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.

Entity-based semantic search on conversational transcripts semantic

This paper describes the implementation of a semantic web search engine on conversation styled transcripts. Our choice of data is Hansard, a publicly available conversation style transcript of parliamentary debates. The current search engine implementation on Hansard is limited to running search queries based on keywords or phrases hence lacks the ability to make semantic inferences from user queries. By making use of knowledge such as the relationship between members of parliament, constituencies, terms of office, as well as topics of debates the search results can be improved in terms of both relevance and coverage. Our contribution is not algorithmic instead we describe how we exploit a collection of external data sources, ontologies, semantic web vocabularies and named entity extraction in the analysis of underlying semantics of user queries as well as the semantic enrichment of the search index thereby improving the quality of results.

Description and process evaluation of pharmacists’ interventions in a pharmacist-led information technology-enabled multicentre cluster randomised controlled trial for reducing medication errors in general practice (PINCER trial)

Objective To undertake a process evaluation of pharmacists' recommendations arising in the context of a complex IT-enabled pharmacist-delivered randomised controlled trial (PINCER trial) to reduce the risk of hazardous medicines management in general practices. Methods PINCER pharmacists manually recorded patients’ demographics, details of interventions recommended, actions undertaken by practice staff and time taken to manage individual cases of hazardous medicines management. Data were coded and double entered into SPSS v15, and then summarised using percentages for categorical data (with 95% CI) and, as appropriate, means (SD) or medians (IQR) for continuous data. Key findings Pharmacists spent a median of 20 minutes (IQR 10, 30) reviewing medical records, recommending interventions and completing actions in each case of hazardous medicines management. Pharmacists judged 72% (95%CI 70, 74) (1463/2026) of cases of hazardous medicines management to be clinically relevant. Pharmacists recommended 2105 interventions in 74% (95%CI 73, 76) (1516/2038) of cases and 1685 actions were taken in 61% (95%CI 59, 63) (1246/2038) of cases; 66% (95%CI 64, 68) (1383/2105) of interventions recommended by pharmacists were completed and 5% (95%CI 4, 6) (104/2105) of recommendations were accepted by general practitioners (GPs), but not completed at the end of the pharmacists’ placement; the remaining recommendations were rejected or considered not relevant by GPs. Conclusions The outcome measures were used to target pharmacist activity in general practice towards patients at risk from hazardous medicines management. Recommendations from trained PINCER pharmacists were found to be broadly acceptable to GPs and led to ameliorative action in the majority of cases. It seems likely that the approach used by the PINCER pharmacists could be employed by other practice pharmacists following appropriate training.

The prevalence and nature of prescribing and monitoring errors in English general practice – a retrospective case note review

Objective To determine the prevalence and nature of prescribing and monitoring errors in general practices in England. Design Retrospective case note review of unique medication items prescribed over a 12 month period to a 2% random sample of patients. Mixed effects logistic regression was used to analyse the data. Setting Fifteen general practices across three primary care trusts in England. Data sources Examination of 6048 unique prescription items prescribed over the previous 12 months for 1777 patients. Main outcome measures Prevalence of prescribing and monitoring errors, and severity of errors, using validated definitions. Results Prescribing and/or monitoring errors were detected in 4.9% (296/6048) of all prescription items (95% confidence interval 4.4 - 5.5%). The vast majority of errors were of mild to moderate severity, with 0.2% (11/6048) of items having a severe error. After adjusting for covariates, patient-related factors associated with an increased risk of prescribing and/or monitoring errors were: age less than 15 (Odds Ratio (OR) 1.87, 1.19 to 2.94, p=0.006) or greater than 64 years (OR 1.68, 1.04 to 2.73, p=0.035), and higher numbers of unique medication items prescribed (OR 1.16, 1.12 to 1.19, p<0.001). Conclusion Prescribing and monitoring errors are common in English general practice, although severe errors are unusual. Many factors increase the risk of error. Having identified the most common and important errors, and the factors associated with these, strategies to prevent future errors should be developed based on the study findings.

