On instabilities in data assimilation algorithms

Data assimilation algorithms are a crucial part of operational systems in numerical weather prediction, hydrology and climate science, but are also important for dynamical reconstruction in medical applications and quality control for manufacturing processes. Usually, a variety of diverse measurement data are employed to determine the state of the atmosphere or to a wider system including land and oceans. Modern data assimilation systems use more and more remote sensing data, in particular radiances measured by satellites, radar data and integrated water vapor measurements via GPS/GNSS signals. The inversion of some of these measurements are ill-posed in the classical sense, i.e. the inverse of the operator H which maps the state onto the data is unbounded. In this case, the use of such data can lead to significant instabilities of data assimilation algorithms. The goal of this work is to provide a rigorous mathematical analysis of the instability of well-known data assimilation methods. Here, we will restrict our attention to particular linear systems, in which the instability can be explicitly analyzed. We investigate the three-dimensional variational assimilation and four-dimensional variational assimilation. A theory for the instability is developed using the classical theory of ill-posed problems in a Banach space framework. Further, we demonstrate by numerical examples that instabilities can and will occur, including an example from dynamic magnetic tomography.

Resolution of sharp fronts in the presence of model error in variational data assimilation

We show that the four-dimensional variational data assimilation method (4DVar) can be interpreted as a form of Tikhonov regularization, a very familiar method for solving ill-posed inverse problems. It is known from image restoration problems that L1-norm penalty regularization recovers sharp edges in the image more accurately than Tikhonov, or L2-norm, penalty regularization. We apply this idea from stationary inverse problems to 4DVar, a dynamical inverse problem, and give examples for an L1-norm penalty approach and a mixed total variation (TV) L1–L2-norm penalty approach. For problems with model error where sharp fronts are present and the background and observation error covariances are known, the mixed TV L1–L2-norm penalty performs better than either the L1-norm method or the strong constraint 4DVar (L2-norm)method. A strength of the mixed TV L1–L2-norm regularization is that in the case where a simplified form of the background error covariance matrix is used it produces a much more accurate analysis than 4DVar. The method thus has the potential in numerical weather prediction to overcome operational problems with poorly tuned background error covariance matrices.

From projects into operations: lessons for data handover

Data from civil engineering projects can inform the operation of built infrastructure. This paper captures lessons for such data handover, from projects into operations, through interviews with leading clients and their supply chain. Clients are found to value receiving accurate and complete data. They recognise opportunities to use high quality information in decision-making about capital and operational expenditure; as well as in ensuring compliance with regulatory requirements. Providing this value to clients is a motivation for information management in projects. However, data handover is difficult as key people leave before project completion; and different data formats and structures are used in project delivery and operations. Lessons learnt from leading practice include defining data requirements at the outset, getting operations teams involved early, shaping the evolution of interoperable systems and standards, developing handover processes to check data rather than documentation, and fostering skills to use and update project data in operations

Prediction of visibility and aerosol within the operational Met Office unified model. I: Model formulation and variational assimilation

The formulation and performance of the Met Office visibility analysis and prediction system are described. The visibility diagnostic within the limited-area Unified Model is a function of humidity and a prognostic aerosol content. The aerosol model includes advection, industrial and general urban sources, plus boundary-layer mixing and removal by rain. The assimilation is a 3-dimensional variational scheme in which the visibility observation operator is a very nonlinear function of humidity, aerosol and temperature. A quality control scheme for visibility data is included. Visibility observations can give rise to humidity increments of significant magnitude compared with the direct impact of humidity observations. We present the results of sensitivity studies which show the contribution of different components of the system to improved skill in visibility forecasts. Visibility assimilation is most important within the first 6-12 hours of the forecast and for visibilities below 1 km, while modelling of aerosol sources and advection is important for slightly higher visibilities (1-5 km) and is still significant at longer forecast times

On the growth of errors in data assimilation systems

Considerable progress has taken place in numerical weather prediction over the last decade. It has been possible to extend predictive skills in the extra-tropics of the Northern Hemisphere during the winter from less than five days to seven days. Similar improvements, albeit on a lower level, have taken place in the Southern Hemisphere. Another example of improvement in the forecasts is the prediction of intense synoptic phenomena such as cyclogenesis which on the whole is quite successful with the most advanced operational models (Bengtsson (1989), Gadd and Kruze (1988)). A careful examination shows that there are no single causes for the improvements in predictive skill, but instead they are due to several different factors encompassing the forecasting system as a whole (Bengtsson, 1985). In this paper we will focus our attention on the role of data-assimilation and the effect it may have on reducing the initial error and hence improving the forecast. The first part of the paper contains a theoretical discussion on error growth in simple data assimilation systems, following Leith (1983). In the second part we will apply the result on actual forecast data from ECMWF. The potential for further forecast improvements within the framework of the present observing system in the two hemispheres will be discussed.

