Initialization of an ENSO Forecast System Using a Parallelized Ensemble Filter

2005 ◽  
Vol 133 (11) ◽  
pp. 3176-3201 ◽  
Author(s):  
S. Zhang ◽  
M. J. Harrison ◽  
A. T. Wittenberg ◽  
A. Rosati ◽  
J. L. Anderson ◽  
...  
Keyword(s):  
Data Assimilation ◽  
Hybrid Model ◽  
Sequential Analysis ◽  
Surface Wind ◽  
Ocean Model ◽  
Consistent Estimate ◽  
State Variables ◽  
Test Bed ◽  
Background Error ◽  
Temperature State

Abstract As a first step toward coupled ocean–atmosphere data assimilation, a parallelized ensemble filter is implemented in a new stochastic hybrid coupled model. The model consists of a global version of the GFDL Modular Ocean Model Version 4 (MOM4), coupled to a statistical atmosphere based on a regression of National Centers for Environmental Prediction (NCEP) reanalysis surface wind stress, heat, and water flux anomalies onto analyzed tropical Pacific SST anomalies from 1979 to 2002. The residual part of the NCEP fluxes not captured by the regression is then treated as stochastic forcing, with different ensemble members feeling the residual fluxes from different years. The model provides a convenient test bed for coupled data assimilation, as well as a prototype for representing uncertainties in the surface forcing. A parallel ensemble adjustment Kalman filter (EAKF) has been designed and implemented in the hybrid model, using a local least squares framework. Comparison experiments demonstrate that the massively parallel processing EAKF (MPPEAKF) produces assimilation results with essentially the same quality as a global sequential analysis. Observed subsurface temperature profiles from expendable bathythermographs (XBTs), Tropical Atmosphere Ocean (TAO) buoys, and Argo floats, along with analyzed SSTs from NCEP, are assimilated into the hybrid model over 1980–2002 using the MPPEAKF. The filtered ensemble of SSTs, ocean heat contents, and thermal structures converge well to the observations, in spite of the imposed stochastic forcings. Several facets of the EAKF algorithm used here have been designed to facilitate comparison to a traditional three-dimensional variational data assimilation (3DVAR) algorithm, for instance, the use of a univariate filter in which observations of temperature only directly impact temperature state variables. Despite these choices that may limit the power of the EAKF, the MPPEAKF solution appears to improve upon an earlier 3DVAR solution, producing a smoother, more physically reasonable analysis that better fits the observational data and produces, to some degree, a self-consistent estimate of analysis uncertainties. Hybrid model ENSO forecasts initialized from the MPPEAKF ensemble mean also appear to outperform those initialized from the 3DVAR analysis. This improvement stems from the EAKF’s utilization of anisotropic background error covariances that may vary in time.

scholarly journals Development of a Hybrid En3DVar Data Assimilation System and Comparisons with 3DVar and EnKF for Radar Data Assimilation with Observing System Simulation Experiments

2018 ◽  
Vol 146 (1) ◽  
pp. 175-198 ◽  
Author(s):  
Rong Kong ◽  
Ming Xue ◽  
Chengsi Liu
Keyword(s):  
Data Assimilation ◽  
System Simulation ◽  
Radar Data ◽  
Mixing Ratio ◽  
State Variables ◽  
Ensemble Size ◽  
Simulation Experiments ◽  
Background Error ◽  
Variational Minimization ◽  
Water Vapor Mixing Ratio

Abstract A hybrid ensemble–3DVar (En3DVar) system is developed and compared with 3DVar, EnKF, “deterministic forecast” EnKF (DfEnKF), and pure En3DVar for assimilating radar data through perfect-model observing system simulation experiments (OSSEs). DfEnKF uses a deterministic forecast as the background and is therefore parallel to pure En3DVar. Different results are found between DfEnKF and pure En3DVar: 1) the serial versus global nature and 2) the variational minimization versus direct filter updating nature of the two algorithms are identified as the main causes for the differences. For 3DVar (EnKF/DfEnKF and En3DVar), optimal decorrelation scales (localization radii) for static (ensemble) background error covariances are obtained and used in hybrid En3DVar. The sensitivity of hybrid En3DVar to covariance weights and ensemble size is examined. On average, when ensemble size is 20 or larger, a 5%–10% static covariance gives the best results, while for smaller ensembles, more static covariance is beneficial. Using an ensemble size of 40, EnKF and DfEnKF perform similarly, and both are better than pure and hybrid En3DVar overall. Using 5% static error covariance, hybrid En3DVar outperforms pure En3DVar for most state variables but underperforms for hydrometeor variables, and the improvement (degradation) is most notable for water vapor mixing ratio qυ (snow mixing ratio qs). Overall, EnKF/DfEnKF performs the best, 3DVar performs the worst, and static covariance only helps slightly via hybrid En3DVar.