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 causes of prescribing errors in English general practices: a qualitative study

Aim: To examine the causes of prescribing and monitoring errors in English general practices and provide recommendations for how they may be overcome. Design: Qualitative interview and focus group study with purposive sampling and thematic analysis informed by Reason’s accident causation model. Participants: General practice staff participated in a combination of semi-structured interviews (n=34) and six focus groups (n=46). Setting: Fifteen general practices across three primary care trusts in England. Results: We identified seven categories of high-level error-producing conditions: the prescriber, the patient, the team, the task, the working environment, the computer system, and the primary-secondary care interface. Each of these was further broken down to reveal various error-producing conditions. The prescriber’s therapeutic training, drug knowledge and experience, knowledge of the patient, perception of risk, and their physical and emotional health, were all identified as possible causes. The patient’s characteristics and the complexity of the individual clinical case were also found to have contributed to prescribing errors. The importance of feeling comfortable within the practice team was highlighted, as well as the safety of general practitioners (GPs) in signing prescriptions generated by nurses when they had not seen the patient for themselves. The working environment with its high workload, time pressures, and interruptions, and computer related issues associated with mis-selecting drugs from electronic pick-lists and overriding alerts, were all highlighted as possible causes of prescribing errors and often interconnected. Conclusion: This study has highlighted the complex underlying causes of prescribing and monitoring errors in general practices, several of which are amenable to intervention.

Phonological therapy in jargon aphasia: effects on naming and neologisms

Background: Jargon aphasia is one of the most intractable forms of aphasia with limited recommendation on amelioration of associated naming difficulties and neologisms. The few naming therapy studies that exist in jargon aphasia have utilized either semantic or phonological approaches but the results have been equivocal. Moreover, the effect of therapy on characteristics of neologisms is less explored. Aims: This study investigates the effectiveness of a phonological naming therapy (i.e., phonological component analysis, PCA) on picture naming abilities and on quantitative and qualitative changes in neologisms for an individual with jargon aphasia (FF). Methods: FF showed evidence of jargon aphasia with severe naming difficulties and produced a very high proportion of neologisms. A single-subject multiple probe design across behaviors was employed to evaluate the effects of PCA therapy on the accuracy for three sets of words. In therapy, a phonological components analysis chart was used to identify five phonological components (i.e., rhymes, first sound, first sound associate, final sound, number of syllables) for each target word. Generalization effects—change in percent accuracy and error pattern—were examined comparing pre-and post-therapy responses on the Philadelphia Naming Test and these responses were analyzed to explore the characteristics of the neologisms. The quantitative change in neologisms was measured by change in the proportion of neologisms from pre- to post-therapy and the qualitative change was indexed by the phonological overlap between target and neologism. Results: As a consequence of PCA therapy, FF showed a significant improvement in his ability to name the treated items. His performance in maintenance and follow-up phases remained comparable to his performance during the therapy phases. Generalization to other naming tasks did not show a change in accuracy but distinct differences in error pattern (an increase in proportion of real word responses and a decrease in proportion of neologisms) were observed. Notably, the decrease in neologisms occurred with a corresponding trend for increase in the phonological similarity between the neologisms and the targets. Conclusions: This study demonstrated the effectiveness of a phonological therapy for improving naming abilities and reducing the amount of neologisms in an individual with severe jargon aphasia. The positive outcome of this research is encouraging, as it provides evidence for effective therapies for jargon aphasia and also emphasizes that use of the quality and quantity of errors may provide a sensitive outcome measure to determine therapy effectiveness, in particular for client groups who are difficult to treat.

