Simulation of Wave Climate for the Arabian Sea and Bay of Bengal Thesis submitted in partial fulfilment of the requirements for the degree of DOCTOR OF PHILOSOPHY . In PHYSICAL OCEANOGRAPHY J. SWAIN Naval Physical and Oceanographic Laboratory Cochin -682021 Cochin University of Science and Technology Cochin-682021 August 1997 Certificate This is to certify that this thesis entitled "Simulation of Wave Climate for the Arabian Sea and Bay of Bengal" is an authentic record of research work carried out by Mr. Jaganath Swain (M.Sc., M.Phi!.), Scientist-D of Naval Physical and Oceanographic Laboratory, Cochin-21 under my supervision and guidance for the Ph.D. degree of Cochin University of Science and Technology, Cochin-22 and no part of this has previously formed the basis for the award of any other degree in any other university. Place: Trivandrum, Dr. M. Baba Date: August 1997. Head, Marine Science Division Centre for Earth Science Studies Akkulam, Thuruvikal P.O. Trivandrum, Kerala. PREFACE One of the well known ocean enviromental disturbances is the wind-induced surface gravity waves which are important in the air-sea interaction process of the coupled ocean atmosphere system. The wave information or the sea state is essential for efficient management and use of our coastal and offshore environment. In the absence of routine operational wave forecasts and long-term wave measurements for the Arabian Sea and Bay of Bengal, visually observed wave data are being used for several applications. Visual observations are routinely made from weather ships, land-based stations and ocean going vessels. They are available over longer periods and cove.r the shipping routes. However, visual observations are only the first estimates and can't be considered as reliable. The next alternative is the establishment of a climatic database through model hindcasting. For the Indian Seas (Arabian Sea and Bay of Bengal) there is a limitation for conducting a long-term wave hindcast due to the scarcity of barotropic data used for the preparation of synoptic weather charts. Therefore, an attempt is made to conduct a simulation experiment for the establishment of deep water wave climate for the Indian Seas using available historical data as input to the wave model. The results of this simulation experiment are presented in this thesis. The thesis consists of seven chapters. The introductory chapter presents some general terminologies and definitions in "Kymatology" (the science of ocean waves) related with the present study and a brief discussion on various sources of data for wave climate establishment. It describes the requirement and availability of visually observed wave data, measured wave data, hindcast wave data, and operational wave forecast data for establishment of a long-term regional wave climate. It. also gives a qualitative assessment of these data from the climatic point of view. The significance of wave climate simulation and the scope of the present work are discussed. Wave prediction techniques which are presently in use are discussed in the second chapter. Brief outline of the numerical wave prediction models including the model used in this study for wave climate simulation are given. It is found that the third generation wave model (W AM) which was developed and demonstrated by the W AMDI group is quite suitable for this study. It represents the physics of the wave evolution in accordance with our present day knowledge, for the full degrees of freedom for the two dimensional wave spectrum. The fundamental equations, numerical scheme, model grid structure, and the input and output options of the model are described in brief. It also describes the model performance and the intercomparisons with various other models. The source of input wind and surface current data used in this study for simulation of regional wave climate is discussed in the third chapter. It provides information on the quality of input data and the methods of estimation of mean monthly fields. It also includes a brief discussion on the summer and winter monsoon wind variability and the monsoonal surface circulation. The most important part of this chapter is the specification of input data to the model. The fourth chapter discusses the wave model implementation for the regional grid system. The input and output specifications to the model are explained. Model execution using "the mean climatic year of winds" and the compilations of various model outputs are also presented. Wave model is executed for the regional grid system (50-100 deg. east, 0-25 deg. north, lxl deg. resolution) assuming appropriate boundary conditions. Representative model runs are carried out for all the months from January to December. The model outputs are stored for all grid points at each time step while the spectral outputs are stored for selected grids. The monthly, seasonal and annual distributions of significant wave height, period and mean direction for total sea and swell are estimated after post processing of model outputs. Validation of simulated wave climate is presented in the fifth chapter. The