Field‐Scale Application of Three Types of Neural Networks to Predict Ground‐Water Levels1

Abstract: In this paper, a field‐scale applicability of three forms of artificial neural network algorithms in forecasting short‐term ground‐water levels at specific control points is presented. These algorithms are the feed‐forward back propagation (FFBP), radial basis networks (RBN), and generalized regression networks (GRN). Ground‐water level predictions from these algorithms are in turn to be used in an Optimized Regional Operations Plan that prescribes scheduled wellfield production for the coming four weeks. These models are up against each other for their accuracy of ground‐water level predictions on lead times ranging from a week to four weeks, ease of implementation, and execution times (mainly training time). In total, 208 networks of each of the three algorithms were developed for the study. It is shown that although learning algorithms have emerged as a viable solution at field scale much larger than previously studied, no single algorithm performs consistently better than others on all the criteria. On average, FFBP networks are 20 and 26%, respectively, more accurate than RBN and GRN in forecasting one week ahead water levels and this advantage drops to 5 and 9% accuracy in forecasting four weeks ahead water levels, whereas GRN posted a training time that is only 5% of the training time taken by that of FFBP networks. This may suggest that in field‐scale applications one may have to trade between the type of algorithm to be used and the degree to which a given objective is honored.     

Spatiotemporal analysis of traffic emissions in over 5000 municipal districts in Brazil

Exposure to traffic emission is harmful to human health. Emission inventories are essential to public health policies aiming at protecting human health, especially in areas with incomplete or nonexistent air pollution monitoring networks. In Brazil, for example, only 1.7% of municipal districts have a monitoring network, and only a few studies have reported data on vehicle emission inventories. No studies have presented emission inventories by municipality. In this study, we predicted vehicular emissions for 5570 municipal districts in Brazil during the period 2001–2012. We used a top-down method to estimate emissions. Carbon dioxide (CO₂) is the pollutant with the highest emissions, with approximately 190 million tons per year during the period 2001–2012). For the other traffic-related pollutants, we predicted annual emissions of 1.5 million tons for carbon monoxide (CO), 1.2 million tons of nitrogen oxides (NO_x), 209,000 tons of nonmethane hydrocarbons (NMHC), 58,000 tons of particulate matter (PM), and 42,000 tons for methane (CH₄). From 2001 to 2012, CO, NMHC, and PM emissions decreased by 41, 33, and 47%, respectively, whereas those CH₄, NO_x, and CO₂ increased by 2, 4, and 84%, respectively. We estimated uncertainties in our study and found that NO_x was the pollutant with the lowest percentage difference, 8%, and NMHC with the highest one, 30%. For CO, CH₄, CO₂, and PM, the values were 22, 14, 21, and 20%, respectively. Finally, we found that during 2001 and 2012 emissions increased in the Northwest and Northeast. In contrast, pollutant emissions, except for CO₂, decreased in the Southeast, South, and part of Midwest. Our predictions can be critical to efforts developing cost-effective public policies tailored to individual municipal districts in Brazil.

Implications: Emission inventories may be an alternative approach to provide data for air quality forecasting in areas where air quality data are not available. This approach can be an effective tool in developing spatially resolved emission inventories.     

Indoor/ambient residential air toxics results in rural western Montana

Indoor and ambient concentrations of 21 volatile organic compounds (including 14 hazardous air pollutants) were measured in the homes of nearly 80 western Montana (Missoula) high school students as part of the 'Air Toxics Under the Big Sky' program during the 2004/2005 and 2005/2006 school years. Target analytes were measured using low flow air sampling pumps and sorbent tubes, with analysis of the exposed samples by thermal desorption/gas chromatography/mass spectrometry (TD/GC/MS). The results reported here present the findings of the first indoor/ambient air toxics monitoring program conducted in a semi-rural valley location located in the Northern Rocky Mountain/Western Montana region. Of all of the air toxics quantified in this study, toluene was found to be the most abundant compound in both the indoor and ambient environments during each of the two school years. Indoor log-transformed mean concentrations were found to be higher when compared with ambient log-transformed mean concentrations at P < 0.001 for the majority of the compounds, supporting the results of previous studies conducted in urban areas. For the air toxics consistently measured throughout this program, concentrations were approximately six times higher inside the student's homes compared to those simultaneously measured directly outside their homes. For the majority of the compounds, there were no significant correlations between indoor and ambient concentrations.     