scholarly journals Stochastic Integration for the Heterogeneous Correlation Modeling Using a Diffusion Equation

2010 ◽  
Vol 138 (8) ◽  
pp. 3356-3365 ◽  
Author(s):  
Olivier Pannekoucke ◽  
Laurent Vezard
Keyword(s):  
Data Assimilation ◽  
Nearest Neighbor ◽  
Integration Scheme ◽  
Square Root ◽  
Stochastic Integration ◽  
Test Bed ◽  
Background Error ◽  
Diffusion Operator ◽  
Background Error Covariance ◽  
Error Covariance

Abstract In this note, a stochastic integration scheme is proposed as an alternative to a deterministic integration scheme, usually employed for the diffusion operator in data assimilation. The stochastic integration scheme is no more than a simple interpolation of the initial condition in lieu of the deterministic integration. Furthermore, this also presents a potential in high performance computing. For the classic preconditioned minimizing problem, the stochastic integration is employed to implement the square root of the background error covariance matrix, while its adjoint is obtained from the adjoint code of the square root code. In a first part, the stochastic integration method and its weak convergence are detailed. Then the practical use of this approach in data assimilation is described. It is illustrated in a 1D test bed, where it is shown to run smoothly for background error covariance modeling, with nearest-neighbor interpolations, and O(100) particles.

scholarly journals Do surface lateral flows matter for data assimilation of soil moisture observations into hyperresolution land models?

2020 ◽  
Author(s):  
Yohei Sawada
Keyword(s):  
Soil Moisture ◽  
Kalman Filter ◽  
Data Assimilation ◽  
Ensemble Kalman Filter ◽  
Land Surface ◽  
Pressure Head ◽  
Surface Flow ◽  
State Variables ◽  
Background Error ◽  
Water Flows

Abstract. It is expected that hyperresolution land modeling substantially innovates the simulation of terrestrial water, energy, and carbon cycles. The major advantage of hyperresolution land models against conventional one-dimensional land surface models is that hyperresolution land models can explicitly simulate lateral water flows. Despite many efforts on data assimilation of hydrological observations into those hyperresolution land models, how surface water flows driven by local topography matter for data assimilation of soil moisture observations has not been fully clarified. Here I perform two minimalist synthetic experiments where soil moisture observations are assimilated into an integrated surface-groundwater land model by an ensemble Kalman filter. I discuss how differently the ensemble Kalman filter works when surface lateral flows are switched on and off. A horizontal background error covariance provided by overland flows is important to adjust the unobserved state variables (pressure head and soil moisture) and parameters (saturated hydraulic conductivity). However, the non-Gaussianity of the background error provided by the nonlinearity of a topography-driven surface flow harms the performance of data assimilation. It is difficult to efficiently constrain model states at the edge of the area where the topography-driven surface flow reaches by linear-Gaussian filters. It brings the new challenge in land data assimilation for hyperresolution land models. This study highlights the importance of surface lateral flows in hydrological data assimilation.

scholarly journals A three-dimensional variational data assimilation system for multiple aerosol species with WRF/Chem and an application to PM<sub>2.5</sub> prediction

2012 ◽  
Vol 12 (5) ◽  
pp. 13515-13552 ◽  
Author(s):  
Z. Li ◽  
Z. Zang ◽  
Q. B. Li ◽  
Y. Chao ◽  
D. Chen ◽  
...  
Keyword(s):  
Los Angeles ◽  
Data Assimilation ◽  
Three Dimensional ◽  
Variational Data Assimilation ◽  
Metal Species ◽  
Individual Species ◽  
Size Distributions ◽  
State Variables ◽  
Background Error ◽  
Mass Concentrations

