  • Kusaka, H., , F. Chen, , M. Tewari, , and H. Hirakuchi, 2005: Impact of using the urban canopy model on the simulation of the heat island. Extended Abstracts, WRF/MM5 Users’ Workshop, Boulder, CO, NCAR.

    • Search Google Scholar
    • Export Citation
  • Laprise, R., 1992: The Euler equations of motion with hydrostatic pressure as independent variable. Mon. Wea. Rev., 120 , 197207.

  • McNider, R. T., , and R. A. Pielke, 1981: Diurnal boundary-layer development over sloping terrain. J. Atmos. Sci., 38 , 21982212.

  • McNider, R. T., , and R. A. Pielke, 1984: Numerical simulation of slope and mountain flows. J. Climate Appl. Meteor., 23 , 14411453.

  • McNider, R. T., , M. D. Moran, , and R. A. Pielke, 1988: Influence of diurnal and inertial boundary layer oscillations on long-range dispersion. Atmos. Environ., 22 , 24452462.

    • Search Google Scholar
    • Export Citation
  • Mellor, G. L., , and T. Yamada, 1982: Development of a turbulence closure model for geophysical fluid problems. Rev. Geophys. Space Phys., 20 , 851875.

    • Search Google Scholar
    • Export Citation
  • Mitchell, K., 2005: The community Noah land-surface model. User’s Guide Public Release Version 2.7.1, Environmental Modeling Center, NOAA/NCEP, 26 pp.

    • Search Google Scholar
    • Export Citation
  • Mlawer, E. J., , S. J. Taubman, , P. D. Brown, , M. J. Iacono, , and S. A. Clough, 1997: Radiative transfer for heterogeneous atmosphere: RRTM, a validated correlated-k model for the long-wave. J. Geophys. Res., 102 , (D14). 1666316682.

    • Search Google Scholar
    • Export Citation
  • Ooyama, K. V., 1990: A thermodynamic foundation for modeling the moist atmosphere. J. Atmos. Sci., 47 , 25802593.

  • Pagowski, M., , J. Hacker, , and J. W. Bao, 2005: Behavior of WRF BL schemes and land surface models in 1d simulations during BAMEX. Extended Abstracts, WRF/MM5 Users’ Workshop, Boulder, CO, NCAR.

    • Search Google Scholar
    • Export Citation
  • Pandey, J. S., , R. Kumar, , and S. Devotta, 2005: Health risks of NO2, SPM and SO2 in Delhi (India). Atmos. Environ., 39 , 68686874.

  • Rama Krishna, T. V. B. P. S., , M. Sharan, , S. G. Gopalakrishnan, , and Aditi, 2003: Mean structure of the nocturnal boundary layer under strong and weak wind conditions: EPRI case study. J. Appl. Meteor., 42 , 952969.

    • Search Google Scholar
    • Export Citation
  • Rogers, E., , T. Black, , B. Ferrier, , Y. Lin, , D. Parrish, , and G. DiMego, 2001: Changes to the NCEP Meso Eta Analysis and Forecast System: Increase in resolution, new cloud microphysics, modified precipitation assimilation, modified 3DVAR analysis. NWS Tech. Procedures Bull., 24 pp. [Available online at http://www.emc.ncep.noaa.gov/mmb/mmbpll/eta12tpb/ and from National Weather Service, Office of Meteorology, 1325 East-West Highway, Silver Spring, MD 20910].

    • Search Google Scholar
    • Export Citation
  • Sharan, M., , R. T. McNider, , S. G. Gopalakrishnan, , and M. P. Singh, 1995: Bhopal gas leak: A numerical simulation of episodic dispersion. Atmos. Environ., 29 , 20612070.

    • Search Google Scholar
    • Export Citation
  • Sharan, M., , S. G. Gopalakrishnan, , R. T. McNider, , and M. P. Singh, 1996: Bhopal gas leak: A numerical investigation of the prevailing meteorological conditions. J. Appl. Meteor., 35 , 16371657.

    • Search Google Scholar
    • Export Citation
  • Sharan, M., , S. G. Gopalakrishnan, , R. T. McNider, , and M. P. Singh, 2000: Bhopal gas leak: A numerical investigation on the possible influence of urban effects on the prevailing meteorological conditions. Atmos. Environ., 34 , 539552.

