The Skill Assessment of Weather and Research Forecasting and WAVEWATCH-III Models During Recent Meteotsunami Event in the Persian Gulf

Mohsen Rahimian,Mostafa Beyramzadeh,Seyed Mostafa Siadatmousavi

doi:10.3389/fmars.2022.834151

Mohsen Rahimian, Mostafa Beyramzadeh + Show 1 more

Open Access

https://doi.org/10.3389/fmars.2022.834151

Copy DOI

Abstract

This study aims to use a fully realistic high-resolution mesoscale atmospheric and wave model to reproduce met-ocean conditions during a meteotsunami in the Persian Gulf. The atmospheric simulations were performed with the Weather and Research Forecasting (WRF) model by varying planetary boundary layer, microphysics, cumulus, and radiations parameterizations. The atmospheric results were compared to the meteorological observations (e.g., air pressure and wind speed) from the coastal and island synoptic and buoy stations of the nearest area to the meteotsunami event. The results show that using Mellor-Yamada-Nakanishi-Niino (MYNN) scheme for planetary boundary and surface layer had the best performance for stations over the water, whereas applying Mellor-Yamada-Janjic scheme for planetary boundary and Eta similarity surface layer had the best performance for stations over the land. For wave simulations, the WAVEWATCH-III model was employed with the well-known WAM-Cycle4 formulation and a more recent ST6 package. Six WRF experiments and ERA5 wind data were used to force the wave models. The new error parameter was introduced to identify the optimum wind data for wave simulation. EXP4 configuration which uses the MYNN scheme for planetary boundary and surface layer was led to minimum error, while ERA5 severely underestimated Hs and Tp parameters. For the first time, the Gaussian Quadrature Method (GQM) was implemented in the WAVEWATCH-III model and combined with a depth scale to be used in the Persian Gulf. This method is more accurate for non-linear wave-wave interaction than the default Discrete Interaction Approximation (DIA) method. Lower coefficients for dissipation term were required for GQM and the resulted bulk wave parameters were improved compared to the DIA method. The calibrated ST6 formulation with GQM resulted in a more realistic prediction of wave spectrum than the default settings of the WAVEWATCH-III.

Highlights

The Persian Gulf is one of the most important oil tanker highways in the world, which has been protected from waves induced by tropical storms and tsunamis over the past few decades (ElSabh and Murty, 1989; Al-Hajri, 1990; Lin and Emanuel, 2016)
This study aims to assess the performance of different parameterizations for physical processes in a high-resolution numerical model in simulating the meteorological characteristics and wind-induced waves during the dominance of meteotsunami in the Persian Gulf
This study skill assessed the performance of the Weather and Research Forecasting (WRF) physics ensemble of a high-resolution modeling system in retrieving atmospheric processes which led to recent meteotsunami in the Persian Gulf

Summary

Introduction

The Persian Gulf is one of the most important oil tanker highways in the world, which has been protected from waves induced by tropical storms and tsunamis over the past few decades (ElSabh and Murty, 1989; Al-Hajri, 1990; Lin and Emanuel, 2016). Tsunami waves produced in the Indian Ocean rarely propagate into the Persian Gulf (Rabinovich and Thomson, 2007; Heidarzadeh et al, 2008); harsh weather is not common in this region (Nadim et al, 2008; Modarress et al, 2012). Salaree et al (2018) conducted a field study on the damaged coastline and reconstructed the initial picture of the whole event. They explained the physical mechanisms generating the strong long waves during this event and concluded that a meteorological tsunami was responsible for this event. The rainfall intensity, maximum reflection, and echo top height images provided by the weather radar confirmed that a strong convergent system, including the middle and upper troposphere, had entered the northern Persian Gulf approximately 4 h before the event and moved to the east (Kazeminezhad et al, 2021). 2 h before its landfall, the convection system deformed into a narrow and long hurricane with 70–130 km length, less than 10 km width, and a transverse speed of 24 m/s

Objectives

Methods

Results

Conclusion