MESOSCALE MODELING FOR MOUNTAIN WEATHER FORECASTING OVER THE

BY SOMESHWAR DAS, S. V. SlNGH, E. N. RAJAGOPAL, AND ROBERT GALL

Severe weather over the Himalayas has calamitous effect, due to complex terrain, poor

development, and a fragile economy, and warrants intensified disaster mitigation through

increased observations from radars and satellites and further application of mesoscale models.

orthwest India and the Himalayas (Fig. 1) are Nparticularly prone to vagaries of severe weather, which claims casualties every year. This region is influ- enced by extratropical dis- turbances that propagate from the west to India during winter, bringing rainfall and chilly weather. These systems severely influence life in the Himalayas by inducing wide- spread rainfall and, at times, very heavy snowfall associ-

FIG. I. Topography over the Indian region.

AFFILIATIONS: DAS, SINGH, AND RAJAGOPAL—National Centre for E-mail: [email protected] Medium Range Weather Forecasting, New Delhi, India; GALL— DOI: 10.1 175/BAMS-84-9-1237 National Center for Atmospheric Research, Boulder, Colorado In final form 7 March 2003 CORRESPONDING AUTHOR: Dr. Someshwar Das, NCMRWF, ©2003 American Meteorological Society INSAT Building, Lodhi Rd., New Delhi I 10 003, India

AMERICAN METEOROLOGICAL SOCIETY SEPTEMBER 2003 BAflSr | 1237

Unauthenticated | Downloaded 09/29/21 12:39 AM UTC TABLE 1. Workshop speakers, affiliations, and presentation titles.

Speaker Affiliation Title of talk

R. Gall NCAR The Observing System Research and Predictability Experiment (THORPEX) and Weather Research Forecast (WRF) model

H. R. Hatwar IMD Development of mountain meteorology over western Himalayas

S. S. Sharma SASE Mesoscale modeling as an aid for site-specific avalanche forecast

S. Das NCMRWF Mountain weather forecasts over the Indian region using MM5 model

H. R. Hatwar IMD Prediction of western disturbances and associated weather over western Himalayas

V. J. Kirubhanidhi IAF The role of katabatic and anabatic winds in the weather over Assam with special reference to Tezpur air field

A. P. Dimri SASE Point-probabilistic prediction of precipitation and quantitative precipitation forecasting in northwestern Himalayas

U. C. Mohanty IITD Mesoscale modeling of intense vortices over India

D. V. Bhaskar Rao AU Numerical model experiments of weather disturbances over India using a mesoscale model

A. R. Kellie NCAR Supercomputing division of NCAR

P. Goswami CMMACS Variable-resolution GCM: A promising tool for to mesoscale simulations

S. S. Vaidya HTM Simulation of tropical systems over Indian region using mesoscale models

E. N. Rajagopal NCMRWF Mesoscale forecasts with Eta Model over Indian region

R. Ramachandran SPL ISRO Numerical simulations over peninsular India during monsoon seasons using a nonhydrostatic mesoscale model

S. Mohandas NCMRWF Sensitivity of land surface parameterization on RSM forecast

N. V. Sam IITD Surface fluxes and related convective activity during MONEX-79: A single-column experiment

S. S. Roy IMD Analysis of thermodynamics of atmosphere over northwestern India during a western disturbance as revealed by model analysis

S. Raghavan IMD Radar observations of mesoscale phenomena and tropical

A. R. Jain NMRF Study of tropical mesoscale convective systems over Gadanki using MST radar and collocated L-band radar

G. Viswanathan RDC ISRO Indigenous development of Doppler weather radar in India

P. C.Joshi SAC ISRO Satellite data for mesoscale studies in the context of Indian scenario

K. Sawyer NCAR Introduction to University Corporation for Atmospheric Research (UCAR)

V. S. Prasad NCMRWF High-resolution satellite datasets for mesoscale applications

R. M. Saxena IAF A satellite study of western disturbances over northwest India during winter months

D. R. Sikka ICRP Mesoscale weather disturbances over India during different seasons

A. K. Das IITD Mesoscale assimilation and comparative study of circulation characteristics of a monsoon depression

G. Srinivasan IMD DST Daily rainfall characteristics from a high-density rain gauge network

O. P. Madan IITD Contrast in meteorological fields of surplus and significant winter seasonal precipitation over western Himalayas