Temporal Trends of Young-of-Year Fishes in Lake Erie and Comparison of Diel Sampling Periods

We explored temporal trends of young-of-year (YOY) fishes caught in bottom trawl hauls at an established offshore monitoring site in Lake Erie in fall during 1961–2001. Sampling was conducted during morning, afternoon, and night in each year. Catches per hour (CPH) of alewife (Alosa pseudoharengus) YOY were relatively low and exhibited no temporal trend. This result was consistent with the species’ intolerance to Lake Erie’s adverse winter water temperatures. Gizzard shad (Dorosoma cepedianum) YOY decreased sharply after 1991, which was consistent with recent oligotrophication of the lake. Following the establishment in 1979 and rapid increase of white perch (Morone americana) YOY, white bass (Morone chrysops) and freshwater drum (Aplodinotus grunniens) YOY decreased. Trout-perch (Percopsis omiscomaycus) YOY decreased during 1986–1991, but recovered to previous levels during 1991–2001. The recovery coincided with the resurgence of mayflies (Ephemoptera) in the lake. CPH of spottail shiner (Notropis hudsonius) and emerald shiner (N. atherinoides) YOY exhibited no temporal trend between 1961 and the late 1970s to early 1980s. CPH of yellow perch (Perca flavescens) YOY decreased during 1961–1988, and walleye (Sander vitreum) YOY increased overall during the time series. These observations were consistent with published studies of adults in the region. CPH of 4 of the 10 species of YOY considered were greatest during night. CPH for walleye YOY was higher in the morning than in the afternoon, but there was no significant difference between night and morning abundances. The results suggest that (1) CPH of YOY fishes may be a useful monitoring tool for Lake Erie, and (2) offshore monitoring programs that do not include night sampling periods may underestimate recruitment for several common species.     

Analysis of two-state multivariate phenotypic change in ecological studies

Analyses of two-state phenotypic change are common in ecological research. Some examples include phenotypic changes due to phenotypic plasticity between two environments, changes due to predator/non-predator character shifts, character displacement via competitive interactions, and patterns of sexual dimorphism. However, methods for analyzing phenotypic change for multivariate data have not been rigorously developed. Here we outline a method for testing vectors of phenotypic change in terms of two important attributes: the magnitude of change (vector length) and the direction of change described by trait covariation (angular difference between vectors). We describe a permutation procedure for testing these attributes, which allows non-targeted sources of variation to be held constant. We provide examples that illustrate the importance of considering vector attributes of phenotypic change in biological studies, and we demonstrate how greater inference can be made than by evaluating variance components with MANOVA alone. Finally, we consider how our method may be extended to more complex data.     

Relatedness,recognition errors,and colony fusion in the termite <Emphasis Type="Italic">Nasutitermes corniger</Emphasis>

Loss of aggression between social groups can have far-reaching effects on the structure of societies and populations. We tested whether variation in the genetic structure of colonies of the termite Nasutitermes corniger affects the probability of aggression toward non-nestmates and the ability of unrelated colonies to fuse. We determined the genotypes of workers and soldiers from 120 colonies at seven polymorphic microsatellite loci. Twenty-seven colonies contained offspring of multiple founding queens or kings, yielding an average within-colony relatedness of 0.33. Genotypes in the remaining 93 colonies were consistent with reproduction by a single queen and king or their progeny, with an average within-colony relatedness of 0.51. In standardized assays, the probability of aggression between workers and soldiers from different colonies was an increasing function of within-colony relatedness. The probability of aggression was not affected significantly by the degree of relatedness between colonies, which was near zero in all cases, or by whether the colonies were neighbors. To test whether these assays of aggression predict the potential for colony fusion in the field, we transplanted selected nests to new locations. Workers and soldiers from colonies that were mutually tolerant in laboratory assays joined their nests without fighting, but workers and soldiers that were mutually aggressive in the assays initiated massive battles. These results suggest that the presence of multiple unrelated queens or kings promotes recognition errors, which can lead to the formation of more complex colony structures.     

Assessment of metal contaminations leaching out from recycling plastic bottles upon treatments

Background, aims, and scope  

Heavy metal contaminants in environment, especially in drinking water, are always of great concern due to their health impact. Due to the use of heavy metals as catalysts during plastic syntheses, particularly antimony, human exposure to metal release from plastic bottles has been a serious concern in recent years. The aim and scope of this study were to assess metal contaminations leaching out from a series of recycling plastic bottles upon treatments.     

