Extraction of wind and temperature information from hybrid 4D-Var assimilation of stratospheric ozone using NAVGEM

Abstract

Extraction of wind and temperature information from stratospheric ozone assimilation is examined within the context of the Navy Global Environmental Model (NAVGEM) hybrid 4-D variational assimilation (4D-Var) data assimilation (DA) system. Ozone can improve the wind and temperature through two different DA mechanisms: (1) through the "flow-of-the-day" ensemble background error covariance that is blended together with the static background error covariance and (2) via the ozone continuity equation in the tangent linear model and adjoint used for minimizing the cost function. All experiments assimilate actual conventional data in order to maintain a similar realistic troposphere. In the stratosphere, the experiments assimilate simulated ozone and/or radiance observations in various combinations. The simulated observations are constructed for a case study based on a 16-day cycling truth experiment (TE), which is an analysis with no stratospheric observations. The impact of ozone on the analysis is evaluated by comparing the experiments to the TE for the last 6 days, allowing for a 10-day spin-up. Ozone assimilation benefits the wind and temperature when data are of sufficient quality and frequency. For example, assimilation of perfect (no applied error) global hourly ozone data constrains the stratospheric wind and temperature to within ∼ 2 m s −1 and ∼ 1 K. This demonstrates that there is dynamical information in the ozone distribution that can potentially be used to improve the stratosphere. This is particularly important for the tropics, where radiance observations have difficulty constraining wind due to breakdown of geostrophic balance. Global ozone assimilation provides the largest benefit when the hybrid blending coefficient is an intermediate value (0.5 was used in this study), rather than 0.0 (no ensemble background error covariance) or 1.0 (no static background er-ror covariance), which is consistent with other hybrid DA studies. When perfect global ozone is assimilated in addition to radiance observations, wind and temperature error decreases of up to ∼ 3 m s −1 and ∼ 1 K occur in the tropical upper stratosphere. Assimilation of noisy global ozone (2 % errors applied) results in error reductions of ∼ 1 m s −1 and ∼ 0.5 K in the tropics and slightly increased temperature errors in the Northern Hemisphere polar region. Reduction of the ozone sampling frequency also reduces the benefit of ozone throughout the stratosphere, with noisy polarorbiting data having only minor impacts on wind and temperature when assimilated with radiances. An examination of ensemble cross-correlations between ozone and other variables shows that a single ozone observation behaves like a potential vorticity (PV) "charge", or a monopole of PV, with rotation about a vertical axis and vertically oriented temperature dipole. Further understanding of this relationship may help in designing observation systems that would optimize the impact of ozone on the dynamics.

High-Altitude (0–100 km) Global Atmospheric Reanalysis System: Description and Application to the 2014 Austral Winter of the Deep Propagating Gravity Wave Experiment (DEEPWAVE)

“…4a and 4b, respectively. The matrix C static specifies various horizontal and vertical correlations among variables, complete descriptions of which can be found in sections 3.8 and 4 and appendix B of Daley and Barker (2001a) and in section 2.2 of Allen et al (2014). Figure 4e shows the vertical correlation of geopotential (temperature) errors, the width of which increases with height, consistent with a similar broadening of model layers with height in Fig.…”

“…The current NRL Atmospheric Variational DAS (NAVDAS) is based around a four-dimensional variational (4DVAR) algorithm solved in observation space using an accelerated representer (AR) method. An overview of NAVDAS-AR relevant to the 0-100-km NAVGEM is provided here; more complete descriptions of specific aspects are provided elsewhere (see, e.g., Daley and Barker 2001a;Xu et al 2005;Kuhl et al 2013;Allen et al 2014).…”

Use of variable ozone in a radiative transfer model for the global Météo‐France 4D‐Var system

“…To summarize, for the two periods considered, we observe that means of analysis differences between the two experiments range from ±2.0 K in the upper stratosphere, ±0.1 K in the lower stratosphere and ±0.25 K F I G U R E 6 Zonal mean cross-section of (a) temperature, (b) relative humidity and (c) zonal wind analysis differences between O3EXP and CONTROL from 12 July to 10 September 2016 F I G U R E 7 As Figure 6, but from 12 November to 31 December 2016 in troposphere for the temperature, ±3.0% in the lower troposphere and ±1.0% in UTLS for the relative humidity and ±2.0 m⋅s −1 in the upper stratosphere, ±0.2 m⋅s −1 in the lower stratosphere and ±0.1 m⋅s −1 in troposphere for the zonal wind component. It is interesting to note that the patterns and observed areas of differences in temperature and zonal wind analyses are similar to those presented in the work by Allen et al (2018) on the use of ozone for temperature and wind information extraction.…”

“…It is interesting to note that the patterns and observed areas of differences in temperature and zonal wind analyses are similar to those presented in the work by Allen et al . (2018) on the use of ozone for temperature and wind information extraction.…”

100 Years of Progress in Forecasting and NWP Applications

“…The retrieval of satellite data and direct assimilation of radiances will play a critical role in this regard. Encouraging results are emerging, such as the positive effect of ozone assimilation on the wind fields (e.g., Allen et al 2018).…”