David S. Bowles

Multivariate nonparametric resampling scheme for generation of daily weather variables

A nonparametric resampling technique for generating daily weather variables at a site is presented. The method samples the original data with replacement while smoothing the empirical conditional distribution function. The technique can be thought of as a smoothed conditional Bootstrap and is equivalent to simulation from a kernel density estimate of the multivariate conditional probability density function. This improves on the classical Bootstrap technique by generating values that have not occurred exactly in the original sample and by alleviating the reproduction of fine spurious details in the data. Precipitation is generated from the nonparametric wet/dry spell model as described in Lall et al. [1995]. A vector of other variables (solar radiation, maximum temperature, minimum temperature, average dew point temperature, and average wind speed) is then simulated by conditioning on the vector of these variables on the preceding day and the precipitation amount on the day of interest. An application of the resampling scheme with 30 years of daily weather data at Salt Lake City, Utah, USA, is provided.

Spatial rainfall estimation by linear and non-linear co-kriging of radar-rainfall and raingage data

The feasibility of linear and nonlinear geostatistical estimation techniques for optimal merging of rainfall data from raingage and radar observations is investigated in this study by use of controlled numerical experiments. Synthetic radar and raingage data are generated with their hypothetical error structures that explicitly account for sampling characteristics of the two sensors. Numerically simulated rainfall fields considered to be ground-truth fields on 4×4 km grids are used in the generation of radar and raingage observations. Ground-truth rainfall fields consist of generated rainfall fields with various climatic characteristics that preserve the space-time covariance function of rainfall events in extratropical cyclonic storms. Optimal mean areal precipitation estimates are obtained based on the minimum variance, unbiased property of kriging techniques under the second order homogeneity assumption of rainfall fields. The evaluation of estimated rainfall fields is done based on the refinement of spatial predictability over what would be provided from each sensor individually. Attention is mainly given to removal of measurement error and bias that are synthetically introduced to radar measurements. The influence of raingage network density on estimated rainfall fields is also examined.

GIS Model for Estimating Dam Failure Life Loss

This paper describes a modular geographical information system (GIS) modelling system for estimating potential loss of life from natural and dam -failure floods. The model provides life-loss estimates for use in dam safety risk assessments. It can also be used to explore options for improving the effectiveness of detection, notification, warning, emergency planning and emergency response. The simulation modelling system comprises the following internal modules: 1) loss of shelter, including prediction of building performance, 2) warning and evacuation, and 3) loss of life based on empirical relationships developed from a wide range of case histories and described in our earlier work [McClelland and Bowles 2000]. Estimated Flood Routing conditions are obtained from an external dam break and flood routing model. Other inputs include a digital elevation model, road layout, and data on populations at risk from readily available sources. Application of the modelling system is demonstrated for sudden and delayed earthquake failures for a large embankment dam. Both the existing warning and evacuation system and an improved system are run. An Uncertainty Mode of the model is also illustrated. Comparisons with the Graham (1999) Method are included. Plans for further model development are summarised, including a Simplified Mode for making preliminary life-loss estimates.

Initial model choice: An operational comparison of stochastic streamflow models for drought

The performance of five stochastic models for generating annual streamflow sequences is evaluated based on applications to four Utah streams. Model performance is evaluated in terms of preservation of annual persistence statistics; cost and ease of model use; magnitude of the economic regret associated with drought-related agricultural losses; and comparison of reservoir capacity and critical drought design parameters. The ARMA-Markov and ARMA models are judged to be best overall in terms of preserving short- and long-term persistence. The broken line model is judged to be best in terms of minimizing economic regret. An initial model choice procedure is proposed.

Risk Assessment of Success Dam, California: Flood Related Potential Failure Modes

Potential seismic deficiencies of Success Dam, California have been identified and a remediation plan for the dam is in the process of development. The dam is currently operated under reservoir pool elevation restriction to reduce risk until the long-term remediation is implemented. Justification of operation restrictions and establishing the priority of remediation activities required preparation of a baseline risk assessment. Seepage and piping failure modes made a significant contribution to the baseline risk, but were little affected by the operating restrictions that were used to reduce the earthquake-related failure modes. This paper describes the role of the flood-induced failure modes in the risk assessment and particularly the evaluation of the piping and seepage failure modes. The USACE Piping and Seepage Toolbox was used as an aid in estimating the probability of failure by the various piping and seepage failure modes.

A Comparison of Methods for Risk Assessment of Dams

Risk-based procedures for assessing appropriate safety levels for new and existing dams have been proposed for use in planning and design of dams and screening of unsafe dams. The risk assessment framework and its application to dams is presented. Approaches for estimating the various types of probabilities and consequences needed to perform a comprehensive dam risk assessment are described. Several methods for planning, screening, and design level risk assessment are summarized and compared. The lecture closes with a discussion of the advantages of, commonly stated objections to, and some unresolved issues related to, risk assessment of dams.

SPATIAL ESTIMATION TECHNIQUES FOR PRECIPITATION ANALYSIS - APPLICATION TO A REGION IN INDIA

Precipitation data from northern India was analyzed using three different spatial estimation techniques, viz. Kriging, Lowess and Smoothing Spline ANOVA. Annual precipitation was considered as a function of latitude and longitude. Various quantitative measures were considered for comparison for these techniques.

Comment on “Stochastic interpolation of rainfall data from rain gages and radar using cokriging: 1, Design of experiments,” and “2, Results,” by Dong‐Jun Seo et al.

The papers by Seo et aI. [1990a, b] are an interesting application of cokriging and the authors are to be commended for that application. However, one of the important obligations of authors and of referees is to discern the extent to which the results are new and to ensure that relevant literature is cited. Unfortunately, these papers failed significantly in acknowledging prior work. Specifically, the development of cokriging, including universal and disjunctive cokriging, is presented as new work whereas it had already been in print for some time before this paper. In particular, see Myers [1982, 1983, 1984, 1985, 1988a, b, c, 1989], Cart et aI. [1985, 1987], and Cart and Myers [I984, 1985]. Not one of these papers is cited. Mathematical Geology is one of the principal journals publishing geostatistical papers, yet there is only one citation to a paper in Mathematical Geology (and that one is unrelated to cokriging), there are no citations to papers in the proceedings from the second international conference on geostatistics (Lake Tahoe, 1983) or the third conference (Avignon 1988). These are significant publications for geostatistics; both authors and reviewers should be aware of them. More specifically, the complete development of the general form of the cokriging estimator including the case of universal cokriging was already given by Myers [1982]. Various additional details were given in other papers listed above. The general case of disjunctive cokriging was given by Myers [1988c]. The authors have only given much more limited results; in particular, they have only considered the case of the estimation of one variable. However, the equations for this special case are not really different (or simpler) from the general case, and an emphasis on the one-variable case obscures important aspects of the development, specifically, the appropriate positive definiteness conditions for