Multiscale Data Assimilation Research Articles

Potential of an EnKF Storm-Scale Data Assimilation System Over Sparse Observation Regions with Complex Orography

High-impact weather events over sparse data regions with complex orography, such as the Mediterranean region, remain a challenge for numerical weather prediction. This study evaluates, for the first time, the ability of a multiscale ensemble-based data assimilation system to reproduce a heavy precipitation episode that occurred during the first Special Observation Period (SOP1) of the Hydrological cycle in the Mediterranean Experiment (HyMeX). During the Intense Observation Period (IOP13) from 14 to 15 October 2012, convective maritime activity associated with an advancing cold front affected coastal areas of southern France, Corsica and Italy. With the main objective of improving forecasts of this weather event, a data assimilation (DA) system using the Ensemble Kalman Filter (EnKF) algorithm is implemented. The potential impact of assimilating conventional in-situ observations (METAR, aircrafts, buoys and rawinsondes) and single-Doppler reflectivity data to improve numerical representation of growing convective maritime structures that will evolve towards coastal populated areas is evaluated. Results indicate that information provided by both observation sources contribute to initiation and subsequent evolution of convective structures not captured by the conventional runs. Notably, data assimilation experiments produce the best quantitative verification scores for the short range (6–8 h) forecasts of accumulated precipitation. Beyond 6–8 h, data assimilation experiments and those without data assimilation are indistinguishable. Sensitivity experiments, evaluating the impact of increasing the length of the radar data assimilation period, reveal the importance of assimilating high-frequency reflectivity data during a mid-term period (6 h approx.) to better depict deep convective structures initiated over the sea that evolve towards populated coastal areas.

Conditional Gaussian Systems for Multiscale Nonlinear Stochastic Systems: Prediction, State Estimation and Uncertainty Quantification.

A conditional Gaussian framework for understanding and predicting complex multiscale nonlinear stochastic systems is developed. Despite the conditional Gaussianity, such systems are nevertheless highly nonlinear and are able to capture the non-Gaussian features of nature. The special structure of the system allows closed analytical formulae for solving the conditional statistics and is thus computationally efficient. A rich gallery of examples of conditional Gaussian systems are illustrated here, which includes data-driven physics-constrained nonlinear stochastic models, stochastically coupled reaction–diffusion models in neuroscience and ecology, and large-scale dynamical models in turbulence, fluids and geophysical flows. Making use of the conditional Gaussian structure, efficient statistically accurate algorithms involving a novel hybrid strategy for different subspaces, a judicious block decomposition and statistical symmetry are developed for solving the Fokker–Planck equation in large dimensions. The conditional Gaussian framework is also applied to develop extremely cheap multiscale data assimilation schemes, such as the stochastic superparameterization, which use particle filters to capture the non-Gaussian statistics on the large-scale part whose dimension is small whereas the statistics of the small-scale part are conditional Gaussian given the large-scale part. Other topics of the conditional Gaussian systems studied here include designing new parameter estimation schemes and understanding model errors.

Understanding the Impact of Radar and In Situ Observations on the Prediction of a Nocturnal Convection Initiation Event on 25 June 2013 Using an Ensemble-Based Multiscale Data Assimilation System

The initiation of new convection at night in the Great Plains contributes to a nocturnal maximum in precipitation and produces localized heavy rainfall and severe weather hazards in the region. Although previous work has evaluated numerical model forecasts and data assimilation (DA) impacts for convection initiation (CI), most previous studies focused only on convection that initiates during the afternoon and not explicitly on nocturnal thunderstorms. In this study, we investigate the impact of assimilating in situ and radar observations for a nocturnal CI event on 25 June 2013 using an ensemble-based DA and forecast system. Results in this study show that a successful CI forecast resulted only when assimilating conventional in situ observations on the inner, convection-allowing domain. Assimilating in situ observations strengthened preexisting convection in southwestern Kansas by enhancing buoyancy and locally strengthening low-level convergence. The enhanced convection produced a cold pool that, together with increased convergence along the northwestern low-level jet (LLJ) terminus near the region of CI, was an important mechanism for lifting parcels to their level of free convection. Gravity waves were also produced atop the cold pool that provided further elevated ascent. Assimilating radar observations further improved the forecast by suppressing spurious convection and reducing the number of ensemble members that produced CI along a spurious outflow boundary. The fact that the successful CI forecasts resulted only when the in situ observations were assimilated suggests that accurately capturing the preconvective environment and specific mesoscale features is especially important for nocturnal CI forecasts.

