Daily scale wintertime sea surface temperature and IPC-Navidad variability in the southern Bay of Biscay from 1981 to 2010

The combination of remotely sensed gappy Sea surface temperature (SST) images with the missing data filling DINEOF (data interpolating empirical orthogonal functions) technique, followed by a principal component analysis of the reconstructed data, has been used to identify the time evolution and the daily scale variability of the wintertime surface signal of the Iberian Poleward Current (IPC), or Navidad, during the 1981-2010 period. An exhaustive comparison with the existing bibliography, and the vertical temperature and salinity profiles related to its extremes over the Bay of Biscay area, show that the obtained time series accurately reflect the IPC-Navidad variability. Once a time series for the evolution of the SST signal of the current over the last decades is well established, this time series is used to propose a physical mechanism in relation to the variability of the IPC-Navidad, involving both atmospheric and oceanic variables. According to the proposed mechanism, an atmospheric circulation anomaly observed in both the 500 hPa and the surface levels generates atmospheric surface level pressure, wind-stress and heat-flux anomalies. In turn, those surface level atmospheric anomalies induce mutually coherent SST and sea level anomalies over the North Atlantic area, and locally, in the Bay of Biscay area. These anomalies, both locally over the Bay of Biscay area and over the North Atlantic, are in agreement with several mechanisms that have separately been related to the variability of the IPC-Navidad, i.e. the south-westerly winds, the joint effect of baroclinicity and relief (JEBAR) effect, the topographic beta effect and a weakened North Atlantic gyre.

The IRYSS-COPD appropriateness study: objectives, methodology, and description of the prospective cohort

Background: Patients with chronic obstructive pulmonary disease (COPD) often experience exacerbations of the disease that require hospitalization. Current guidelines offer little guidance for identifying patients whose clinical situation is appropriate for admission to the hospital, and properly developed and validated severity scores for COPD exacerbations are lacking. To address these important gaps in clinical care, we created the IRYSS-COPD Appropriateness Study. Methods/Design: The RAND/UCLA Appropriateness Methodology was used to identify appropriate and inappropriate scenarios for hospital admission for patients experiencing COPD exacerbations. These scenarios were then applied to a prospective cohort of patients attending the emergency departments (ED) of 16 participating hospitals. Information was recorded during the time the patient was evaluated in the ED, at the time a decision was made to admit the patient to the hospital or discharge home, and during follow-up after admission or discharge home. While complete data were generally available at the time of ED admission, data were often missing at the time of decision making. Predefined assumptions were used to impute much of the missing data. Discussion: The IRYSS-COPD Appropriateness Study will validate the appropriateness criteria developed by the RAND/UCLA Appropriateness Methodology and thus better delineate the requirements for admission or discharge of patients experiencing exacerbations of COPD. The study will also provide a better understanding of the determinants of outcomes of COPD exacerbations, and evaluate the equity and variability in access and outcomes in these patients.

An investigation of imputation methods for discrete databases and multi-variate time series

When it comes to information sets in real life, often pieces of the whole set may not be available. This problem can find its origin in various reasons, describing therefore different patterns. In the literature, this problem is known as Missing Data. This issue can be fixed in various ways, from not taking into consideration incomplete observations, to guessing what those values originally were, or just ignoring the fact that some values are missing. The methods used to estimate missing data are called Imputation Methods. The work presented in this thesis has two main goals. The first one is to determine whether any kind of interactions exists between Missing Data, Imputation Methods and Supervised Classification algorithms, when they are applied together. For this first problem we consider a scenario in which the databases used are discrete, understanding discrete as that it is assumed that there is no relation between observations. These datasets underwent processes involving different combina- tions of the three components mentioned. The outcome showed that the missing data pattern strongly influences the outcome produced by a classifier. Also, in some of the cases, the complex imputation techniques investigated in the thesis were able to obtain better results than simple ones. The second goal of this work is to propose a new imputation strategy, but this time we constrain the specifications of the previous problem to a special kind of datasets, the multivariate Time Series. We designed new imputation techniques for this particular domain, and combined them with some of the contrasted strategies tested in the pre- vious chapter of this thesis. The time series also were subjected to processes involving missing data and imputation to finally propose an overall better imputation method. In the final chapter of this work, a real-world example is presented, describing a wa- ter quality prediction problem. The databases that characterized this problem had their own original latent values, which provides a real-world benchmark to test the algorithms developed in this thesis.