simulation experiment has been carried out in spite of limited measurements available for the Indian Seas. Hence, a complete validation of the simulation results is not possible with the available measurements. However, a qualitative validation of the height and period parameters are carried out using available ship-borne wave recorder data and ship reported visually observed data. The wave height data are compared with the Geosat altimeter data for two selected locations. The results and discussion on the regional wave climate based on the simulation experiment are presented in the sixth chapter. The spatial distributions of significant wave parameters are discussed (monthly, seasonal and annual) in detail. The bivariate and cumulative distributions of significant wave height and period have been analysed for the Arabian Sea and Bay of Bengal during rough weather (May-September) and fair weather (October-April) seasons. Secondly, a comparative study of the wave climate between (1) the Arabian Sea and Bay of Bengal and (2) East and west coast of India are presented in this chapter. The spectral characteristics for a selected site are also discussed. Finally, the limitations of the simulated wave climate are explained.· The last chapter summarises the important results of the simulation experiment. The study suggests that the model could successfully simulate the wave climate for the Indian Seas based on the mean climatic year of winds and the simulated database can be utilized for several applications. ii CONTENTS Page LI ST OF FI GURES .................................................. vi LIST OF TABLES ................................................... ix 1CIHlAIPlrlEIR I. INTRODUCTION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-11 1.1 General terminologies ..................................... 1 1.2 Defini tion of wave climate ................................ 3 1.3 Data source for wave climate 4 1.3.1 Quality and duration of wave data .................... 4 1.3.2 Visually observed wave data .......................... 5 1.3.3 Measured wave data 6 I. 3. 4 Hindcast wave data 7 1.3.5 Operational wave forecast data ....................... 8 1.4 Significance of wave climate simulation ................... 8 1.4.1 Concept of "mean climatic year of winds" ............. 9 1.4.2 The simulation process ............................... 9 1.5 Scope of the present work ................................. 10 1CIHlAIPlrIEIR I I. SELECTI ON OF WAVE MODEL ............................ . 12-29 11.1 Introduction ............................................ . 12 11.2 Early wave prediction techniques 12 11.3 Numerical wave prediction models 13 11. 4 Wave model intercomparison study 14 11.5 The present model ........................................ 16 11. 5.1 Background of the model ............................. 17 11. 5. 2 Fundamental equations ............................... 18 11.5.3 Numerical scheme.................................... 21 11.5.4 Model grid structure................................ 24 11.5.5 Model system........................................ 25 11.5.6 Model options and user inputs ....................... 25 iii 11. 5. 7 nodel outputs 27 11.6 Performance of the model ................................ . 28 Ill. INPUT DATA FOR SIMULATION OF VAVE CLIMATE ......... . 30-41 C~A~l[~ 111.1 Introduction............................................ 30 111.2 Source of wind and surface current data ................. 30 111.3 Assessment of data quality 31 111.4 Estimation of mean monthly wind fields .................. 33 111.5 Discussion on the Indian monsoon and surface circulation 35 111.5.1 Summer and winter monsoon winds .................... 35 Ill. 5. 2 nonsoonal surface circulation...................... 36 111.6 Input data specification for the present study.......... 37 IV. SIMULATION OF VAVE CLIMATE ......................... . 42-46 C~A~l[~ IV.1 Introduction..... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 IV.2 Vave model implementation 42 IV.2.1 Regional grid system 43 IV.2.2 Input and output specifications ..................... 43 IV.3 Vave model execution 44 IV.4 Compilation of model outputs ............................. 46 C~A~i[~ V. VALIDATION OF SIMULATED VAVE CLIMATE ................ . 47-51 V.1 Introduction.............................................. 47 V.2 Validation using visually observed wave data .............. 47 V.3 Validation using measured wave data ...................... . 50 C~A~i[~ VI. SIMULATED VAVE CLIMATE FOR THE INDIAN SEAS ......... . 52-67 VI.1 Introduction ............................................ . 52 VI.2 Spatial distribution of wave parameters .................. 52 VI. 2.1 nonthly distribution.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 VI.2.2 Seasonal distribution ............................... 60 VI. 2. 3 Annual distribution................................. 61 iv VI.3 Statistical distribution of wave height and period ....... 62 VI.3.1 Bivariate distribution .............................. 62 VI. 3. 2 Cumulative distribution............................. 62 VI.4 Spectral characteristics ................................. 