Formulations for Estimating Spatial Variations of Analysis Error Variance to Improve Multiscale and Multistep Variational Data Assimilation

When the coarse-resolution observations used in the first step of multiscale and multistep variational data assimilation become increasingly nonuniform and/or sparse, the error variance of the first-step analysis tends to have increasingly large spatial variations. However, the analysis error variance computed from the previously developed spectral formulations is constant and thus limited to represent only the spatially averaged error variance. To overcome this limitation, analytic formulations are constructed to efficiently estimate the spatial variation of analysis error variance and associated spatial variation in analysis error covariance. First, a suite of formulations is constructed to efficiently estimate the error variance reduction produced by analyzing the coarse-resolution observations in one- and two-dimensional spaces with increased complexity and generality (from uniformly distributed observations with periodic extension to nonuniformly distributed observations without periodic extension). Then, three different formulations are constructed for using the estimated analysis error variance to modify the analysis error covariance computed from the spectral formulations. The successively improved accuracies of these three formulations and their increasingly positive impacts on the two-step variational analysis (or multistep variational analysis in first two steps) are demonstrated by idealized experiments.

On sequential multiscale inversion and data assimilation

Multiscale approaches are very popular for example for solving partial differential equations and in many applied fields dealing with phenomena which take place on different levels of detail. The broad idea of a multiscale approach is to decompose your problem into different scales or levels and to use these decompositions either for constructing appropriate approximations or to solve smaller problems on each of these levels, leading to increased stability or increased efficiency. The idea of sequential multiscale is to first solve the problem in a large-scale subspace and then successively move to finer scale spaces. Our goal is to analyzethe sequential multiscale approach applied to an inversion or state estimation problem. We work in a generic setup given by a Hilbert space environment. We work out the analysis both for an unregularized and a regularized sequential multiscale inversion. In general the sequential multiscale approach is not equivalent to a full solution, but we show that under appropriate assumptions we obtain convergence of an iterative sequential multiscale version of the method. For the regularized case we develop a strategy to appropriately adapt the regularization when an iterative approach is taken. We demonstrate the validity of the iterative sequential multiscale approach by testing the method on an integral equation as it appears for atmospheric temperature retrieval from infrared satellite radiances.

State estimation and prediction using clustered particle filters

Particle filtering is an essential tool to improve uncertain model predictions by incorporating noisy observational data from complex systems including non-Gaussian features. A class of particle filters, clustered particle filters, is introduced for high-dimensional nonlinear systems, which uses relatively few particles compared with the standard particle filter. The clustered particle filter captures non-Gaussian features of the true signal, which are typical in complex nonlinear dynamical systems such as geophysical systems. The method is also robust in the difficult regime of high-quality sparse and infrequent observations. The key features of the clustered particle filtering are coarse-grained localization through the clustering of the state variables and particle adjustment to stabilize the method; each observation affects only neighbor state variables through clustering and particles are adjusted to prevent particle collapse due to high-quality observations. The clustered particle filter is tested for the 40-dimensional Lorenz 96 model with several dynamical regimes including strongly non-Gaussian statistics. The clustered particle filter shows robust skill in both achieving accurate filter results and capturing non-Gaussian statistics of the true signal. It is further extended to multiscale data assimilation, which provides the large-scale estimation by combining a cheap reduced-order forecast model and mixed observations of the large- and small-scale variables. This approach enables the use of a larger number of particles due to the computational savings in the forecast model. The multiscale clustered particle filter is tested for one-dimensional dispersive wave turbulence using a forecast model with model errors.