Retrievals of sea surface temperature from infrared imagery: origin and form of systematic errors

We show that retrievals of sea surface temperature from satellite infrared imagery are prone to two forms of systematic error: prior error (familiar from the theory of atmospheric sounding) and error arising from nonlinearity. These errors have different complex geographical variations, related to the differing geographical distributions of the main geophysical variables that determine clear-sky brightness-temperatures over the oceans. We show that such errors arise as an intrinsic consequence of the form of the retrieval (rather than as a consequence of sub-optimally specified retrieval coefficients, as is often assumed) and that the pattern of observed errors can be simulated in detail using radiative-transfer modelling. The prior error has the linear form familiar from atmospheric sounding. A quadratic equation for nonlinearity error is derived, and it is verified that the nonlinearity error exhibits predominantly quadratic behaviour in this case.

Broadly speaking: vocabulary in semantic dementia shifts towards general, semantically diverse words

One of the cardinal features of semantic dementia (SD) is a steady reduction in expressive vocabulary. We investigated the nature of this breakdown by assessing the psycholinguistic characteristics of words produced spontaneously by SD patients during an autobiographical memory interview. Speech was analysed with respect to frequency and imageability, and a recently-developed measure called semantic diversity. This measure quantifies the degree to which a word can be used in a broad range of different linguistic contexts. We used this measure in a formal exploration of the tendency for SD patients to replace specific terms with more vague and general words, on the assumption that more specific words are used in a more constrained set of contexts. Relative to healthy controls, patients were less likely to produce low-frequency, high-imageability words, and more likely to produce highly frequent, abstract words. These changes in the lexical-semantic landscape were related to semantic diversity: the highly frequent and abstract words most prevalent in the patients' speech were also the most semantically diverse. In fact, when the speech samples of healthy controls were artificially engineered such that low semantic diversity words (e.g., garage, spanner) were replaced with broader terms (e.g., place, thing), the characteristics of their speech production came to closely resemble that of SD patients. A similar simulation in which low-frequency words were replaced was less successful in replicating the patient data. These findings indicate systematic biases in the deterioration of lexical-semantic space in SD. As conceptual knowledge degrades, speech increasingly consists of general terms that can be applied in a broad range of linguistic contexts and convey less specific information.

Care homes’ use of medicines study: prevalence, causes and potential harm of medication errors in care homes for older people

Introduction: Care home residents are at particular risk from medication errors, and our objective was to determine the prevalence and potential harm of prescribing, monitoring, dispensing and administration errors in UK care homes, and to identify their causes. Methods: A prospective study of a random sample of residents within a purposive sample of homes in three areas. Errors were identified by patient interview, note review, observation of practice and examination of dispensed items. Causes were understood by observation and from theoretically framed interviews with home staff, doctors and pharmacists. Potential harm from errors was assessed by expert judgement. Results: The 256 residents recruited in 55 homes were taking a mean of 8.0 medicines. One hundred and seventy-eight (69.5%) of residents had one or more errors. The mean number per resident was 1.9 errors. The mean potential harm from prescribing, monitoring, administration and dispensing errors was 2.6, 3.7, 2.1 and 2.0 (0 = no harm, 10 = death), respectively. Contributing factors from the 89 interviews included doctors who were not accessible, did not know the residents and lacked information in homes when prescribing; home staff’s high workload, lack of medicines training and drug round interruptions; lack of team work among home, practice and pharmacy; inefficient ordering systems; inaccurate medicine records and prevalence of verbal communication; and difficult to fill (and check) medication administration systems. Conclusions: That two thirds of residents were exposed to one or more medication errors is of concern. The will to improve exists, but there is a lack of overall responsibility. Action is required from all concerned.

A generalised Bayesian instrumental variable approach under student t-distributed errors with application

Bayesian analysis is given of an instrumental variable model that allows for heteroscedasticity in both the structural equation and the instrument equation. Specifically, the approach for dealing with heteroscedastic errors in Geweke (1993) is extended to the Bayesian instrumental variable estimator outlined in Rossi et al. (2005). Heteroscedasticity is treated by modelling the variance for each error using a hierarchical prior that is Gamma distributed. The computation is carried out by using a Markov chain Monte Carlo sampling algorithm with an augmented draw for the heteroscedastic case. An example using real data illustrates the approach and shows that ignoring heteroscedasticity in the instrument equation when it exists may lead to biased estimates.