    • Search Google Scholar
    • Export Citation
  • Singh, M. P., , R. T. McNider, , and J. T. Lin, 1993: An analytical study of diurnal wind-structure variation in the boundary layer and the low-level nocturnal jet. Bound.-Layer Meteor., 63 , 397423.

    • Search Google Scholar
    • Export Citation
  • Skamarock, W. C., , J. B. Klemp, , J. Dudhia, , D. O. Gill, , D. M. Barker, , W. Wang, , and J. G. Powers, 2005: A description of the advanced research WRF version 2. NCAR Tech. Note NCAR/TN-468+STR, 100 pp.

    • Search Google Scholar
    • Export Citation
  • Smirnova, T. G., , J. M. Brown, , and S. G. Benjamin, 1997: Performance of different soil model configurations in simulating ground surface temperature and surface fluxes. Mon. Wea. Rev., 125 , 18701884.

    • Search Google Scholar
    • Export Citation
  • Smirnova, T. G., , J. M. Brown, , S. G. Benjamin, , and D. Kim, 2000: Parameterization of cold-season processes in the MAPS land-surface scheme. J. Geophys. Res., 105 , (D3). 40774086.

    • Search Google Scholar
    • Export Citation
  • Stull, R., 1993: Review of transilient turbulence theory and non-local mixing. Bound.-Layer Meteor., 62 , 2196.

  • Stull, R., , and A. G. M. Driedonks, 1987: Applications of the transilient turbulence parameterization to atmospheric boundary-layer simulations. Bound.-Layer Meteor., 40 , 209239.

    • Search Google Scholar
    • Export Citation
  • Takemi, T., 2006: Impacts of moisture profile on the evolution and organization of midlatitude squall lines under various shear conditions. Atmos. Res., 82 , 3754.

    • Search Google Scholar
    • Export Citation
  • Wu, D., , X. Tie, , C. Li, , Z. Ying, , A. K. Lau, , J. Huang, , X. Deng, , and X. Bi, 2005: An extremely low visibility event over the Guangzhou region: A case study. Atmos. Environ., 39 , 65686577.

    • Search Google Scholar
    • Export Citation
  • Yadav, S., , M. S. Chauhan, , and A. Sharma, 2007: Characterisation of bio-aerosols during dust storm period in N-NW India. Atmos. Environ., 41 , 60636073.

    • Search Google Scholar
    • Export Citation
  • Zhang, Q., , C. Zhao, , X. Tie, , Q. Wei, , M. Huang, , G. Li, , Z. Ying, , and C. Li, 2006: Characterizations of aerosols over the Beijing region: A case study of aircraft measurements. Atmos. Environ., 40 , 45134527.

    • Search Google Scholar
    • Export Citation
  • Zilitinkevich, S. S., 1995: Non-local turbulent transport: Pollution dispersion aspects of coherent structure of convective flows. Air Pollution Theory and Simulation, H. Power, N. Moussiopoulos, and C. A. Brebbia, Eds., Vol. 1, Air Pollution III, Computational Mechanics Publications, 53–60.

    • Search Google Scholar
    • Export Citation

Fig. 1.
Fig. 1.

The selected domain over the northern Indian region. Here the shaded area with asterisks represents Himalayan region, the shaded area with crosses represents Indo-Gangetic plain, and the shaded area with plus signs represent the Thar Desert. 1: Gujarat, 2: Madhya Pradesh, 3: Uttar Pradesh, 4: Uttaranchal, 5: Himachal Pradesh, 6: Jammu and Kashmir, 7: Punjab, 8: Haryana, and 9: Rajasthan; AH: Ahmedabad, DE: Delhi, and JO: Jodhpur.

Citation: Journal of Applied Meteorology and Climatology 48, 11; 10.1175/2009JAMC1926.1

Fig. 2.
Fig. 2.

Regional land surface characteristics: (a) topography, where the contours signify height in meters, (b) soil type, where the numerical values signify 19 USGS soil categories, (c) vegetation fraction during the summer case, (d) vegetation fraction during the winter case, (e) albedo during the summer case, and (f) albedo during the winter case.

Citation: Journal of Applied Meteorology and Climatology 48, 11; 10.1175/2009JAMC1926.1

Fig. 3.
Fig. 3.

Diurnal variation of surface temperature during summer case over three northern Indian cities: (a) Delhi, (b) Ahmedabad, and (c) Jodhpur for the control simulation (thin solid line), idealized simulation (thin dashed line), YSU+Noah simulation (thick dashed line), MYJ+RUC simulation (thick solid line), and observations from IMD (filled circles).