A Study of Multiscale Initial Condition Perturbation Methods for Convection-Permitting Ensemble Forecasts

AbstractThe impacts of multiscale flow-dependent initial condition (IC) perturbations for storm-scale ensemble forecasts of midlatitude convection are investigated using perfect-model observing system simulation experiments. Several diverse cases are used to quantitatively and qualitatively understand the impacts of different IC perturbations on ensemble forecast skill. Scale dependence of the results is assessed by evaluating 2-h storm-scale reflectivity forecasts separately from hourly accumulated mesoscale precipitation forecasts.Forecasts are initialized with different IC ensembles, including an ensemble of multiscale perturbations produced by a multiscale data assimilation system, mesoscale perturbations produced at a coarser resolution, and filtered multiscale perturbations. Mesoscale precipitation forecasts initialized with the multiscale perturbations are more skillful than the forecasts initialized with the mesoscale perturbations at several lead times. This multiscale advantage is due to greater consistency between the IC perturbations and IC uncertainty. This advantage also affects the short-term, smaller-scale forecasts. Reflectivity forecasts on very small scales and very short lead times are more skillful with the multiscale perturbations as a direct result of the smaller-scale IC perturbation energy. The small-scale IC perturbations also contribute to some improvements to the mesoscale precipitation forecasts after the ~5-h lead time. Altogether, these results suggest that the multiscale IC perturbations provided by ensemble data assimilation on the convection-permitting grid can improve storm-scale ensemble forecasts by improving the sampling of IC uncertainty, compared to downscaling of IC perturbations from a coarser-resolution ensemble.

Coarse-grained sensitivity for multiscale data assimilation.

We show that the effective average action and its gradient are useful for solving multiscale data assimilation problems. We also present a procedure for numerically evaluating the gradient of the effective average action and demonstrate that the variational problem for slow degrees of freedom can be solved properly using the effective gradient.

A Multiscale Variational Data Assimilation Scheme: Formulation and Illustration

Abstract A multiscale data assimilation (MS-DA) scheme is formulated for fine-resolution models. A decomposition of the cost function is derived for a set of distinct spatial scales. The decomposed cost function allows for the background error covariance to be estimated separately for the distinct spatial scales, and multi-decorrelation scales to be explicitly incorporated in the background error covariance. MS-DA minimizes the partitioned cost functions sequentially from large to small scales. The multi-decorrelation length scale background error covariance enhances the spreading of sparse observations and prevents fine structures in high-resolution observations from being overly smoothed. The decomposition of the cost function also provides an avenue for mitigating the effects of scale aliasing and representativeness errors that inherently exist in a multiscale system, thus further improving the effectiveness of the assimilation of high-resolution observations. A set of one-dimensional experiments is performed to examine the properties of the MS-DA scheme. Emphasis is placed on the assimilation of patchy high-resolution observations representing radar and satellite measurements, alongside sparse observations representing those from conventional in situ platforms. The results illustrate how MS-DA improves the effectiveness of the assimilation of both these types of observations simultaneously.

RACORO continental boundary layer cloud investigations: 1. Case study development and ensemble large‐scale forcings

AbstractObservation‐based modeling case studies of continental boundary layer clouds have been developed to study cloudy boundary layers, aerosol influences upon them, and their representation in cloud‐ and global‐scale models. Three 60 h case study periods span the temporal evolution of cumulus, stratiform, and drizzling boundary layer cloud systems, representing mixed and transitional states rather than idealized or canonical cases. Based on in situ measurements from the Routine AAF (Atmospheric Radiation Measurement (ARM) Aerial Facility) CLOWD (Clouds with Low Optical Water Depth) Optical Radiative Observations (RACORO) field campaign and remote sensing observations, the cases are designed with a modular configuration to simplify use in large‐eddy simulations (LES) and single‐column models. Aircraft measurements of aerosol number size distribution are fit to lognormal functions for concise representation in models. Values of the aerosol hygroscopicity parameter, κ, are derived from observations to be ~0.10, which are lower than the 0.3 typical over continents and suggestive of a large aerosol organic fraction. Ensemble large‐scale forcing data sets are derived from the ARM variational analysis, European Centre for Medium‐Range Weather Forecasts, and a multiscale data assimilation system. The forcings are assessed through comparison of measured bulk atmospheric and cloud properties to those computed in “trial” large‐eddy simulations, where more efficient run times are enabled through modest reductions in grid resolution and domain size compared to the full‐sized LES grid. Simulations capture many of the general features observed, but the state‐of‐the‐art forcings were limited at representing details of cloud onset, and tight gradients and high‐resolution transients of importance. Methods for improving the initial conditions and forcings are discussed. The cases developed are available to the general modeling community for studying continental boundary clouds.