63 VI.5 A comparative study of wave climate ...................... 63 VI.5.1 Arabian Sea and Bay of Bengal....................... 64 VI.5.2 East and West coast of India ........................ 64 VI.6 Limitations of simulated wave climate .................... 65 VI. 7 Future outlook 66 VII. SUMMARY AND CONCLUSIONS ........................... . 68-71 C~A~i[~ A~[IN[) 0} X -A 3g-W'AM FLOW' CHARTS ................................... . 72-77 3g-W'AM PROGRAMS, SUBROUTINES AND FUNCTIONS ........... . 78-82 A~[IN[)O}X-B A~[IN[)O}X-C INCLUDE FILES (PARAMETERS FOR ARRAY DIMENSITIONS) .... . 83-85 AnlN[)O}x-D USER INPUT FILES ..................................... . 86-98 IR[IF[IR[INC[S 99-106 WAW[S ............................. 107-109 A~i~~IR'S ~~~lOCAiO~INS ~IN ~C[AIN LI ST OF FI GURES Next to Page Fig. 1 Location map for Indian Seas. 11 Fig.2A Wind data distribution for Indian Seas (IMD, 1961-70). 30 Fig.2B Wind data distribution for Indian Seas (H&L, 1911-70). 31 Fig.2C Surface current data distribution for Indian Seas 31 (1954-94). Fig.3A Mean surface wind fields for Indian Seas (January to April). 35 Fig.3B Same as Fig.3A for May to August. 35 Fig.3C Same as Fig.3A for September to December. 35 Fig.4A Observed joint probability distribution for u and v 35 components of winds (January to March). Fig.4B Same as Fig.4A for April to June. 35 Fig.4C Same as Fig.4A for July to September. 35 Fig.4D Same as Fig.4A for October to December. 35 Fig.5A Mean surface current fields for Indian Seas (January to 36 April. Fig.5B Same as Fig.5A for May to August. 36 Fig.5C Same as Fig.5A for September to December. 36 Fig.6 Observed joint probability density for u and v components 39 of wind during July. Fig.1 Schematic representation for wind estimation from observed 39 joint probability distribution of u and v components. ° ° Fig.8A The mean climatic year of winds (67.5 E, 12.5 ). 40 0 Fig.8B Same as Fig.8A (87.5 E, 12.5°). 40 Fig.9 Model grid system for Indian Seas. 43 Fig. 10 Evolution of wave spectrum for 72 hours of model run using 46 estimated mean climatic wind fields for the month of July. vi Fig.llA Visually observed windsea data distribution for Indian 47 Seas (1961-70). Fig.llB Visually observed swell data distribution for Indian Seas 47 (1961-70). Fig. 12 Visually observed significant wave height versus swell wave 48 height in the Indian Seas. Fig.13A Comparison between simulated and visually observed Hs for 49 Arabian Sea and Bay of Bengal during rough weather and fair weather seasons respectively. ......................... ... . . . Fig.13B Same as Fig. 13A for Ts. 49 ... ... . . ...... . .. . Fig.14A Same as Fig. 13A for Hsw. 49 Fig.14B Same as Fig. 13A for Tsw. ...... ..... . . ...... ... . . . 49 Fig.15A Measured wave data distribution for Indian Seas using 50 ship-borne wave recorder (1976-93, rough weather season). Fig.15B Same as Fig. 15A for fair weather season. 50 Fig.16A Comparison between simulated and measured (SBWR) Hs for 50 the Arabian Sea and Bay of Bengal during rough and fair weather seasons respectively. Fig.16B Same as Fig.16A for Ts. 50 Fig.17 Comparisons of results obtained from wave climate simulation 51 and GEOSAT satellite mission for two selected sites. Fig.18A [a] The significant wave height, [b] Significant wave period 53 and windsea direction (January). Fig.18B [ a] The swell wave height, [b] swell wave period and 53 direction (January) . ............................ Fig.19A Same as Fig.18A for February. 53 ............................ Fig.19B Same as Fig.18B for February. 53 ............................... Fig.20A Same as Fig.18A for March. 53 ............................... Fig.20B Same as Fig.18B for March. 53 ............................... Fig.21A Same as Fig.18A for Apri 1. 53 ............................... Fig.21B Same as Fig.18B for Apri 1. 53 ................................. Fig.22A Same as Fig.18A for May. 53 vii Fig.22B Same as Fig. 18B for May. 53 Fig.23A Same as Fig.18A for June. 53 Fig.23B Same as Fig.18B for June. 53 Fig.24A Same as Fig.18A for July. 53 Fig.24B Same as Fig.18B for July. 53 Fig.25A Same as Fig.18A for August. 53 Fig.25B Same as Fig.18B for August. 53 Fig.26A Same as Fig.18A for September. 53 Fig.26B Same as Fig.18B for September. 53 Fig.27A Same as Fig.18A for October. 53 Fig.27B Same as Fig.18B for October. 53 Fig.28A Same as Fig.18A for November. 53 Fig.28B Same as Fig.18B for November. 53 Fig.29A Same as Fig. 18A for December. 53 Fig.29B Same as Fig.18B for December. 53 Fig. 30 Same as Fig.18A for May-September. 61 Fig. 31 Same as Fig.18A for October-April. 61 Fig. 32 Same as Fig.18A for January-December. 61 Fig. 33 Bivariate distribution for Hs and Is for the Arabian Sea 62 and Bay of Bengal during rough weather and fair weather seasons. Fig.34A Cumulative distribution of Hs for the Arabian Sea and Bay 62 of Bengal during rough weather and fair weather seasons. Fig.34B Same as Fig.34A for Is. 62 Fig.35A Average wave spectrum for a selected site in the Arabian 63 Sea during rough weather season. Fig.35B Same as Fig.35A during fair weather season. 63 viii