Abstract. A three-dimensional variational data assimilation (3-DVAR) algorithm for aerosols in a WRF/Chem model is presented. The WRF/Chem model uses the MOSAIC (Model for Simulating Aerosol Interactions and Chemistry) scheme, which explicitly treats eight major species (elemental/black carbon, organic carbon, nitrate, sulfate, chloride, ammonium, sodium, and the sum of other inorganic, inert mineral and metal species) and represents size distributions using a sectional method with four size bins. The 3-DVAR scheme is formulated to take advantage of the MOSAIC scheme in providing comprehensive analyses of specie concentrations and size distributions. To treat the large number of state variables associated with the MOSAIC scheme, this 3-DVAR algorithm first determines the analysis increments of the total mass concentrations of the eight species, defined as the sum of the mass concentrations across all size bins, and then distributes the analysis increments over four size bins according to the background error variances. The number concentrations for each size bin are adjusted based on the ratios between the mass and number concentrations of the background state. This system has been applied to the analysis and prediction of PM2.5 in the Los Angeles basin during the CalNex 2010 field experiment, with assimilation of surface PM2.5 and speciated concentration observations. The results demonstrate that the data assimilation significantly reduces the errors in comparison with a down scaling simulation and improved forecasts of the concentrations of PM2.5 as well as individual species for up to 24 h. Some implementation difficulties and limitations of the system are also discussed.

scholarly journals Do surface lateral flows matter for data assimilation of soil moisture observations into hyperresolution land models?

2020 ◽  
Vol 24 (8) ◽  
pp. 3881-3898
Author(s):  
Yohei Sawada
Keyword(s):  
Soil Moisture ◽  
Kalman Filter ◽  
Data Assimilation ◽  
Ensemble Kalman Filter ◽  
Land Surface ◽  
Pressure Head ◽  
Surface Flow ◽  
State Variables ◽  
Background Error ◽  
Water Flows

Abstract. It is expected that hyperresolution land modeling substantially innovates the simulation of terrestrial water, energy, and carbon cycles. The major advantage of hyperresolution land models against conventional 1-D land surface models is that hyperresolution land models can explicitly simulate lateral water flows. Despite many efforts on data assimilation of hydrological observations into those hyperresolution land models, how surface water flows driven by local topography matter for data assimilation of soil moisture observations has not been fully clarified. Here I perform two minimalist synthetic experiments where soil moisture observations are assimilated into an integrated surface–groundwater land model by an ensemble Kalman filter. I discuss how differently the ensemble Kalman filter works when surface lateral flows are switched on and off. A horizontal background error covariance provided by overland flows is important for adjusting the unobserved state variables (pressure head and soil moisture) and parameters (saturated hydraulic conductivity). However, the non-Gaussianity of the background error provided by the nonlinearity of a topography-driven surface flow harms the performance of data assimilation. It is difficult to efficiently constrain model states at the edge of the area where the topography-driven surface flow reaches by linear-Gaussian filters. It brings the new challenge in land data assimilation for hyperresolution land models. This study highlights the importance of surface lateral flows in hydrological data assimilation.

scholarly journals Correlation-Cutoff Method for Covariance Localization in Strongly Coupled Data Assimilation

2018 ◽  
Vol 146 (9) ◽  
pp. 2881-2889 ◽  
Author(s):  
Takuma Yoshida ◽  
Eugenia Kalnay
Keyword(s):  
Kalman Filter ◽  
Data Assimilation ◽  
Coupled Model ◽  
Error Variance ◽  
State Variables ◽  
Background Error ◽  
Strongly Coupled ◽  
Error Correlation ◽  
Analysis Error ◽  
Coupled Data Assimilation

Abstract Strongly coupled data assimilation (SCDA), where observations of one component of a coupled model are allowed to directly impact the analysis of other components, sometimes fails to improve the analysis accuracy with an ensemble Kalman filter (EnKF) as compared with weakly coupled data assimilation (WCDA). It is well known that an observation’s area of influence should be localized in EnKFs since the assimilation of distant observations often degrades the analysis because of spurious correlations. This study derives a method to estimate the reduction of the analysis error variance by using estimates of the cross covariances between the background errors of the state variables in an idealized situation. It is shown that the reduction of analysis error variance is proportional to the squared background error correlation between the analyzed and observed variables. From this, the authors propose an offline method to systematically select which observations should be assimilated into which model state variable by cutting off the assimilation of observations when the squared background error correlation between the observed and analyzed variables is small. The proposed method is tested with the local ensemble transform Kalman filter (LETKF) and a nine-variable coupled model, in which three Lorenz models with different time scales are coupled with each other. The covariance localization with the correlation-cutoff method achieves an analysis more accurate than either the full SCDA or the WCDA methods, especially with smaller ensemble sizes.