Data assimilation with correlated observation errors: experiments with a 1-D shallow water model

Remote sensing observations often have correlated errors, but the correlations are typically ignored in data assimilation for numerical weather prediction. The assumption of zero correlations is often used with data thinning methods, resulting in a loss of information. As operational centres move towards higher-resolution forecasting, there is a requirement to retain data providing detail on appropriate scales. Thus an alternative approach to dealing with observation error correlations is needed. In this article, we consider several approaches to approximating observation error correlation matrices: diagonal approximations, eigendecomposition approximations and Markov matrices. These approximations are applied in incremental variational assimilation experiments with a 1-D shallow water model using synthetic observations. Our experiments quantify analysis accuracy in comparison with a reference or ‘truth’ trajectory, as well as with analyses using the ‘true’ observation error covariance matrix. We show that it is often better to include an approximate correlation structure in the observation error covariance matrix than to incorrectly assume error independence. Furthermore, by choosing a suitable matrix approximation, it is feasible and computationally cheap to include error correlation structure in a variational data assimilation algorithm.

The global distribution of phytoplankton size spectrum and size classes from their light-absorption spectra derived from satellite data

The absorption spectra of phytoplankton in the visible domain hold implicit information on the phytoplankton community structure. Here we use this information to retrieve quantitative information on phytoplankton size structure by developing a novel method to compute the exponent of an assumed power-law for their particle-size spectrum. This quantity, in combination with total chlorophyll-a concentration, can be used to estimate the fractional concentration of chlorophyll in any arbitrarily-defined size class of phytoplankton. We further define and derive expressions for two distinct measures of cell size of mixed populations, namely, the average spherical diameter of a bio-optically equivalent homogeneous population of cells of equal size, and the average equivalent spherical diameter of a population of cells that follow a power-law particle-size distribution. The method relies on measurements of two quantities of a phytoplankton sample: the concentration of chlorophyll-a, which is an operational index of phytoplankton biomass, and the total absorption coefficient of phytoplankton in the red peak of visible spectrum at 676 nm. A sensitivity analysis confirms that the relative errors in the estimates of the exponent of particle size spectra are reasonably low. The exponents of phytoplankton size spectra, estimated for a large set of in situ data from a variety of oceanic environments (~ 2400 samples), are within a reasonable range; and the estimated fractions of chlorophyll in pico-, nano- and micro-phytoplankton are generally consistent with those obtained by an independent, indirect method based on diagnostic pigments determined using high-performance liquid chromatography. The estimates of cell size for in situ samples dominated by different phytoplankton types (diatoms, prymnesiophytes, Prochlorococcus, other cyanobacteria and green algae) yield nominal sizes consistent with the taxonomic classification. To estimate the same quantities from satellite-derived ocean-colour data, we combine our method with algorithms for obtaining inherent optical properties from remote sensing. The spatial distribution of the size-spectrum exponent and the chlorophyll fractions of pico-, nano- and micro-phytoplankton estimated from satellite remote sensing are in agreement with the current understanding of the biogeography of phytoplankton functional types in the global oceans. This study contributes to our understanding of the distribution and time evolution of phytoplankton size structure in the global oceans.

Sea surface temperature estimation from the Geostationary Operational Environmental Satellite 12 (GOES-12)

This paper describes the techniques used to obtain sea surface temperature (SST) retrievals from the Geostationary Operational Environmental Satellite 12 (GOES-12) at the National Oceanic and Atmospheric Administration’s Office of Satellite Data Processing and Distribution. Previous SST retrieval techniques relying on channels at 11 and 12 μm are not applicable because GOES-12 lacks the latter channel. Cloud detection is performed using a Bayesian method exploiting fast-forward modeling of prior clear-sky radiances using numerical weather predictions. The basic retrieval algorithm used at nighttime is based on a linear combination of brightness temperatures at 3.9 and 11 μm. In comparison with traditional split window SSTs (using 11- and 12-μm channels), simulations show that this combination has maximum scatter when observing drier colder scenes, with a comparable overall performance. For daytime retrieval, the same algorithm is applied after estimating and removing the contribution to brightness temperature in the 3.9-μm channel from solar irradiance. The correction is based on radiative transfer simulations and comprises a parameterization for atmospheric scattering and a calculation of ocean surface reflected radiance. Potential use of the 13-μm channel for SST is shown in a simulation study: in conjunction with the 3.9-μm channel, it can reduce the retrieval error by 30%. Some validation results are shown while a companion paper by Maturi et al. shows a detailed analysis of the validation results for the operational algorithms described in this present article.