Lesions impairing regular versus irregular past tense production

We investigated selective impairments in the production of regular and irregular past tense by examining language performance and lesion sites in a sample of twelve stroke patients. A disadvantage in regular past tense production was observed in six patients when phonological complexity was greater for regular than irregular verbs, and in three patients when phonological complexity was closely matched across regularity. These deficits were not consistently related to grammatical difficulties or phonological errors but were consistently related to lesion site. All six patients with a regular past tense disadvantage had damage to the left ventral pars opercularis (in the inferior frontal cortex), an area associated with articulatory sequencing in prior functional imaging studies. In addition, those that maintained a disadvantage for regular verbs when phonological complexity was controlled had damage to the left ventral supramarginal gyrus (in the inferior parietal lobe), an area associated with phonological short-term memory. When these frontal and parietal regions were spared in patients who had damage to subcortical (n = 2) or posterior temporo-parietal regions (n = 3), past tense production was relatively unimpaired for both regular and irregular forms. The remaining (12th) patient was impaired in producing regular past tense but was significantly less accurate when producing irregular past tense. This patient had frontal, parietal, subcortical and posterior temporo-parietal damage, but was distinguished from the other patients by damage to the left anterior temporal cortex, an area associated with semantic processing. We consider how our lesion site and behavioral observations have implications for theoretical accounts of past tense production.

An historical perspective on forecast errors

Using annual observations on industrial production over the last three centuries, and on GDP over a 100-year period, we seek an historical perspective on the forecastability of these UK output measures. The series are dominated by strong upward trends, so we consider various specifications of this, including the local linear trend structural time-series model, which allows the level and slope of the trend to vary. Our results are not unduly sensitive to how the trend in the series is modelled: the average sizes of the forecast errors of all models, and the wide span of prediction intervals, attests to a great deal of uncertainty in the economic environment. It appears that, from an historical perspective, the postwar period has been relatively more forecastable.

A systematic approach to identify the sources of tropical SST errors in coupled models using the adjustment of initialised experiments

Understanding the sources of systematic errors in climate models is challenging because of coupled feedbacks and errors compensation. The developing seamless approach proposes that the identification and the correction of short term climate model errors have the potential to improve the modeled climate on longer time scales. In previous studies, initialised atmospheric simulations of a few days have been used to compare fast physics processes (convection, cloud processes) among models. The present study explores how initialised seasonal to decadal hindcasts (re-forecasts) relate transient week-to-month errors of the ocean and atmospheric components to the coupled model long-term pervasive SST errors. A protocol is designed to attribute the SST biases to the source processes. It includes five steps: (1) identify and describe biases in a coupled stabilized simulation, (2) determine the time scale of the advent of the bias and its propagation, (3) find the geographical origin of the bias, (4) evaluate the degree of coupling in the development of the bias, (5) find the field responsible for the bias. This strategy has been implemented with a set of experiments based on the initial adjustment of initialised simulations and exploring various degrees of coupling. In particular, hindcasts give the time scale of biases advent, regionally restored experiments show the geographical origin and ocean-only simulations isolate the field responsible for the bias and evaluate the degree of coupling in the bias development. This strategy is applied to four prominent SST biases of the IPSLCM5A-LR coupled model in the tropical Pacific, that are largely shared by other coupled models, including the Southeast Pacific warm bias and the equatorial cold tongue bias. Using the proposed protocol, we demonstrate that the East Pacific warm bias appears in a few months and is caused by a lack of upwelling due to too weak meridional coastal winds off Peru. The cold equatorial bias, which surprisingly takes 30 years to develop, is the result of an equatorward advection of midlatitude cold SST errors. Despite large development efforts, the current generation of coupled models shows only little improvement. The strategy proposed in this study is a further step to move from the current random ad hoc approach, to a bias-targeted, priority setting, systematic model development approach.