R. M. Saxena IAF Mesoscale climatology of Ladakh area

O. P. Sharma IITD Mesoscale circulations forced by topographic heterogeneity

M. Mandal IITD Mesoscale analysis and tropical forecast: A case study

Y. V. Rama Rao IMD Cyclone-track prediction with IMD limited-area model and quasi-Lagrangian limited-area model

P. K. Mohanty BU A study on the role of mesoscale phenomena on the incidence of wet and dry conditions over Orissa

K. Srinivasan SASE A case study of western disturbance using MM5 model

M. R. Sakya NMS Weather forecasting in Nepal

1238 | BAflS- SEPTEMBER 2003

Unauthenticated | Downloaded 09/29/21 12:39 AM UTC TABLE 2. Acronyms of affiliations.

Acronym Location

AU Andhra University, Waltair, India BU Berhampur University, Berhampur, India CMMACS Centre for Mathematical Modeling and Computer Simulations, Bangalore, India DST Department of Science and Technology, New Delhi, India IAF Indian Air Force, New Delhi, India ICRP Indian Climate Research Program, New Delhi, India IITD Indian Institute of Technology, Delhi, India HTM Indian Institute of Tropical Meteorology, Pune, India IMD India Meteorological Department, New Delhi, India ISRO Indian Space Research Organization, Bangalare, India NCAR National Center for Atmospheric Research, Boulder, Colorado NCMRWF National Centre for Medium Range Weather Forecasting, New Delhi, India NMRF National MST Radar Facility, Andhra Pradesh, India NMS Nepal Meteorological Service, Kathmandu, Nepal RDC Radar Development Cell, India SAC Space Application Centre, Gujarat, India SASE Snow and Avalanche Study Establishment, Manali, India SPL Space Physics Laboratory, Trivandrum, India VSSC Vikram Saravai Space Centre, Trivandrum, India

ated with squall winds, hail, and severe cold waves. disasters of this nature, and to develop the capacity Gale winds and heavy rain/snowfall result in ava- to predict them, observations and modeling on a lanches and landslides. smaller spatial scale, that is, mesoscale, are necessary. Nor'westers (squall lines that affect northeastern The understanding and prediction of weather are India during the premonsoon season), cloudbursts, advanced through analysis of observations and nu- windstorms, hailstorms, and lightning have high im- merical modeling of physical processes. pact on human life and the economy. These events, The year 2002 was the International Year of Moun- occurring on a smaller spatiotemporal scale, invite tains, encouraging governments to ensure sustainable application of regional models with fine-grid resolu- mountain development. In view of the large impact tion. The squall that hit Delhi during 27 May 2002 of mountain weather, a workshop on mesoscale mod- uprooted around 2000 trees, leading to traffic jams, eling, with special emphasis on mountain weather water logging, and electric supply disruption. On forecasting, was jointly organized by the National 14 November 1986, the combined effect of weather Centre for Medium Range Weather Forecasting and avalanches claimed 70 lives and property worth (NCMRWF) of India and the National Center for At- millions of rupees on the Srinagar-Leh Highway. On mospheric Research (NCAR) on 29-30 July 2002 in 16 January 1995, the Jammu-Srinagar Highway was New Delhi (Tables 1 and 2). The workshop was divided hit by a number of avalanches, stranding more than into five sessions: Mountain Weather Forecasting, 300 vehicles and 400 people; 67 lives and property Mesoscale Atmospheric Modeling, Mesoscale Obser- worth billions of rupees were lost. The Malpa land- vations and Data Assimilation, Radar and Satellite Ap- slide disaster in the hills of Himachal Pradesh, owing plications, and Severe Weather Phenomena. The to heavy rainfall on 27 August 1997, is still fresh in scientists discussed various aspects of mesoscale mod- our memory. During the Amarnath pilgrimage in eling; a panel discussion at the end produced recom- August 1996, 194 people died because of rough mendations for improving severe weather forecasts over weather. Innumerable incidents of this nature take the mountains, in particular, over the monsoon region place over hilly areas. To combat weather-induced (Table 3). We present a brief summary of each session.

AMERICAN METEOROLOGICAL SOCIETY SEPTEMBER 2003 BAflS* | 83

Unauthenticated | Downloaded 09/29/21 12:39 AM UTC TABLE 3. Key issues and recommendations.