Development of fine‐resolution analyses and expanded large‐scale forcing properties: 1. Methodology and evaluation

AbstractWe produce fine‐resolution, three‐dimensional fields of meteorological and other variables for the U.S. Department of Energy's Atmospheric Radiation Measurement (ARM) Southern Great Plains site. The Community Gridpoint Statistical Interpolation system is implemented in a multiscale data assimilation (MS‐DA) framework that is used within the Weather Research and Forecasting model at a cloud‐resolving resolution of 2 km. The MS‐DA algorithm uses existing reanalysis products and constrains fine‐scale atmospheric properties by assimilating high‐resolution observations. A set of experiments show that the data assimilation analysis realistically reproduces the intensity, structure, and time evolution of clouds and precipitation associated with a mesoscale convective system. Evaluations also show that the large‐scale forcing derived from the fine‐resolution analysis has an overall accuracy comparable to the existing ARM operational product. For enhanced applications, the fine‐resolution fields are used to characterize the contribution of subgrid variability to the large‐scale forcing and to derive hydrometeor forcing, which are presented in companion papers.

Development of fine‐resolution analyses and expanded large‐scale forcing properties: 2. Scale awareness and application to single‐column model experiments

AbstractFine‐resolution three‐dimensional fields have been produced using the Community Gridpoint Statistical Interpolation (GSI) data assimilation system for the U.S. Department of Energy's Atmospheric Radiation Measurement Program (ARM) Southern Great Plains region. The GSI system is implemented in a multiscale data assimilation framework using the Weather Research and Forecasting model at a cloud‐resolving resolution of 2 km. From the fine‐resolution three‐dimensional fields, large‐scale forcing is derived explicitly at grid‐scale resolution; a subgrid‐scale dynamic component is derived separately, representing subgrid‐scale horizontal dynamic processes. Analyses show that the subgrid‐scale dynamic component is often a major component over the large‐scale forcing for grid scales larger than 200 km. The single‐column model (SCM) of the Community Atmospheric Model version 5 is used to examine the impact of the grid‐scale and subgrid‐scale dynamic components on simulated precipitation and cloud fields associated with a mesoscale convective system. It is found that grid‐scale size impacts simulated precipitation, resulting in an overestimation for grid scales of about 200 km but an underestimation for smaller grids. The subgrid‐scale dynamic component has an appreciable impact on the simulations, suggesting that grid‐scale and subgrid‐scale dynamic components should be considered in the interpretation of SCM simulations.

Multiscale Methods for Data Assimilation in Turbulent Systems

Data assimilation of turbulent signals is an important challenging problem because of the extremely complicated large dimension of the signals and incomplete partial noisy observations which usually mix the large scale mean flow and small scale fluctuations. Due to the limited computing power in the foreseeable future, it is desirable to use multiscale forecast models which are cheap and fast to mitigate the curse of dimensionality in turbulent systems; thus model errors from imperfect forecast models are unavoidable in the development of a data assimilation method in turbulence. Here we propose a suite of multiscale data assimilation methods which use stochastic Superparameterization as the forecast model. Superparameterization is a seamless multiscale method for parameterizing the effect of small scales by cheap local problems embedded in a coarse grid. The key ingredient of the multiscale data assimilation methods is the systematic use of conditional Gaussian mixtures which make the methods efficient by filtering a subspace whose dimension is smaller than the full state. The multiscale data assimilation methods proposed here are tested on a six dimensional conceptual dynamical model for turbulence which mimics interesting features of anisotropic turbulence including two way coupling between the large and small scale parts, intermittencies, and extreme events in the smaller scale fluctuations. Numerical results show that suitable multiscale data assimilation methods have high skill in estimating the most energetic modes of turbulent signals even with infrequent observation times.