A Flow-dependent Targeted Observation Method

2020 ◽  
Author(s):  
Youmin Tang ◽  
Yaling Wu
Keyword(s):  
Data Assimilation ◽  
Tropical Indian Ocean ◽  
Seasonal Prediction ◽  
Test Bed ◽  
Error Statistics ◽  
Background Error ◽  
Observation Method ◽  
Targeted Observation ◽  
Data Assimilation System ◽  
Assimilation System

&lt;p&gt;In this study, we developed a flow-dependent ensemble-based targeted observation method, by minimizing the analysis error variance under the framework of Ensemble Kalman filter (EnKF) data assimilation system. This method estimates the background error statistics as a flow dependent function. The covariance localization is also introduced for computing efficiency and alleviating the spurious observations.&amp;#160;&amp;#160;As a test bed, an &amp;#160;optimal observation array of sea level anomalies (SLA) is designed for its seasonal prediction over the tropical Indian Ocean (TIO) region. &amp;#160;Furthermore, the observing system simulation experiments (OSSEs) is used to verify the resultant optimal observational array using our recently developed coupled data assimilation system. A comparison between this flow-dependent method and the traditional method is also given. &amp;#8203;&lt;/p&gt;

scholarly journals Data assimilation of Argo profiles in a northwestern Pacific model

2017 ◽  
Vol 17 (1) ◽  
pp. 17-30 ◽  
Author(s):  
Zhaoyi Wang ◽  
Andrea Storto ◽  
Nadia Pinardi ◽  
Guimei Liu ◽  
Hui Wang
Keyword(s):  
Data Assimilation ◽  
Positive Impact ◽  
Three Dimensional ◽  
Predictive Performance ◽  
Ocean Model ◽  
Northwestern Pacific ◽  
Background Error ◽  
Salinity Data ◽  
Argo Profiles ◽  
Assimilation Scheme

Abstract. Based on a novel estimation of background-error covariances for assimilating Argo profiles, an oceanographic three-dimensional variational (3DVAR) data assimilation scheme was developed for the northwestern Pacific Ocean model (NwPM) for potential use in operational predictions and maritime safety applications. Temperature and salinity data extracted from Argo profiles from January to December 2010 were assimilated into the NwPM. The results show that the average daily temperature (salinity) root mean square error (RMSE) decreased from 0.99 °C (0.10 psu) to 0.62 °C (0.07 psu) in assimilation experiments throughout the northwestern Pacific, which represents a 37.2 % (27.6 %) reduction in the error. The temperature (salinity) RMSE decreased by  ∼  0.60 °C ( ∼  0.05 psu) for the upper 900 m (1000 m). Sea level, temperature and salinity were in better agreement with in situ and satellite datasets after data assimilation than before. In addition, a 1-month experiment with daily analysis cycles and 5-day forecasts explored the performance of the system in an operational configuration. The results highlighted the positive impact of the 3DVAR initialization at all forecast ranges compared to the non-assimilative experiment. Therefore, the 3DVAR scheme proposed here, coupled to ROMS, shows a good predictive performance and can be used as an assimilation scheme for operational forecasting.

scholarly journals A Comparison between 3DVAR and EnKF for Data Assimilation Effects on the Yellow Sea Fog Forecast

2018 ◽  
Author(s):  
Xiaoyu Gao ◽  
Shanhong Gao ◽  
Yue Yang
Keyword(s):  
Data Assimilation ◽  
Yellow Sea ◽  
Initial Conditions ◽  
Surface Wind ◽  
Probability Of Detection ◽  
The Other ◽  
Moisture Distribution ◽  
The Yellow Sea ◽  
Background Error ◽  
Sea Fog

The data assimilation method to improve sea fog forecast over the Yellow Sea is usually three-dimensional variational assimilation (3DVAR), whereas ensemble Kalman filter (EnKF) has not yet been applied on this weather phenomenon. In this paper, two sea fog cases over the Yellow sea, one spread widely and the other spread narrowly along the coastal area, are studied in detail by a series of numerical experiments with 3DVAR and EnKF based on the Grid-point Statistical Interpolation (GSI) system and the Weather Research and Forecasting (WRF) model. The results show that the assimilation effect of EnKF outperforms that of 3DVAR: for the widespread-fog case, the probability of detection and equitable threat scores of the forecasted sea fog area get improved respectively by ~57.9% and ~55.5%; the sea fog of the other case completely mis-forecasted by 3DVAR is produced successfully by EnKF. These improvements of EnKF relative to 3DVAR are benefited from its flow-dependent background error, resulting in more realistic depiction of sea surface wind for the widespread-fog case and better moisture distribution for the other case in the initial conditions. More importantly, the correlation between temperature and humidity in the background error of EnKF plays a vital role in the response of moisture to the assimilation of temperature, which leads to a great improvement on the initial moisture conditions for sea fog forecast.

Sign in / Sign up

Export Citation Format

Share Document