Issues Recommendations

What does it take to make • Input observations at very high resolution a better weather forecast • A mesoscale model of high resolution (< 1 —10 km) with appropriate physics over the mountains? • Appropriate 3D/4D variational data assimilation system

How are better initial • Set up mesoscale network of automatic weather stations along the entire conditions generated for Himalayan belt mesoscale forecasting • Install Doppler weather radars covering the Himalayan range over the Himalayas? • Utilize satellite observations over the region

How is accuracy • Conduct sensitivity experiments with the physical processes of forecasts over the • Examine the forecasts at different resolutions in the horizontal and vertical Himalayas increased?

Do we understand the • Conduct special field experiments at suitable locations to study severe physical processes over , cloud microphysics, and other mesoscale convective complexes the mountains? • Study air flow, mass, and moisture fields over and around the mountain ranges under various synoptic conditions, western disturbances, and nor'westers • Study the precipitation patterns and orographic contributions at different altitude ranges during summer and winter seasons

What are the dynamical • Study dynamic and physical processes associated with orographically forced effects of mountains on thunderstorms the weather systems? • Improve understanding of mountain waves, lee , gravity wave drag, and parameterization of subgrid-scale orography in NWP models

MOUNTAIN WEATHER FORECASTING. mesoscale model outputs, global model products, in The importance of the Himalayas to the monsoon situ observations, and satellite observations, along circulation over the Indian subcontinent is well with synoptic conditions by collaborative efforts be- known (Desai 1968; Godbole 1973; Hahn and Manabe tween NCMRWF, IMD, and SASE. 1975; Das and Bedi 1978, Yanai et al. 1992). The In addition to these techniques, mesoscale model- Himalayan snow cover during winter is a key factor ing can aid in site-specific avalanche forecasts. To in long-range forecasts of monsoon rainfall (Vernekar predict an avalanche accurately, a forecaster must et al. 1995). However, real-time forecasting of severe keep a continuous record of the weather elements that weather, including western disturbances over the contribute to metamorphism and, also, of the exces- mountains, remains a challenging task, especially with sive loading of the snow cover due to snow precipita- sparse data and difficult communications. To enhance tion, cornice formation and its collapse, wind trans- meteorological observations over the western portation, and more. The information for these factors Himalayas, 26 surface observatories and two upper- is sufficient on a synoptic scale for a period-specific air stations have been set up. A mountain meteorol- forecast for a general area, but it is not sufficient for ogy program has been started at NCMRWF, the India site- and time-specific avalanche forecasts. This, to a Meteorological Department (IMD), and the Snow certain extent, can be achieved through mesoscale and Avalanche Study Establishment (SASE) at Manali modeling of various weather parameters, such as tem- for the purpose of forecasting, providing training, and perature, wind speed/direction, and humidity. development of snow climatology, among other Another important forecast parameter is the type of things. Currently, mountain weather forecasting over precipitation (snow, ice, rain) and, if available, the ex- the western Himalayas is carried out through a com- tent of crystal formation. The forecasts at 10-km reso- bination of various products, including regional/ lution in the horizontal and 100-m resolution in the

