Error Propagation

This paper analyses the survey configuration, instrumentation and error propagation in the control survey technique known as "wall station traversing". It utilises wall mounted survey points as an alternative to roof (backs) mounted points. Recently, this technique has gained widespread acceptance in underground metalliferous mines in Western Australia. The authors analyse and compare error propagation of the "wall stations" technique

The reliability of file systems depends in part on how well they propagate errors. We develop a static analysis technique, EDP, that analyzes how file systems and storage device drivers propagate error codes. Running our EDP analysis on all file systems and 3 major storage device drivers in Linux 2.6, we find that errors are often incorrectly propagated; 1153 calls (13%) drop an error code without handling it.

The overarching goal of the research described in this dissertation is to understand the hydrologic implications of error propagation from satellite precipitation products to hydrologic simulations. The complex interaction between precipitation error and corresponding hydrologic response is examined following a numerical- and an analytical-based approach. The application of a hydrologic model forced by various satellite precipitation products is adopted as the numerical-based framework that is used to identify the properties of error propagation with respect to a number of factors (e.g. basin scale, seasonality, severity of rainfall and flow). The investigation is conducted in complex terrain basins of the Eastern Italian Alps. Results show better consistency between gauges for events occurred over larger scale basins during warm season months that are associated with moderate intensity of rain and flow rate. Furthermore, an event -based error analysis is conducted focusing on the evaluation of satellite-simulated flood event characteristics for different flood types. Results revealed that on average systematic rainfall error is reduced in simulated runoff, highlighting the dampening effect on error during the rainfall-runoff transformation. The dampening effect on random error was decreasing with increasing runoff coefficient. In addition to the numerical investigation, an analytical framework is developed to decompose the error propagation into space and time components. This essentially allows to assess the relative contribution of the different processes of catchment flood response on error propagation. Demonstration of the analytical framework is conducted based on 180 rainfall-runoff events that occurred over the Tar River basin in North Carolina. It is shown that error in timing of flood event is attributed equally to error in runoff generation and routing time. Error in hydrograph shape is mainly controlled by the error in the variability of runoff generation time while error in flood volume is predominantly controlled by the error in rainfall volume. Overall, these investigations provide important information for the hydrologic modelers to choose the appropriate precipitation products for the hydrologic-related practice. It also serves as guidance for the satellite precipitation-product developers on the designs of more advance retrieval algorithms.

I discuss our current understanding of parton distributions. I begin with the underlying theoretical framework, and the way in which different data sets constrain different partons, highlighting recent developments. The methods of examining the uncertainties on the distributions and those physical quantities dependent on them is analysed. Finally I look at the evidence that additional theoretical corrections beyond NLO perturbative QCD may be necessary, what type of corrections are indicated and the impact these may have on the uncertainties.

This thesis extends the error propagation network to deal with time varying or dynamic patterns.  Examples are given of supervised, reinforcement driven and unsupervised learning.

Chapter 1 presents an overview of connectionist models.

Chapter 2 introduces the error propagation algorithm for general node types.

Chapter 3 discusses the issue of data representation in connectionist models.

Chapter 4 describes the use of several types of networks applied to the problem of the recognition of steady state vowels from multiple speakers.

Chapter 5 extends the error propagation algorithm to deal with time varying input.  Three possible architectures are explored which deal with learning sequences of known length and sequences of unknown and possibly indefinite length.  Several simple examples are given.

Chapter 6 describes the use of two dynamic nets to form a speech coder.  The popular method of Differential Pulse Code Modulation for speech coding employs two linear filters to encode and decode speech.  By generalising these to non-linear filters, implemented as dynamic nets, a reduction in the noise imposed by a limited bandwidth channel is achieved.

Chapter 7 describes the application of a dynamic net to the recognition of a large subset of the phonemes of English from continuous speech.  The dynamic net is found to give a higher recognition rate both in comparison with a fixed window net and with the established k nearest neighbour technique.

Chapter 8 describes a further development of dynamic nets which allows them to be trained by a reinforcement signal which expresses the correctness of the output of the net.  Two possible architectures are given and an example of learning to play the game of noughts and crosses is presented.

The vast advancements in satellite rainfall observations over the last years have opened new horizons in hydrological applications at global scale. Specifically, newly available high resolution (4-8 km, 1-3 hourly) satellite precipitation estimates (HPE) have allowed researchers to consider their potential integration with hydrologic models for flood monitoring applications. However, performance evaluation of HPEs in cases of major flash flood-inducing

This work focuses on the stochastic properties of boresight determi nation between a strapdown IMU and a frame- based imaging sensor. The core of the stochastic model is a rigorous error propagation of the estimated input accuracies and their correlations. This subsequently yields mor e realistic estimates of boresight accuracy for every image event. The nature of INS/GPS integration makes

Although environmental modelling is increasingly performed within a GIS framework, analysis of the associated error is far from routine, and rarely presented with the results. An important benefit of performing error analysis is its value in determining which elements of a vulnerability assessment framework need improving. With this in mind, it was decided to examine the extent to which error might propagate through a model of groundwater vulnerability to pesticide contamination. A pesticide leaching model was developed and incorporated into an assessment of groundwater contamination risk from normal agricultural use of the herbicide isoproturon, in a 30 km x 37 km area of river catchment to the north-west of London, England. The model, which comprised two main components accounting for (i) degradation and (ii) attenuation of the pesticide, was based on conventional contaminant transport calculations, combined with existing soil, rainfall, hydrogeological and depth to water table da...

The autocorrelation function of error propagation for a one-tap decision-feedback equalizer is derived. The autocorrelation is shown to be a geometric sequence which may or may not have alternating sign depending on the channel impulse response

A new, high-resolution and high-precision geoid has been computed for the whole of Canada and part of the U.S., ranging from 35°N to about 90°N in latitude and 210°E to 320°E in longitude. The OSU91A geopotential model complete to degree and order 360 was combined with a 5′ × 5′ mean gravity anomaly grid and 1km × 1km topographical information to generate the geoid file. The remove-restore technique was adopted for the computation of terrain effects by Helmert's condensation reduction. The contribution of the local gravity data to the geoid was computed strictly by the 1D-FFT technique, which allows for the evaluation of the discrete spherical Stokes integral without any approximation, parallel by parallel. The indirect effects of up to second order were considered. The internal precision of the geoid, i.e. the contribution of the gravity data and the model coefficients noise, was also evaluated through error propagation by FFT. In a relative sense, these errors seem to agree quite well with the external errors and show clearly the weak areas of the geoid which are mostly due to insufficient gravity data coverage. Comparison of the gravimetric geoid with the GPS/levelling-derived geoidal heights of eight local GPS networks with a total of about 900 stations shows that the absolute agreement with respect to the GPS/levelling datum is generally better than 10 cm RMS and the relative agreement ranges, in most cases, from 4 to 1 ppm over short distances of about 20 to 100km, 1 to 0.5 ppm over distances of about 100 to 200 km, and 0.5 to 0.1 ppm for baselines of 200 to over 1000 km. Other existing geoids, such as UNB90, GEOID90 and GSD91, were also included in the comparison, showing that the new geoid achieves the best agreement with the GPS/levelling data.