11:00 AM - 11:15 AM
[MGI29-08] Wave assimilation system of JMA and its improvement plan
★Invited Papers
Keywords:data assimilation, waves, analysis
JMA developed a wave analysis and assimilation system which is operationally used for years. The system consists of wave analysis and wave assimilation component. We will introduce the system and our future plans.
In recent decades, orbital satellites become a powerful tool for wave observation, and observed data, which is mainly significant wave heights, are used in assimilation systems in many National Meteorological Services (NMS). In JMA, an Objective Wave Analysis System (OWAS) was developed at first. The system analyzes wave heights with Optimal Interpolation (OI), by referring wave heights observed by satellites, buoys, coastal recorders, and ships reports. The analysis is carried out in global region except the polar regions with 0.5-degree resolution. The results are issued as wave analysis charts.
A wave assimilation system followed to be developed, by using the wave analysis results. In wave models, wave height is not sufficient information because wave spectrum is the basic physical value, and more detailed information on each spectrum are required. Therefore, some special treatment for modification is used in wave assimilation. In the JMA wave assimilation scheme, wave spectra are divided to wind-sea and swell by a partitioning method and both values are modified in different way. If swell is predominant, spectra are modified equally as the rate of energy of first guess (model original) and analysis. When wind-sea is predominant, the spectra are modified with Toba’s power law, and residual swell spectra are modified so that the total wave energy will same as analysis.
In general, wave assimilation impact does not keep to later forecast, but our system shows some good impact till later time although it is not so much. The merit of wave assimilation would be rather in analysis and early forecasts, and it will be useful especially for reanalysis. Since current system use simple OI, we have a plan to improve assimilation scheme, such as introducing a three-dimensional variational method (3D-Var), which will give more reasonable wave fields. However, considering of assimilation impact and resource, more sophisticated scheme like a four-dimensional variational method (4D-Var) will not be necessary for operational use. Recently satellite wave spectrum observations become available, and assimilation of wave spectra is also introduced in some NMS. The development of a wave spectrum assimilation scheme is also our future plan.
In recent decades, orbital satellites become a powerful tool for wave observation, and observed data, which is mainly significant wave heights, are used in assimilation systems in many National Meteorological Services (NMS). In JMA, an Objective Wave Analysis System (OWAS) was developed at first. The system analyzes wave heights with Optimal Interpolation (OI), by referring wave heights observed by satellites, buoys, coastal recorders, and ships reports. The analysis is carried out in global region except the polar regions with 0.5-degree resolution. The results are issued as wave analysis charts.
A wave assimilation system followed to be developed, by using the wave analysis results. In wave models, wave height is not sufficient information because wave spectrum is the basic physical value, and more detailed information on each spectrum are required. Therefore, some special treatment for modification is used in wave assimilation. In the JMA wave assimilation scheme, wave spectra are divided to wind-sea and swell by a partitioning method and both values are modified in different way. If swell is predominant, spectra are modified equally as the rate of energy of first guess (model original) and analysis. When wind-sea is predominant, the spectra are modified with Toba’s power law, and residual swell spectra are modified so that the total wave energy will same as analysis.
In general, wave assimilation impact does not keep to later forecast, but our system shows some good impact till later time although it is not so much. The merit of wave assimilation would be rather in analysis and early forecasts, and it will be useful especially for reanalysis. Since current system use simple OI, we have a plan to improve assimilation scheme, such as introducing a three-dimensional variational method (3D-Var), which will give more reasonable wave fields. However, considering of assimilation impact and resource, more sophisticated scheme like a four-dimensional variational method (4D-Var) will not be necessary for operational use. Recently satellite wave spectrum observations become available, and assimilation of wave spectra is also introduced in some NMS. The development of a wave spectrum assimilation scheme is also our future plan.