1240 | BAflS- SEPTEMBER 2003

Unauthenticated | Downloaded 09/29/21 12:39 AM UTC vertical (up to 4-5 km) are good enough for making The Eta Model is presently run at 48-km resolu- avalanche forecasts. Attempts have been made for tion using the boundary conditions from the T80 glo- point-probabilistic prediction of precipitation and bal model. Model forecasts of western disturbances quantitative precipitation forecasting (QPF) in the and heavy rainfall are fairly good up to 72 h (Rajagopal northwestern Himalayas. Studies based on the Na- and Iyengar 2001). Sensitivity studies with two differ- tional Centers for Environmental Prediction (NCEP) ent land surface parameterizations, using a Regional and NCMRWF analyses/forecasts, and observations Spectral Model (RSM) have been carried out at from IMD indicate an accuracy of 80% for 24-h lead NCMRWF. The first scheme has a two-layer struc- time forecast at Sonamarg in the western Himalayas. ture for the soil moisture, whereas the second scheme Also important to weather over the Himalayas are has a single-layer soil moisture parameterization. the katabatic and anabatic winds on the mountain Results using the two-layer soil moisture scheme show slopes. Observations from 1991-2000 during Janu- less easterly wind bias at lower levels. ary-April show a high diurnal variation of rainfall in Attempts have been made to simulate a severe cy- the Assam valley in the eastern Himalayas. The most clonic over the Bay of Bengal using a Regional probable time for rain is between 2100 and 1000 LST. Atmospheric Modeling System (RAMS) at the Indian Institute of Tropical Meteorology (IITM). Initial con- MESOSCALE ATMOSPHERIC MODELING. ditions were prepared from the Global Energy and Mesoscale modeling for real-time forecasting is evolv- Water Cycle Experiment (GEWEX) Asian Monsoon ing in India. At NCMRWF, high-resolution mesos- Experiment (GAME) analysis. Available radiosonde/ cale models, such as the fifth-generation (Pennsylva- rawinsonde (RS/RW) and surface data were assimi- nia State University) PSU-NCAR Mesoscale Model lated into the analysis. The predicted storm track (MM5) and Eta Model are run on a real-time basis compared well with the GAME analyses and IMD for forecasting mesoscale systems, such as the west- observation through 24 h by using the RAMS at ern disturbances, severe thunderstorms, tropical cy- 50-km resolution. clones, and heavy rainfall episodes. The MM5 model At IITM an Advanced Regional Prediction System is run on triple-nested domains (at 90-, 30-, and (ARPS) at 50-km resolution over the Indian region 10-km resolutions), using initial conditions from the has been used to simulate the monsoon depression, T80 global model of NCMRWF. For the purpose of mesoscale circulations, and its time evolution over mountain weather forecasting, four inner domains at peninsular India. The group working at Vikram 10-km resolution are placed over the northwest, cen- Sarabhai Space Centre (VSSC) initializes the model tral, and northeast Himalayas, and western Ghat with vertical profiles of temperature, relative humid- Mountains on an experimental basis. Case studies of ity, and winds available at Thumba. western disturbances over the Himalayan range and The rich and complex scale interactions in the heavy rainfall events associated with the active mon- Tropics make simulation and forecasting a challenge. soon phase have shown that the model has reasonably Both up- and down-scale energy transfer takes place. good capability of forecasting severe weather up to Thus, while small-scale (including subgrid scale) pro- 72 h in advance (Das 2002). However, heavy rainfall cesses play a crucial role in monsoon dynamics, the events are underestimated; further studies will be large-scale fields also seem to significantly affect small needed to improve the physical processes and model (mesoscale) systems. It is, therefore, desirable to em- resolution. Scarcity of observations over the ploy a global environment for simulating mesoscale Himalayas remains a major issue. systems, as well. A main drawback of a mesoscale fore- Studies of intense atmospheric vortices, particu- cast model is that the data for the boundary condi- larly tropical cyclones, have been carried out using tions must be supplied externally. The artificial lat- MM5 by groups working at NCMRWF, the Indian eral boundaries in a mesoscale model arise because of Institute of Technology, Delhi (IITD), and Andhra its limited domain. A general circulation model University. Results indicate that the cyclone track and (GCM) can generate the boundary conditions, but the rainfall forecasts by MM5 are improved using satel- quality of the mesoscale simulation then strongly de- lite data in the input analysis. Experiments using a pends on their quality. Preliminary results of mon- synthetic vortex and four-dimensional data assimila- soon simulations at the Centre for Mathematical tion (4DDA) indicate that the mesoscale data assimi- Modeling and Computer Simulations (CMMACS) lation, appropriate parameterization of land surface have shown that with a carefully chosen configuration, processes, and the mesonetwork of observations are the variable-resolution GCM can be effective in mod- very important for improving the quality of forecasts. eling scales from global to monsoon to the mesoscale.