Multiscale Data Assimilation and Forecasting

Multiscale Data Assimilation and Forecasting

Integrating ASCAT surface soil moisture and GEOV1 leaf area index into the SURFEX modelling platform: a land data assimilation application over France

Abstract. The land monitoring service of the European Copernicus programme has developed a set of satellite-based biogeophysical products, including surface soil moisture (SSM) and leaf area index (LAI). This study investigates the impact of joint assimilation of remotely sensed SSM derived from Advanced Scatterometer (ASCAT) backscatter data and the Copernicus Global Land GEOV1 satellite-based LAI product into the the vegetation growth version of the Interactions between Soil Biosphere Atmosphere (ISBA-A-gs) land surface model within the the externalised surface model (SURFEX) modelling platform of Météo-France. The ASCAT data were bias corrected with respect to the model climatology by using a seasonal-based CDF (Cumulative Distribution Function) matching technique. A multivariate multi-scale land data assimilation system (LDAS) based on the extended Kalman Filter (EKF) is used for monitoring the soil moisture, terrestrial vegetation, surface carbon and energy fluxes across the domain of France at a spatial resolution of 8 km. Each model grid box is divided into a number of land covers, each having its own set of prognostic variables. The filter algorithm is designed to provide a distinct analysis for each land cover while using one observation per grid box. The updated values are aggregated by computing a weighted average. In this study, it is demonstrated that the assimilation scheme works effectively within the ISBA-A-gs model over a four-year period (2008–2011). The EKF is able to extract useful information from the data signal at the grid scale and distribute the root-zone soil moisture and LAI increments throughout the mosaic structure of the model. The impact of the assimilation on the vegetation phenology and on the water and carbon fluxes varies from one season to another. The spring drought of 2011 is an interesting case study of the potential of the assimilation to improve drought monitoring. A comparison between simulated and in situ soil moisture gathered at the twelve SMOSMANIA (Soil Moisture Observing System–Meteorological Automatic Network Integrated Application) stations shows improved anomaly correlations for eight stations.

New perspectives on superparameterization for geophysical turbulence

This is a research expository paper regarding superparameterization, a class of multi-scale numerical methods designed to cope with the intermittent multi-scale effects of inhomogeneous geophysical turbulence where energy often inverse-cascades from the unresolved scales to the large scales through the effects of waves, jets, vortices, and latent heat release from moist processes. Original as well as sparse space–time superparameterization algorithms are discussed for the important case of moist atmospheric convection including the role of multi-scale asymptotic methods in providing self-consistent constraints on superparameterization algorithms and related deterministic and stochastic multi-cloud parameterizations. Test models for the statistical numerical analysis of superparameterization algorithms are discussed both to elucidate the performance of the basic algorithms and to test their potential role in efficient multi-scale data assimilation. The very recent development of grid-free seamless stochastic superparameterization methods for geophysical turbulence appropriate for “eddy-permitting” mesoscale ocean turbulence is presented here including a general formulation and illustrative applications to two-layer quasigeostrophic turbulence, and another difficult test case involving one-dimensional models of dispersive wave turbulence. This last test case has randomly generated solitons as coherent structures which collapse and radiate wave energy back to the larger scales, resulting in strong direct and inverse turbulent energy cascades.