NOAA's sea surface temperature products from operational geostationary satellites

NOAA's National Environmental Satellite, Data, and Information Service (NESDIS) has generated sea surface temperature (SST) products from Geostationary Operational Environmental Satellite (GOES)-East (E) and GOES-West (W) on an operational basis since December 2000. Since that time, a process of continual development has produced steady improvements in product accuracy. Recent improvements extended the capability to permit generation of operational SST retrievals from the Japanese Multifunction Transport Satellite (MTSAT)-1R and the European Meteosat Second Generation (MSG) satellite, thereby extending spatial coverage. The four geostationary satellites (at longitudes of 75°W, 135°W, 140°E, and 0°) provide high temporal SST retrievals for most of the tropics and midlatitudes, with the exception of a region between 60° and 80°E. Because of ongoing development, the quality of these retrievals now approaches that of SST products from the polar-orbiting Advanced Very High Resolution Radiometer (AVHRR). These products from GOES provide hourly regional imagery, 3-hourly hemispheric imagery, 24-h merged composites, a GOES SST level 2 preprocessed product every 1/2 h for each hemisphere, and a match-up data file for each product. The MTSAT and the MSG products include hourly, 3-hourly, and 24-h merged composites. These products provide the user community with a reliable source of SST observations, with improved accuracy and increased coverage in important oceanographic, meteorological, and climatic regions.

Estimating interchannel observation-error correlations for IASI radiance data in the Met Office system

The optimal utilisation of hyper-spectral satellite observations in numerical weather prediction is often inhibited by incorrectly assuming independent interchannel observation errors. However, in order to represent these observation-error covariance structures, an accurate knowledge of the true variances and correlations is needed. This structure is likely to vary with observation type and assimilation system. The work in this article presents the initial results for the estimation of IASI interchannel observation-error correlations when the data are processed in the Met Office one-dimensional (1D-Var) and four-dimensional (4D-Var) variational assimilation systems. The method used to calculate the observation errors is a post-analysis diagnostic which utilises the background and analysis departures from the two systems. The results show significant differences in the source and structure of the observation errors when processed in the two different assimilation systems, but also highlight some common features. When the observations are processed in 1D-Var, the diagnosed error variances are approximately half the size of the error variances used in the current operational system and are very close in size to the instrument noise, suggesting that this is the main source of error. The errors contain no consistent correlations, with the exception of a handful of spectrally close channels. When the observations are processed in 4D-Var, we again find that the observation errors are being overestimated operationally, but the overestimation is significantly larger for many channels. In contrast to 1D-Var, the diagnosed error variances are often larger than the instrument noise in 4D-Var. It is postulated that horizontal errors of representation, not seen in 1D-Var, are a significant contributor to the overall error here. Finally, observation errors diagnosed from 4D-Var are found to contain strong, consistent correlation structures for channels sensitive to water vapour and surface properties.

RMetS special interest group meeting: high resolution data assimilation

Data assimilation (DA) systems are evolving to meet the demands of convection-permitting models in the field of weather forecasting. On 19 April 2013 a special interest group meeting of the Royal Meteorological Society brought together UK researchers looking at different aspects of the data assimilation problem at high resolution, from theory to applications, and researchers creating our future high resolution observational networks. The meeting was chaired by Dr Sarah Dance of the University of Reading and Dr Cristina Charlton-Perez from the MetOffice@Reading. The purpose of the meeting was to help define the current state of high resolution data assimilation in the UK. The workshop assembled three main types of scientists: observational network specialists, operational numerical weather prediction researchers and those developing the fundamental mathematical theory behind data assimilation and the underlying models. These three working areas are intrinsically linked; therefore, a holistic view must be taken when discussing the potential to make advances in high resolution data assimilation.

Challenges of operational river forecasting

Skillful and timely streamflow forecasts are critically important to water managers and emergency protection services. To provide these forecasts, hydrologists must predict the behavior of complex coupled human–natural systems using incomplete and uncertain information and imperfect models. Moreover, operational predictions often integrate anecdotal information and unmodeled factors. Forecasting agencies face four key challenges: 1) making the most of available data, 2) making accurate predictions using models, 3) turning hydrometeorological forecasts into effective warnings, and 4) administering an operational service. Each challenge presents a variety of research opportunities, including the development of automated quality-control algorithms for the myriad of data used in operational streamflow forecasts, data assimilation, and ensemble forecasting techniques that allow for forecaster input, methods for using human-generated weather forecasts quantitatively, and quantification of human interference in the hydrologic cycle. Furthermore, much can be done to improve the communication of probabilistic forecasts and to design a forecasting paradigm that effectively combines increasingly sophisticated forecasting technology with subjective forecaster expertise. These areas are described in detail to share a real-world perspective and focus for ongoing research endeavors.