Citation: Journal of Applied Meteorology and Climatology 48, 11; 10.1175/2009JAMC1926.1

Fig. 4.
Fig. 4.

Diurnal variation of surface temperature during winter case over three northern Indian cities: (a) Delhi, (b) Ahmedabad, and (c) Jodhpur for the control simulation (thin solid line), idealized simulation (thin dashed line), and observations from IMD (filled circles).

Citation: Journal of Applied Meteorology and Climatology 48, 11; 10.1175/2009JAMC1926.1

Fig. 5.
Fig. 5.

Diurnal variation of surface wind during summer case over three northern Indian cities: (a) wind speed over Delhi, (b) wind direction over Delhi, (c) wind speed over Ahmedabad, (d) wind direction over Ahmedabad, (e) wind speed over Jodhpur, and (f) wind direction over Jodhpur for the control simulation (thin solid line), idealized simulation (thin dashed line), YSU+Noah simulation (thick dashed line), MYJ+RUC simulation (thick solid line), and observations from IMD (filled circles).

Citation: Journal of Applied Meteorology and Climatology 48, 11; 10.1175/2009JAMC1926.1

Fig. 6.
Fig. 6.

Diurnal variation of surface wind during winter case over three northern Indian cities: (a) wind speed over Delhi, (b) wind direction over Delhi, (c) wind speed over Ahmedabad, (d) wind direction over Ahmedabad, (e) wind speed over Jodhpur, and (f) wind direction over Jodhpur for the control simulation (thin solid line), idealized simulation (thin dashed line), and observations from IMD (filled circles).

Citation: Journal of Applied Meteorology and Climatology 48, 11; 10.1175/2009JAMC1926.1

Fig. 7.
Fig. 7.

Vertical potential temperature profiles over three northern Indian stations during summer case: (a) Delhi at 0000 UTC 21 May, (b) Delhi at 1200 UTC 21 May, (c) Ahmedabad at 1200 UTC 20 May, (d) Ahmedabad at 0000 UTC 21 May, (e) Jodhpur at 1200 UTC 20 May, and (f) Jodhpur at 0000 UTC 21 May for the control simulation (thin solid line), idealized simulation (thin dashed line), YSU+Noah simulation (thick dashed line), MYJ+RUC simulation (thick solid line), and observations from Wyoming Weather Web data archive (filled circles).

Citation: Journal of Applied Meteorology and Climatology 48, 11; 10.1175/2009JAMC1926.1

Fig. 8.
Fig. 8.

Vertical profiles of potential temperature over three northern Indian stations during winter case: (a) Delhi at 1200 UTC 9 Dec, (b) Delhi at 0000 UTC 10 Dec, (c) Ahmedabad at 1200 UTC 9 Dec, (d) Ahmedabad at 0000 UTC 10 Dec, (e) Jodhpur at 1200 UTC 9 Dec, and (f) Jodhpur at 0000 UTC 10 Dec for the control simulation (thin solid line), idealized simulation (thin dashed line), and observations from Wyoming Weather Web data archive (filled circles).

Citation: Journal of Applied Meteorology and Climatology 48, 11; 10.1175/2009JAMC1926.1

Fig. 9.
Fig. 9.

Vertical profiles of wind speed over three northern Indian stations during summer case: (a) Delhi at 0000 UTC 21 May, (b) Delhi at 1200 UTC 21 May, (c) Ahmedabad at 1200 UTC 20 May, (d) Ahmedabad at 0000 UTC 21 May, (e) Jodhpur at 1200 UTC 20 May, and (f) Jodhpur at 0000 UTC 21 May for the control simulation (thin solid line), idealized simulation (thin dashed line), YSU+Noah simulation (thick dashed line), MYJ+RUC simulation (thick solid line), and observations from Wyoming Weather Web data archive (filled circles).

Citation: Journal of Applied Meteorology and Climatology 48, 11; 10.1175/2009JAMC1926.1

Fig. 10.
Fig. 10.

As in Fig. 9 but for wind direction. The angles within the range 360°–450° lie in the first quadrant.

Citation: Journal of Applied Meteorology and Climatology 48, 11; 10.1175/2009JAMC1926.1

Fig. 11.
Fig. 11.

As in Fig. 8 but for wind speed.