Application of the Weather Research and Forecasting Model for Air Quality Modeling in the San Francisco Bay Area

AbstractThe Weather Research and Forecasting (WRF) model is evaluated by conducting various sensitivity experiments over central California including the San Francisco Bay Area (SFBA), with the goal of establishing a WRF model configuration to be used by the Bay Area Air Quality Management District (BAAQMD) for its air quality applications. For the two selected cases, a winter particulate matter case and a summer ozone case, WRF solutions are evaluated both quantitatively by comparing the error statistics and qualitatively by analyzing the model-simulated mesoscale features. Model evaluation is also performed for the SFBA, Sacramento Valley, and San Joaquin Valley subregions. The recommended WRF configuration includes use of the Rapid Radiative Transfer Model/Dudhia (or RRTMG) radiation schemes and the Pleim–Xiu land surface physics, combined with a multiscale four-dimensional data assimilation strategy throughout the simulation period to assimilate the available observations, including standard observations from the World Meteorological Organization and local special observations. With the recommended model configuration, WRF is able to simulate the meteorological variables with reasonable error, with the added value, although relatively small, of assimilating the additional BAAQMD local special observations. Mesoscale features, simulated reasonably well for both cases, include the upslope and downslope flows that occur along the mountains that surround the Central Valley of California, as well as the mesoscale eddies that develop within the valley.

Forecasting mesoscale precipitation using the MM5 model with the four-dimensional data assimilation (FDDA) technique

The two basic forms of multi-scale data assimilation procedures (FDDA), based on Newtonian relaxation, of analysis and observations nudging have been applied for precipitation event period occurred over Portugal during summer season, using the Fifth Generation Mesoscale Model (MM5) developed and maintained by the Pennsylvania State University and National Center for Atmospheric Research (PSU/NCAR). The model has been configured for three nested grid domains covering part of the Eastern part of North Atlantic region evolving the Portugal, with 35 vertical levels, from surface up to 100 hPa top level. The model forecasting have been conducted employing daily available data from surface observational network, radio-sounding from Lisbon/Portugal and NOAA-16 polar orbiting satellite retrieved vertical profiles data. The three integration domains of MM5 model have been processed using, as boundary and first guess fields, the global atmospheric forecast NCEP-NWS/AVN model data gathered through the Unidata Local Data Manager (LDM)/Unidata Internet Data Distribution (IDD) system. All daily forecasting, with FDDA and with no FDDA, have been run for 60 hours forecast, with 30 minutes interval model data output to provide enough timely detailed results. The FDDA analysis presented a quite reasonable data ingesting volume of almost all available satellite data, with the exception of humidity data retrieved for high levels, above around 500 hPa. The obtained results indicate that, even using weak FDDA constraint coefficient values, presents a significant improvement in the numerical prognosis in the precipitation field, on both space and time integration levels. The results also presented an enhancement of the physics of the convective mesoscale system development, particularly over mountain region, indicating that it would be interesting to conduct an experiment with a dense data collecting platform coverage focused on events which occur in some prevailing mountain region of Portugal.

A data-assimilative ocean forecasting system for the Prince William sound and an evaluation of its performance during sound Predictions 2009