AMERICAN METEOROLOGICAL SOCIETY SEPTEMBER 2003 BAflS* | 1241

Unauthenticated | Downloaded 09/29/21 12:39 AM UTC MESOSCALE OBSERVATIONS AND DATA cipitation intensity and distribution showed that the ASSIMILATION. Spells of very heavy rainfall along model responded positively after data assimilation. the west coast of India occur over an area of 200 km Analysis of limited observations over the Ladakh x 200 km and may last for 1-2 days. The terrain of region in the western Himalayas has been carried out the west coast of India rises to about 2695 m in the based on some special data. Ladakh is popularly western Ghat Mountains. There is clear evidence that termed as a land between earth and sky at an eleva- deep convection is embedded within the large rain- tion of 10 680 feet. It is in a remote and rugged land fall events, favoring spots close to the high terrain and deep in the Himalayas on the border of Tibet, in the horizontal wind shear in the low-level jet. These Sinkiang, and Kashmir. Rocky defiles of grotesque rainfall events coincide with the formation of a north- proportion, broad valleys cut by lazy rivers, and arid south over the coast, sometimes with an em- expanses of rock merge suddenly with moist, upland bedded onset vortex. A high-density gauge network meadows with dense shrubs and alpine flowers. Arctic in 1996-2000 in the western Ghats, in and around and desert conditions make its climate among the Mumbai, indicates that precipitation events exceed- most peculiar in the world. Based on limited obser- ing 70 mm constituted from 29% to 77% of the aver- vations, a climatology of thundershowers, drizzle/rain age annual rainfall. The percentages were the largest showers, snow/sleet, poor visibility, low clouds, hail, for gauges located on the western Ghats and the least and haze has been constructed. Observations indicate for the foothills of the Ghats. They were moderately that visibility can diminish to < 800 m during Septem- large along the coast. These results suggest that dif- ber between 0600 and 1200 LST; in Thoise, from April ferent climatic controls influence heavy precipitation to May, and December between 1500 and 1800 LST; over these regions. Understanding of the mesoscale in Jammu, throughout the year between 0600 and organization of deep convection along the west coast 0900 LST; in Leh, during January, February, and of India is a goal of the Arabian Sea Monsoon Experi- November between 0600 and 0900 LST; and in ment (ARMEX). There are 25 institutions/organiza- Srinagar, throughout the year between 0600 and tions, including the Indian Air Force and Navy, par- 0900 LST. Observations show that the maximum oc- ticipating in this field campaign. Unfortunately the currence of low clouds (< 60 m) and fog is observed 2002 monsoon was not very active. Heavy rain fell over Jammu and Srinagar. Low clouds occur mostly only twice through the end of July. These episodes in March and April in Srinagar, and during July and were very well predicted by the mesoscale models at August in Jammu between 0900 and 1200 LST. Low NCMRWF. clouds are not observed in Leh and Kargil. Fog occurs Prior to ARMEX, another field experiment called mostly between 0600 and 0900 LST in the region. Thoise the Bay of Bengal Monsoon Experiment (BOBMEX) and Kargil do not experience fog. Haze occurs in was organized during July-August 1999. Here, MM5 Jammu and Srinagar throughout the year, and in 3-4 was used to study the circulation of a monsoon de- months in a year in Thoise and Leh, and only in winter pression during BOBMEX (26-29 July 1999). Using in Kargil on few occasions. It generally occurs between two global datasets from NCMRWF and the NCEP- 0600 and 0900 LST. The maximum occurrence of thun- NCAR reanalysis separately as a first guess, a derstorms is observed in Jammu and Srinagar, while the multiquadric (MQD) interpolation technique was least occurrence is in Leh and Thoise (and not at all used for objective analysis at a horizontal resolution in Kargil). Thunderstorms mostly occur in May and of 30 km. Synoptic surface and upper-air observations July, between 0300 and 0600, and 1500 and 1800 LST. plus the European Meteorological Satellite

(METEOSAT) and Special Sensor Microwave Imager RADAR AND SATELLITE APPLICATIONS. (SSM/I), ORV Sugar Kanya, and coastal RS/RW data One of the major problems of mesoscale modeling is were utilized to improve the mesoscale analysis. A the paucity of data. The extent of operational obser- simple data assimilation scheme, based on nudging vations is well below the scales required by mesoscale (Newtonian relaxation), dynamically constrained the models. Mesoscale analyses, therefore, increasingly model. The model was able to simulate the structure rely on new and high-resolution satellite data and of mesoconvective organization and the prominent other remote sensing systems, such as radar. In addi- synoptic features associated with monsoon depres- tion, increasing use is made of proxy data, such as sion. The objective verification and error analysis of moisture profile and adiabatic heating from infrared different simulations from two reanalysis datasets il- imagery and bogus data. Present-day lustrated that the mesoscale assimilation improved the space technology is able to circumvent these problems performance of the model. Investigation of the pre- to some extent.