Tetrabromobisphenol A (TBBPA) and hexabromocyclododecanes (HBCDs) in tissues of humans, dolphins, and sharks from the United States

Concentrations of tetrabromobisphenol A (TBBPA) and alpha-, beta-, and gamma-isomers of hexabromocyclododecanes (HBCDs) were determined by liquid chromatography-tandem mass spectrometry (LC-ESI-MS/MS) in human adipose tissue obtained in New York City, and in three marine top-level predators--bottlenose dolphin, bull shark, and Atlantic sharpnose shark--collected from coastal waters of Florida, USA. The overall mean concentrations (mean+/-SD) of TBBPA and HBCDs were 0.048+/-0.102 and 0.333+/-0.571 ng/g lipid wt in human adipose tissue samples, 1.2+/-3 and 7.38+/-18 ng/g lipid wt in bottlenose dolphin blubber, 9.5+/-12 and 77.7+/-128 ng/g lipid wt in bull shark muscle, and 0.872+/-0.5 and 54.5+/-88 ng/g lipid wt in Atlantic sharpnose shark muscle. Overall mean concentrations of HBCDs were 5-10-fold higher than mean TBBPA concentrations, in all of the samples analyzed. The highest concentrations of TBBPA and HBCDs were detected in the bull shark muscle at concentrations of 35.6 and 413 ng/g, lipid wt, respectively. Concentrations of TBBPA and HBCDs, after log-transformation, were significantly correlated with each other in human adipose tissue and bottlenose dolphin blubber, but not in bull shark muscle samples. In the human adipose tissue samples, the concentrations of HBCDs were 3-4 orders of magnitude lower than the concentrations of polybrominated diphenyl ethers (PBDEs) previously reported for the same set of tissue samples. Concentrations of HBCDs in human samples from the United States were 1-5-fold lower than the concentrations reported from several European countries. HBCD concentrations in bottlenose dolphins from the United States were 1-2 orders of magnitude lower than the concentrations reported for other cetacean species from Europe. The present report is the first to determine levels of TBBPA and HBCDs in humans, bottlenose dolphins, and sharks from the United States.     

Modeling and experiments of polydisperse particle clouds

A model for polydisperse particle clouds has been developed in this study. We extended the monodisperse particle cloud model of Lai et al. (Environ Fluid Mech 13(5):435–463, 2013) to the case of polydisperse particles. The particle cloud is first considered to be a thermal or buoyant vortex ring, with the thermal induced velocity field modeled by an expanding spherical Hill’s vortex. The buoyancy of the composite thermal is assumed to be the sum of buoyancy contributed by the all particles inside the thermal. Individual particles (of different particle properties) in the cloud are then tracked by the particle tracking equation using the computed induced velocity field. The turbulent dispersion effect is also accounted for by using a random walk model. Experiments of polydisperse particle clouds were carried out to validate the model. The agreement between model predictions and experiments was reasonable. We further validate our model by comparing it with the LES study of Wang et al. (J Hydraul Eng ASCE 141(7):06015006, 2014). The limitations of our model are then discussed with reference to the comparison. Overall, although some flow details are not captured by our model, the simplicity and generality of the model makes it useful in engineering applications.     

Weather effects on avian breeding performance and implications of climate change

The influence of recent climate change on the world's biota has manifested broadly, resulting in latitudinal range shifts, advancing dates of arrival of migrants and onset of breeding, and altered community relationships. Climate change elevates conservation concerns worldwide because it will likely exacerbate a broad range of identified threats to animal populations. In the past few decades, grassland birds have declined faster than other North American avifauna, largely due to habitat threats such as the intensification of agriculture. We examine the effects of local climatic variations on the breeding performance of a bird endemic to the shortgrass prairie, the Lark Bunting (Calamospiza melanocorys) and discuss the implications of our findings relative to future climate predictions. Clutch size, nest survival, and productivity all positively covaried with seasonal precipitation; yet relatively intense daily precipitation events temporarily depressed daily survival of nests. Nest survival was positively related to average temperatures during the breeding season. Declining summer precipitation may reduce the likelihood that Lark Buntings can maintain stable breeding populations in eastern Colorado although average temperature increases of up to 3 degrees C (within the range of this study) may ameliorate declines in survival expected with drier conditions. Historic climate variability in the Great Plains selects for a degree of vagility and opportunism rather than strong site fidelity and specific adaptation to local environments. These traits may lead to northerly shifts in distribution if climatic and habitat conditions become less favorable in the drying southern regions of the Great Plains. Distributional shifts in Lark Buntings could be constrained by future changes in land use, agricultural practices, or vegetative communities that result in further loss of shortgrass prairie habitats.