The development and implementation of a three-dimensional ocean modeling system for the Prince William Sound (PWS) is described. The system consists of a regional ocean model component (ROMS) forced by output from a regional atmospheric model component (the Weather Research and Forecasting Model, WRF). The ROMS ocean model component has a horizontal resolution of 1km within PWS and utilizes a recently-developed multi-scale 3DVAR data assimilation methodology along with freshwater runoff from land obtained via real-time execution of a digital elevation model. During the Sound Predictions Field Experiment (July 19–August 3, 2009) the system was run in real-time to support operations and incorporated all available real-time streams of data. Nowcasts were produced every 6h and a 48-h forecast was performed once a day. In addition, a sixteen-member ensemble of forecasts was executed on most days. All results were published at a web portal (http://ourocean.jpl.nasa.gov/PWS) in real time to support decision making.The performance of the system during Sound Predictions 2009 is evaluated. The ROMS results are first compared with the assimilated data as a consistency check. RMS differences of about 0.7°C were found between the ROMS temperatures and the observed vertical profiles of temperature that are assimilated. The ROMS salinities show greater discrepancies, tending to be too salty near the surface. The overall circulation patterns observed throughout the Sound are qualitatively reproduced, including the following evolution in time. During the first week of the experiment, the weather was quite stormy with strong southeasterly winds. This resulted in strong north to northwestward surface flow in much of the central PWS. Both the observed drifter trajectories and the ROMS nowcasts showed strong surface inflow into the Sound through the Hinchinbrook Entrance and strong generally northward to northwestward flow in the central Sound that was exiting through the Knight Island Passage and Montague Strait entrance. During the latter part of the second week when surface winds were light and southwesterly, the mean surface flow at the Hinchinbrook Entrance reversed to weak outflow and a cyclonic eddy formed in the central Sound. Overall, RMS differences between ROMS surface currents and observed HF radar surface currents in the central Sound were generally between 5 and 10cm/s, about 20–40% of the time mean current speeds.The ROMS reanalysis is then validated against independent observations. A comparison of the ROMS currents with observed vertical current profiles from moored ADCPs in the Hinchinbrook Entrance and Montague Strait shows good qualitative agreement and confirms the evolution of the near surface inflow/outflow at these locations described above. A comparison of the ROMS surface currents with drifter trajectories provided additional confirmation that the evolution of the surface flow described above was realistic. Forecasts of drifter locations had RMS errors of less than 10km for up to 36h. One and two-day forecasts of surface temperature, salinity and current fields were more skillful than persistence forecasts. In addition, ensemble mean forecasts were found to be slightly more skillful than single forecasts. Two case studies demonstrated the system’s qualitative skill in predicting subsurface changes within the mixed layer measured by ships and autonomous underwater vehicles. In summary, the system is capable of producing a realistic evolution of the near-surface circulation within PWS including forecasts of up to two days of this evolution. Use of the products provided by the system during the experiment as part of the asset deployment decision making process demonstrated the value of accurate regional ocean forecasts in support of field experiments.

Multivariate and Multiscale Data Assimilation in Terrestrial Systems: A Review

More and more terrestrial observational networks are being established to monitor climatic, hydrological and land-use changes in different regions of the World. In these networks, time series of states and fluxes are recorded in an automated manner, often with a high temporal resolution. These data are important for the understanding of water, energy, and/or matter fluxes, as well as their biological and physical drivers and interactions with and within the terrestrial system. Similarly, the number and accuracy of variables, which can be observed by spaceborne sensors, are increasing. Data assimilation (DA) methods utilize these observations in terrestrial models in order to increase process knowledge as well as to improve forecasts for the system being studied. The widely implemented automation in observing environmental states and fluxes makes an operational computation more and more feasible, and it opens the perspective of short-time forecasts of the state of terrestrial systems. In this paper, we review the state of the art with respect to DA focusing on the joint assimilation of observational data precedents from different spatial scales and different data types. An introduction is given to different DA methods, such as the Ensemble Kalman Filter (EnKF), Particle Filter (PF) and variational methods (3/4D-VAR). In this review, we distinguish between four major DA approaches: (1) univariate single-scale DA (UVSS), which is the approach used in the majority of published DA applications, (2) univariate multiscale DA (UVMS) referring to a methodology which acknowledges that at least some of the assimilated data are measured at a different scale than the computational grid scale, (3) multivariate single-scale DA (MVSS) dealing with the assimilation of at least two different data types, and (4) combined multivariate multiscale DA (MVMS). Finally, we conclude with a discussion on the advantages and disadvantages of the assimilation of multiple data types in a simulation model. Existing approaches can be used to simultaneously update several model states and model parameters if applicable. In other words, the basic principles for multivariate data assimilation are already available. We argue that a better understanding of the measurement errors for different observation types, improved estimates of observation bias and improved multiscale assimilation methods for data which scale nonlinearly is important to properly weight them in multiscale multivariate data assimilation. In this context, improved cross-validation of different data types, and increased ground truth verification of remote sensing products are required.