1242 | BAflS- SEPTEMBER 2003

Unauthenticated | Downloaded 09/29/21 12:39 AM UTC Recently, IMD has begun to replace the existing soscale forecasts. Also, the prediction of movements cyclone detection radars with state-of-the-art Doppler of tropical cyclones is significantly improved with the radars. Three have been installed at Chennai, Kolkata, use of scatterometer data, for wind vectors over and Sriharikota on the east coast of India. The last of oceans. Work is under way in India to receive data these was developed by the Indian Space Research from the Moderate Resolution Imaging Spectrora- Organization (ISRO). With Doppler technology in diometer (MODIS) on board the Terra/Aqua satel- IMD and an expansion to cover the entire coastline lites of the National Aeronautics and Space Admin- and areas prone to storms, cyclone warnings and istration (NASA). This promises very high resolution short-range forecasts are expected to improve. The humidity profiles. Megha Tropique (to be launched radars are expected to support numerical modelers. in the year 2005) will provide high-resolution and While Doppler wind measurements provide a direct frequent data coverage over the Tropics. A sensor estimate of storm intensity, the availability in real time called Geostationary Imaging Fourier Transform of data and imagery in composite form through net- Spectrometer (GIFTS) is presently under active inves- working will make cyclone warnings more efficient. tigation. This would provide humidity and tempera- Aviation forecasting and storm warning will get a ture with very high vertical and horizontal resolution. boost from various hydrological and wind products Increasing efforts are being made at NCMRWF to derived instantaneously. generate and study high-resolution satellite databases Besides the Doppler radar, a mesosphere-strato- for use in the planned mesoscale analysis system. The sphere-troposphere (MST) radar has been opera- advanced satellite datasets from SSM/I, Quick tional in India at Gadanki (13.47°N, 79.4°E). Several Scatterometer (QuickSCAT), TRMM Microwave field campaigns have used the MST radar to study Imager (TMI), Advanced Television Infrared Obser- deep convective systems. Two important physical vation Satellite (TIROS) Operational Vertical processes—tropopause weakening and gravity waves Sounder (ATOVS), Atmospheric Infrared Sounder triggered by mesoscale convective systems—have (AIRS), and India's own satellites: Multichannel Scan- been examined to get an idea of related dynamical ning Microwave Radiometer (MSMR), INSAT, and processes. The convective storms are generally fol- METSAT are now being actively studied and lowed by stratiform precipitation. The height-time archived. Some of these datasets are already being structure of vertical velocity during these two types used operationally in the NCMRWF global data as- of precipitating events can help in distinguishing such similation and forecasting system. systems. The composite height profiles of vertical ve- locity in convective and stratiform precipitation in- SEVERE WEATHER. Mesoscale weather forecast- dicate height-time structure of diabatic heating. ing continues to be a challenging problem. The Coun- Collocated L-band radar observations at Gadanki are cil of Scientific and Industrial Research has recognized used to classify the observed precipitating systems into it as a problem where interinstitutional cooperation convective and stratiform regions and, thus, to obtain is essential and, therefore, launched an ambitious composite height profiles of vertical velocities. Results project on mesoscale modeling for monsoon-related from these studies can be used to prepare initial con- predictions. ditions and verify output from mesoscale models. The Orissa super cyclone of 29-30 October 1999 Shortly, three geostationary Indian satellites, resulted in more than 10,000 deaths, damaged namely the Indian National Satellite (INSAT)-2E, 1,828,532 houses, and 1,810,091 hectares of agricul- Meteorological Satellite (METSAT), and INSAT-3A tural land and significant infrastructure destruction will be available over the Indian Ocean region. A Very affecting one-third of the population of the state. The High Resolution Radiometer (VHRR) on board these storm, with winds of over 250 km h_1 (160 mph), ba- satellites gives information on cloud motion vectors, sically wiped out everything in its path. cloud cover, outgoing longwave radiation, precipita- Due to inadequate data coverage over the ocean, tion, and more. Charge coupled devices (CCD) on the intensity of such cyclones is poorly represented board INS A T-2E/INSA T-3A add cirrus detection, in global analyses/reanalyses. Most of the global analy- vegetation monitoring, and continuous monitoring of ses/reanalyses also have large initial positional errors. mesoscale systems. Microwave radiometers on board Investigators at IITD used satellite (METEOSAT, the Defense Meteorological Satellite Program SSM/I, MSMR), surface, and upper-air observations (DMSP) and Tropical Rainfall Measuring Mission over India with the NCEP reanalysis as the back- (TRMM) satellites give the water vapor content and ground field. The improved analysis was used to in- precipitation rates, which significantly improve me- tegrate the model (MM5) for a 5-day forecast of the