Citation: Journal of Applied Meteorology and Climatology 48, 11; 10.1175/2009JAMC1926.1

Fig. 12.
Fig. 12.

As in Fig. 8 but for wind direction. The angles within the range 360°–450° lie in the first quadrant.

Citation: Journal of Applied Meteorology and Climatology 48, 11; 10.1175/2009JAMC1926.1

Fig. 13.
Fig. 13.

Synoptic situations over Indian region from Indian daily weather map at 0300 UTC (a) 20 May 2005 and (b) 9 Dec 2004, and (c) symbols used in (a) and (b).

Citation: Journal of Applied Meteorology and Climatology 48, 11; 10.1175/2009JAMC1926.1

Fig. 14.
Fig. 14.

Surface flow fields during control simulation in summer case at (a) 0600 UTC 20 May, (b) 1200 UTC 20 May, (c) 0600 UTC 21 May, and (d) 1200 UTC 21 May.

Citation: Journal of Applied Meteorology and Climatology 48, 11; 10.1175/2009JAMC1926.1

Fig. 15.
Fig. 15.

Difference of surface control flow fields with respect to analysis during the summer case at (a) 0600 UTC 20 May, (b) 1200 UTC 20 May, (c) 0600 UTC 21 May, and (d) 1200 UTC 21 May.

Citation: Journal of Applied Meteorology and Climatology 48, 11; 10.1175/2009JAMC1926.1

Fig. 16.
Fig. 16.

Surface flow fields during summer case in YSU+Noah simulation at (a) 0600 UTC 20 May and (b) 0600 UTC 21 May, and in MYJ+RUC simulation at (c) 0600 UTC 20 May and (d) 0600 UTC 21 May.

Citation: Journal of Applied Meteorology and Climatology 48, 11; 10.1175/2009JAMC1926.1

Fig. 17.
Fig. 17.

Surface flow fields during the control simulation in winter case at (a) 0600 UTC 9 Dec, (b) 1200 UTC 9 Dec, (c) 1800 UTC 9 Dec, and (d) 0000 UTC 10 Dec.

Citation: Journal of Applied Meteorology and Climatology 48, 11; 10.1175/2009JAMC1926.1

Fig. 18.
Fig. 18.

Difference of the surface control flow fields with respect to analysis during winter case at (a) 0600 UTC 9 Dec, (b) 1200 UTC 9 Dec, (c) 1800 UTC 9 Dec, and (d) 0000 UTC 10 Dec.

Citation: Journal of Applied Meteorology and Climatology 48, 11; 10.1175/2009JAMC1926.1

Fig. 19.
Fig. 19.

Surface flow fields during the summer case in the idealized simulation at (a) 0600 UTC 21 May and (b) 1200 UTC 21 May, and the corresponding difference in surface flow fields with respect to the control simulation at (c) 0600 UTC 21 May and (d) 1200 UTC 21 May.

Citation: Journal of Applied Meteorology and Climatology 48, 11; 10.1175/2009JAMC1926.1

Fig. 20.
Fig. 20.

Near-surface forward trajectories during summer case from (a) Jodhpur starting at 0000 UTC, (b) Jodhpur starting at 1200 UTC, (c) OA (27°N, 72°E) starting at 0000 UTC, (d) OA (27°N, 72°E) starting at 1200 UTC, (e) OB (28°N, 72.5°E) starting at 0000 UTC, (f) OB (28°N, 72.5°E) starting at 1200 UTC, (g) OC (29°N, 75°E) starting at 0000 UTC, and (h) OC (29°N, 75°E) starting at 1200 UTC simulation. Here, the filled square indicates origin of the dust and the filled diamond represents Delhi. The solid line represents the control simulation and dotted line represents idealized simulation.

Citation: Journal of Applied Meteorology and Climatology 48, 11; 10.1175/2009JAMC1926.1

Fig. 21.
Fig. 21.

Near-surface forward trajectories during 26–28 May 2005 case from different points (same as that of Fig. 20) over the Thar region starting at (a) 0000 and (b) 1200 UTC 26 May.

Citation: Journal of Applied Meteorology and Climatology 48, 11; 10.1175/2009JAMC1926.1

Table 1.

The physics schemes in WRF used during the simulations.

Table 1.
Table 2.

Correlation coefficients between computed surface temperatures and IMD observations during control and idealized simulations.

Table 2.
Save