Forecasting peak daily ozone levels--I. A regression with time series errors model having a principal component trigger to fit 1991 ozone levels

This research was motivated by the need to warn the population of Milwaukee, WI, on high-ozone days. A statistical model for the peak daily 1-hr ozone level is proposed. A Regression with Time Series Errors (RTSE) model, which includes a principal component (PC) trigger, is the basis for forecasting the peak daily 1-hr ozone level. The RTSE model, with a PC trigger, is first employed to estimate daily peak ozone measured at the University of Wisconsin, Milwaukee-North (UWM-N), during the 1991 ozone season. The RTSE model uses peak daily temperature, morning vector average wind direction, and the PC trigger as predictor variables. The PC trigger was designed to summarize atmospheric circumstances when peak ozone was greater than 100 parts per billion (ppb). It is verified that the RTSE model, with a PC trigger, significantly improves the prediction of peak daily ozone, particularly peak ozone greater than 100 ppb. In comparison with the RTSE model without the PC trigger, the RTSE model with a PC trigger raised the R2 from 0.680 to 0.809. It is suggested that the RTSE model, with the PC trigger, is an adequate statistical model that has the potential for real-time ozone forecasting.     

Determination of the source areas contributing to regionally high warm season PM2.5 in eastern North America

An ensemble-trajectory analysis technique known as Quantitative Transport Bias Analysis was applied to determine which geographic areas systematically contributed to above- and below-average fine particle mass (PM2.5) over eastern North America. Six-hour average measurements from 12 rural or suburban locations in eastern North America collected using a tapered element oscillating microbalance were individually associated with corresponding 3-day back-trajectories for the warm seasons (May-September) of 2000 and 2001. Much of the populated areas of northeastern North America were implicated in the build-up of PM2.5 to above-average concentrations. The finer structure of the Quantitative Transport Bias Analysis pattern indicated that transport from the Ohio River Valley, particularly the eastern portion of this area, was most often associated with the highest PM2.5 concentrations. In addition, air masses originating over a relatively large area from southeast Ohio to the western part of Virginia and the western Kentucky to central Tennessee area tended to result in relatively high PM2.5 concentrations over northeastern North America. These observation-based findings were consistent with the spatial distribution of the main sulfur dioxide emissions sources and the major oxides of nitrogen point sources.     

Preliminary on-road measurement of the effect of oxygenated fuel on CO emissions near Las Vegas, Nevada

A preliminary measurement of an oxygenated fuel effect for in-use vehicles travelling at freeway speed was conducted near Las Vegas, NV, in January 1991 and May 1992. The experimental design was based on two principal factors: (1) the large volume of traffic that visits Las Vegas from Southern California on three-day holiday weekends and (2) the fact that, at the time of the study, the Las Vegas area used oxygenated fuels in the winter and Southern California did not. Measurements were conducted at carefully selected sites 20 km southwest of Las Vegas near Sloan, NV, and were accomplished with the Fuel Efficiency Automoblie Test (FEAT) remote sensor developed at the University of Denver. The January 1991 measurements were made during the Las Vegas oxyfuel season, while the May 1992 control measurements were made outside the oxyfuel season. Over 24,500 individual CO concentrations were measured; registration data from over 5,500 of these vehicles were obtained from the license plate numbers. After corrections for differences in velocity and mean age, the Las Vegas outbound (oxyfuel) CO emissions on Monday morning of the January holiday weekend showed a difference of -18% +/- 11% compared to the inbound (non-oxyfuel) CO emissions on Friday evening preceding the holiday weekend.     

Distribution of trace element pollutants in a contaminated grassland ecosystem established on metalliferous fluorspar tailings. 2: Zinc

Concentrations of zinc in vegetation, invertebrates and small mammals in a grassland ecosystem derived from a mine waste vegetation scheme indicated the high concentration of zinc in the original tailings. However, the considerable differences in soil and vegetation zinc concentrations between the tailings and an uncontaminated site were not reflected in the invertebrate and small mammal communities. Invertebrate zinc levels and total body concentrations in Microtus agrestis L. (field vole) and Sorex araneus L. (common shrew) suggested a notable buffering of zinc absorption and tissue accumulation in the presence of substantially elevated dietary zinc. In the small mammals the skeleton appeared to function as the major storage site for zinc, possibly contributing to homeostasis in the soft tissues.     

Storm flow export of metolachlor from a coastal plain watershed

During an 18-month (1994-1995) survey of the surface water in an Atlantic Coastal Plain watershed, metolachlor was most frequently detected during storm flow events. Therefore, a sampling procedure, focused on storm flow, was implemented in June of 1996. During 1996, three tropical cyclones made landfall within 150 km of the watershed. These storms, as well as several summer thunderstorms, produced six distinct storm flow events within the watershed. Metolachlor was detected leaving the watershed during each event. In early September, Hurricane Fran produced the largest storm flow event and accounted for the majority of the metolachlor exports. During the storm event triggered by Hurricane Fran, the highest daily average flow (7.5 m2 s-1) and highest concentration (5.1 micrograms L-1) ever measured at the watershed outlet were recorded. Storm flow exports leaving the watershed represented 0.1 g ha-1 or about 0.04% of active ingredient applied.     

Modelling Replicated Weed Growth Data using Spatially-varying Growth Curves

Weed growth in agricultural fields constitutes a major deterrent to the growth of crops, often resulting in low productivity and huge losses for the farmers. Therefore, proper understanding of patterns in weed growth is vital to agricultural research. Recent advances in Geographical Information Systems (GIS) now allow geocoding of agricultural data, which enable more sophisticated spatial analysis. Our current application concerns the development of statistical models for conducting spatial analysis of growth patterns in weeds. Our data comes from an experiment conducted in Waseca, Minnesota, that recorded growth of the weed Setariaspp. We capture the spatial variation in Setaria spp. growth using spatially-varying growth curves. An added challenge is that these designs are spatially replicated, with each plot being a lattice of sub-plots. Therefore, spatial variation may exist at different resolutions – a macro level variation between the plots and micro level variation between the sub-plots nested within each plot. We develop a Bayesian hierarchical framework for this setting. Flexible classes of models result which are fitted using simulation-based methods.