AMERICAN METEOROLOGICAL SOCIETY SEPTEMBER 2003 BAflS* | 1243

Unauthenticated | Downloaded 09/29/21 12:39 AM UTC storm from 0000 UTC 26 October to 0000 UTC 31 October 1999. The mesoscale analysis improved REFERENCES the prediction skill of the model both for track and Das, P. K., and H. S. Bedi, 1978: Inclusion of Himalayas intensity of the cyclone. in a PE model. Indian J. Meteor. Hydrol. Geophys., 29, The Orissa super cyclone was also investigated by 373-383. two different models used at IMD for regional Das, S., 2002: Evaluation and verification of MM5 fore- weather and cyclone prediction. The first model is a casts over the Indian region. Proc. 12th PSU/NCAR limited-area model (LAM), which includes a synthetic Mesoscale Model Users Workshop, Boulder, CO, vortex for tropical cyclones. The LAM is a semi- NCAR, 77-81. Lagrangian, semi-implicit, multilayer primitive equa- Desai, B. N., 1968: Influence of topographic features of tion model on sigma coordinates. The model has hori- Indian sub-continent on its weather and climate. zontal resolution of 0.75° lat x 0.75° long and 16 sigma Geogr. Rev., 30, 33-44. levels in the vertical. The second model, adapted from Godbole, R. V., 1973: Numerical simulation of the In- NCEP for hurricane forecasting, is a quasi-Lagrangian dian summer monsoon. Indian J. Meteor. Hydrol model (QLM). The QLM is a multilevel fine-mesh Geophys., 24, 1-14. model with a horizontal resolution of 40 km and Hahn, D. G., and S. Manabe, 1975: The role of moun- 16 sigma levels in the vertical, with a variable domain tains in the south Asian monsoon circulation. /. depending on the center of the vortex. In this model Atmos. Sci., 32, 1515-1541. a more complex vortex circulation defined by an ana- Rajagopal, E. N., and G. R. Iyengar, 2001: Implementa- lytical expression is added after the objective analy- tion of NCEP's Eta Model over Indian region. sis, but before the model initialization. Results from Research activities in atmospheric and oceanic mod- the LAM and QLM show that the track prediction eling Rep. 31, WMO/TD No. 1064, 5.32-5.33. errors in both the models are of comparable magni- Vernekar, A. D., J. Zhou, and J. Shukla, 1995: The effects tude. The 24-h model forecast error was about 80 km of Eurasian snow cover on the Indian monsoon. /. in both models. Climate, 8, 248-266. Yanai, M., C. Li, and Z. Song, 1992: Seasonal heating of ACKNOWLEDGMENTS. The workshop was con- the Tibetan Plateau and its effects on the evolution ducted under grants provided by the Department of Sci- of the Asian summer monsoon. /. Meteor. Soc. Japan, ence and Technology, Government of India. 70, 319-351.

Mid-Latitude Weather Systems is the first text to make extensive use of conventional weather charts and equations to fully illustrate the behavior and evolution of weather patterns. Presenting a fusion between the mathematical and descriptive fields of meteorology and integrated coverage of synoptic and dynamic ap- proaches, Mid-Latitude Weather Systems provides students with an in- valuable course text and reference source to gain an unclouded ap- preciation of the underlying processes and behavior of mid-latitude weather patterns.

Mid-Latitude Weather Systems is available for $47/list, $37/AMS members, or $27/ student members, by sending prepaid orders to: Order Department, AMS, 45 Bea- con Street, Boston, MA 02108-3693. Please make checks payable to the American Meteorological Society, or call 617-227-2425 to order by phone (Mastercard, VISA, and American Express accepted).

1244 | BAflS- SEPTEMBER 2003

Unauthenticated | Downloaded 09/29/21 12:39 AM UTC