5:15 PM - 6:45 PM
[AAS10-P04] Long-term variations in atmospheric tides
Keywords:Atmospheric Tide, Long-term variation, Reanalysis, Earth System Model
Atmospheric thermal tides are daily fluctuations of the atmosphere driven by the periodic heating of the atmosphere,. Their 24-hr and 12-hr harmonic components are known as diurnal tides, semidiurnal tides, respectively. Tides have been studied for many years and the observational characteristics of long-term averages and seasonal variations are now well known. By contrast, long-term (annual to multi-decadal) variations have been less studied except for the variations associated with quasi-biennial oscillation (QBO), primarily due to the lack of long-term, diurnally resolving datasets. In this study, we investigate long-term variability of barometric tides using latest data from a latest reanalysis and an Earth System Model.
Surface pressure data from ERA5 reanalysis data and long-term integrated data from the Meteorological Research Institute Earth System Model (MRI-ESM v2.0) were used in the analysis. For the latter, two types of data were used: an experiment that reproduced the realistic Earth’s history (hereafter referred to as Historical), and an experiment in which only the amount of solar radiation was varied and the factors involving humans, such as the amount of carbon dioxide, were fixed from before the industrial revolution (hereafter referred to as CO2fixed). Monthly mean tidal components were calculated using 68 years of hourly surface pressure data from 1950 to 2017 for ERA5, 165 years of 3-hourly surface pressure data from 1850 to 2014 for Historical, and 171 years of 3-hourly surface pressure data from 1850 to 2020 for CO2fixed. In this study, we focus on the sun-synchronous component that is moving with the sun in the longitude direction (westward wavenumber 1 for the diuenal tide and westward wavenumber 2 for the semidiurnal tide). Since both diurnal and semidiurnal tides have large amplitudes in the tropics, the following discussion is based on components averaged over the tropics (10°S to 10°N).
We first examined the long-term trend and found that the ERA5 results for both the semidiurnal and diurnal tides differed significantly from the model results. It is known that various trends seen in reanalysis data are easily influenced by qualitative and quantitative changes in the assimilated observational data, and so further study(e.g., the analysis of pure barometric data) would be necessary. On the other hand, when comparing two model experiments (Historical and CO2fixed), it was found that the amplitude of semidiurnal tides in Historical run shows an increasing trend from around 1940 compared to CO2fixed run, whereas that trend in diurnal tides does not. This suggests that the amplitude of semidiurnal tides is easily affected by increases in greenhouse gases. Specifically, two factors may be possible: an increase in solar radiation absorption in the troposphere (water vapor heating) as temperature rises and/or a change in the temperature structure of the atmosphere. Note that if the former is the cause, an upward trend would also be seen in the amplitude of diurnal tide, whose main source is water vapor heating. It is thus inferred that changes in the temperature structure of the atmosphere may be more effective.
We also investigated the annual fluctuations associated with QBO. We detected signals not only in the semidiurnal tides, which was reported in many previous stuides, but also in the diurnal tides, although these were only based on model results.
In the future, we will investigate the reliability of the above results by analyzing other reanalysis data, model data, and even actual observation data, and clarify the physical process behind the above results using the framework of classical tidal theory.
Surface pressure data from ERA5 reanalysis data and long-term integrated data from the Meteorological Research Institute Earth System Model (MRI-ESM v2.0) were used in the analysis. For the latter, two types of data were used: an experiment that reproduced the realistic Earth’s history (hereafter referred to as Historical), and an experiment in which only the amount of solar radiation was varied and the factors involving humans, such as the amount of carbon dioxide, were fixed from before the industrial revolution (hereafter referred to as CO2fixed). Monthly mean tidal components were calculated using 68 years of hourly surface pressure data from 1950 to 2017 for ERA5, 165 years of 3-hourly surface pressure data from 1850 to 2014 for Historical, and 171 years of 3-hourly surface pressure data from 1850 to 2020 for CO2fixed. In this study, we focus on the sun-synchronous component that is moving with the sun in the longitude direction (westward wavenumber 1 for the diuenal tide and westward wavenumber 2 for the semidiurnal tide). Since both diurnal and semidiurnal tides have large amplitudes in the tropics, the following discussion is based on components averaged over the tropics (10°S to 10°N).
We first examined the long-term trend and found that the ERA5 results for both the semidiurnal and diurnal tides differed significantly from the model results. It is known that various trends seen in reanalysis data are easily influenced by qualitative and quantitative changes in the assimilated observational data, and so further study(e.g., the analysis of pure barometric data) would be necessary. On the other hand, when comparing two model experiments (Historical and CO2fixed), it was found that the amplitude of semidiurnal tides in Historical run shows an increasing trend from around 1940 compared to CO2fixed run, whereas that trend in diurnal tides does not. This suggests that the amplitude of semidiurnal tides is easily affected by increases in greenhouse gases. Specifically, two factors may be possible: an increase in solar radiation absorption in the troposphere (water vapor heating) as temperature rises and/or a change in the temperature structure of the atmosphere. Note that if the former is the cause, an upward trend would also be seen in the amplitude of diurnal tide, whose main source is water vapor heating. It is thus inferred that changes in the temperature structure of the atmosphere may be more effective.
We also investigated the annual fluctuations associated with QBO. We detected signals not only in the semidiurnal tides, which was reported in many previous stuides, but also in the diurnal tides, although these were only based on model results.
In the future, we will investigate the reliability of the above results by analyzing other reanalysis data, model data, and even actual observation data, and clarify the physical process behind the above results using the framework of classical tidal theory.