Standard factorial designs may sometimes be inadequate for experiments that aim to estimate a generalized linear model, for example, for describing a binary response in terms of several variables. A method is proposed for finding exact designs for such experiments which uses a criterion that allows for uncertainty in the link function, the linear predictor or the model parameters, together with a design search. Designs are assessed and compared by simulation of the distribution of efficiencies relative to locally optimal designs over a space of possible models. Exact designs are investigated for two applications and their advantages over factorial and central composite designs are demonstrated.

The construction of decision-theoretic Bayesian designs for realistically-complex nonlinear models is computationally challenging, as it requires the optimization of analytically intractable expected utility functions over high-dimensional design spaces. We provide the most general solution to date for this problem through a novel approximate coordinate exchange algorithm. This methodology uses a Gaussian process emulator to approximate the expected utility as a function of a single design coordinate in a series of conditional optimization steps. It has flexibility to address problems for any choice of utility function and for a wide range of statistical models with different numbers of variables, numbers of runs and randomization restrictions. In contrast to existing approaches to Bayesian design, the method can find multi-variable designs in large numbers of runs without resorting to asymptotic approximations to the posterior distribution or expected utility. The methodology is demonstrated on a variety of challenging examples of practical importance, including design for pharmacokinetic models and design for mixed models with discrete data. For many of these models, Bayesian designs are not currently available. Comparisons are made to results from the literature, and to designs obtained from asymptotic approximations.

No abstract

Halite particles about 2 micrometers in size were collected by a quartz crystal microbalance cascade impactor from the El Chichón eruption cloud in the lower stratosphere during April and May 1982. These particles are probably derived from the erupted chloride-rich, alkalic magma. Enrichments of hydrogen chloride and increases in optical depolarization in the eruption cloud observed by lidar measurements may reflect the influence of the halite particles. There is evidence that the halite particles reacted with sulfuric acid after about 1 month, releasing gaseous hydrogen chloride, which can influence the catalytic destruction of ozone in the stratosphere.

Various design and model selection methods are available for supersaturated designs having more factors than runs but little research is available on their comparison and evaluation. In this paper, simulated experiments are used to evaluate the use of E(s2)-optimal and Bayesian D-optimal designs, and to compare three analysis strategies representing regression, shrinkage Various design and model selection methods are available for supersaturated designs having more factors than runs but little research is available on their comparison and evaluation. In this paper, simulated experiments are used to evaluate the use of E(s 2 )-optimal and Bayesian D-optimal designs, and to compare three analysis strategies representing regression, shrinkage and a novel model-averaging procedure. Suggestions are made for choosing the values of the tuning constants for each approach. Findings include that (i) the preferred analysis is via shrinkage; (ii) designs with similar numbers of runs and factors can be effective for a considerable number of active effects of only moderate size; and (iii) unbalanced designs can perform well. Some comments are made on the performance of the design and analysis methods when effect sparsity does not hold.

Highly resolved aerosol size distributions measured from high-altitude aircraft can be used to describe the effect of the 1991 eruption of Mount Pinatubo on the stratospheric aerosol. In some air masses, aerosol mass mixing ratios increased by factors exceeding 100 and aerosol surface area concentrations increased by factors of 30 or more. Increases in aerosol surface area concentration were accompanied by increases in chlorine monoxide at mid-latitudes when confounding factors were controlled. This observation supports the assertion that reactions occurring on the aerosol can increase the fraction of stratospheric chlorine that occurs in ozone-destroying forms.

No abstract

A ground‐based lidar system at NASA Langley Research Center in Hampton, Virginia, has monitored the stratospheric aerosol vertical distribution and loading since 1974. The eruption of Mt. Pinatubo in June 1991 produced the largest enhancement of stratospheric aerosol loading ever observed by lidar over this mid‐latitude location. Low altitude layers (<20 km) were the first to arrive over Hampton in early August, the result of transport associated with a tropospheric anticyclonic cell over North America. The maximum peak scattering ratio, 34 at 22.4 km, and the maximum stratospheric integrated backscatter of 0.0053 sr−1, both at 694 nm, observed since the eruption were measured on February 20, 1992. After decreasing during the spring and summer of 1992, the aerosol burden increased significantly during the winter of 1992–3, evidence of poleward winter transport from the equatorial reservoir. Over the period from February 1992 to February 1994, the stratospheric aerosol loading decreased with an average 1/e decay time of 10.1 months. The vertical distribution, intensity, and transport of Pinatubo aerosols over Hampton, Virginia, are described in detail and compared with similar measurements after El Chichon.