Organic soils promote the efficacy of entomopathogenic nematodes, with different foraging strategies, in the control of a major forest pest: a meta-analysis of field trial data

The large pine weevil, Hylobius abietis, is a serious pest of reforestation in northern Europe. However, weevils developing in stumps of felled trees can be killed by entomopathogenic nematodes applied to soil around the stumps and this method of control has been used at an operational level in the UK and Ireland. We investigated the factors affecting the efficacy of entomopathogenic nematodes in the control of the large pine weevil spanning 10 years of field experiments, by means of a meta-analysis of published studies and previously unpublished data. We investigated two species with different foraging strategies, the ‘ambusher’ Steinernema carpocapsae, the species most often used at an operational level, and the ‘cruiser’ Heterorhabditis downesi. Efficacy was measured both by percentage reduction in numbers of adults emerging relative to untreated controls and by percentage parasitism of developing weevils in the stump. Both measures were significantly higher with H. downesi compared to S. carpocapsae. General linear models were constructed for each nematode species separately, using substrate type (peat versus mineral soil) and tree species (pine versus spruce) as fixed factors, weevil abundance (from the mean of untreated stumps) as a covariate and percentage reduction or percentage parasitism as the response variable. For both nematode species, the most significant and parsimonious models showed that substrate type was consistently, but not always, the most significant variable, whether replicates were at a site or stump level, and that peaty soils significantly promote the efficacy of both species. Efficacy, in terms of percentage parasitism, was not density dependent.

Exploring strategies for coupled 4D-Var data assimilation using an idealised atmosphere-ocean model

Operational forecasting centres are currently developing data assimilation systems for coupled atmosphere-ocean models. Strongly coupled assimilation, in which a single assimilation system is applied to a coupled model, presents significant technical and scientific challenges. Hence weakly coupled assimilation systems are being developed as a first step, in which the coupled model is used to compare the current state estimate with observations, but corrections to the atmosphere and ocean initial conditions are then calculated independently. In this paper we provide a comprehensive description of the different coupled assimilation methodologies in the context of four dimensional variational assimilation (4D-Var) and use an idealised framework to assess the expected benefits of moving towards coupled data assimilation. We implement an incremental 4D-Var system within an idealised single column atmosphere-ocean model. The system has the capability to run both strongly and weakly coupled assimilations as well as uncoupled atmosphere or ocean only assimilations, thus allowing a systematic comparison of the different strategies for treating the coupled data assimilation problem. We present results from a series of identical twin experiments devised to investigate the behaviour and sensitivities of the different approaches. Overall, our study demonstrates the potential benefits that may be expected from coupled data assimilation. When compared to uncoupled initialisation, coupled assimilation is able to produce more balanced initial analysis fields, thus reducing initialisation shock and its impact on the subsequent forecast. Single observation experiments demonstrate how coupled assimilation systems are able to pass information between the atmosphere and ocean and therefore use near-surface data to greater effect. We show that much of this benefit may also be gained from a weakly coupled assimilation system, but that this can be sensitive to the parameters used in the assimilation.

Design of a unified data with business rules storage model for OLTP and OLAP systems

This paper reviews the literature concerning the practice of using Online Analytical Processing (OLAP) systems to recall information stored by Online Transactional Processing (OLTP) systems. Such a review provides a basis for discussion on the need for the information that are recalled through OLAP systems to maintain the contexts of transactions with the data captured by the respective OLTP system. The paper observes an industry trend involving the use of OLTP systems to process information into data, which are then stored in databases without the business rules that were used to process information and data stored in OLTP databases without associated business rules. This includes the necessitation of a practice, whereby, sets of business rules are used to extract, cleanse, transform and load data from disparate OLTP systems into OLAP databases to support the requirements for complex reporting and analytics. These sets of business rules are usually not the same as business rules used to capture data in particular OLTP systems. The paper argues that, differences between the business rules used to interpret these same data sets, risk gaps in semantics between information captured by OLTP systems and information recalled through OLAP systems. Literature concerning the modeling of business transaction information as facts with context as part of the modelling of information systems were reviewed to identify design trends that are contributing to the design quality of OLTP and OLAP systems. The paper then argues that; the quality of OLTP and OLAP systems design has a critical dependency on the capture of facts with associated context, encoding facts with contexts into data with business rules, storage and sourcing of data with business rules, decoding data with business rules into the facts with the context and recall of facts with associated contexts. The paper proposes UBIRQ, a design model to aid the co-design of data with business rules storage for OLTP and OLAP purposes. The proposed design model provides the opportunity for the implementation and use of multi-purpose databases, and business rules stores for OLTP and OLAP systems. Such implementations would enable the use of OLTP systems to record and store data with executions of business rules, which will allow for the use of OLTP and OLAP systems to query data with business rules used to capture the data. Thereby ensuring information recalled via OLAP systems preserves the contexts of transactions as per the data captured by the respective OLTP system.