METHODOLOGIES FOR CALIBRATION AND PREDICTIVE ANALYSIS OF A WATERSHED MODEL1

ABSTRACT: The use of a fitted parameter watershed model to address water quantity and quality management issues requires that it be calibrated under a wide range of hydrologic conditions. However, rarely does model calibration result in a unique parameter set. Parameter nonuniqueness can lead to predictive nonuniqueness. The extent of model predictive uncertainty should be investigated if management decisions are to be based on model projections. Using models built for four neighboring watersheds in the Neuse River Basin of North Carolina, the application of the automated parameter optimization software PEST in conjunction with the Hydrologic Simulation Program Fortran (HSPF) is demonstrated. Parameter nonuniqueness is illustrated, and a method is presented for calculating many different sets of parameters, all of which acceptably calibrate a watershed model. A regularization methodology is discussed in which models for similar watersheds can be calibrated simultaneously. Using this method, parameter differences between watershed models can be minimized while maintaining fit between model outputs and field observations. In recognition of the fact that parameter nonuniqueness and predictive uncertainty are inherent to the modeling process, PEST's nonlinear predictive analysis functionality is then used to explore the extent of model predictive uncertainty.     

Respiration and lipid content of the Arctic copepod<Emphasis Type="Italic"> Calanus hyperboreus</Emphasis> overwintering 1 m above the seafloor at 2,300 m water depth in the Fram Strait

In August 2000 high concentrations of the dominant herbivorous copepod Calanus hyperboreus were detected in the Arctic Fram Strait, west of Spitsbergen, 1 m above the seafloor at 2,290 m water depth. Individuals from that layer were sampled by a hyper-benthic net attached to the frame of an epi-benthic sledge. For comparison, the vertical distribution of C. hyperboreus in the water column was studied simultaneously by a multiple opening/closing net haul from 2,250 m depth to the surface. Maximum abundance was found close to the surface with 6.6 and 10.0 ind. m^?3 at 0–50 m and 50–100 m depth, respectively. However, the major fraction of the population (>40%) occurred between 1,000 and 1,500 m depth. In the deepest layer (2,000–2,250 m) abundance measured 2.2 ind. m^?3 and was twice as high as between 100 and 1,000 m depth. In comparison to individuals from surface waters, copepods from the hyper-benthic layer were torpid and did not react to mechanical stimuli. Stage CV copepodids and females from the deep sample contained 4–10% less lipid and showed significantly reduced respiration rates of 0.24 and 0.26 ml O₂ h^?1 g^?1 dry mass (DM) as compared to surface samples (0.49 and 0.43 ml O₂ h^?1 g^?1 DM). All these observations indicate that the hyper-benthic part of the population had already started a dormant overwintering phase at great depth. Based on the lipid deposits and energy demands, the potential maximum duration of the non-feeding dormant phase was estimated at 76–110 days for females and at 98–137 days for CV copepodids, depending on what indispensable minimum lipid content was assumed. In any case, the estimated times could not meet the necessary requirements for a starvation period of >6 months until the next phytoplankton bloom in the following spring. The ecological implications of these results are discussed with respect to the life cycle and eco-physiological adaptations of C. hyperboreus to its high-Arctic habitat.     

Global sensitivity analysis of a process-based model for ammonia emissions from manure storage and treatment structures

The development of process-based models to estimate ammonia emissions from animal feeding operations (AFOSs) is sought to replace costly and time-consuming direct measurements. Critical to process-based model development is conducting sensitivity analysis to determine the input parameters and their interactions that contribute most to the variance of the model output. Global and relative sensitivity analyses were applied to a process-based model for predicting ammonia emissions from the surface of anaerobic lagoons for treating and storing manure. The objectives were to compare global sensitivity analysis (GSA) to relative (local) sensitivity analysis (RSA) on a process-based model for ammonia emissions. Based on the first-order coefficient, both GSA and RSA showed the model input parameters in order of importance in process model for ammonia emissions from lagoon surfaces were: (i) pH, (ii) lagoon liquid temperature, (iii) wind speed above the lagoon surface, and (iv) the concentration of ammoniacal nitrogen in the lagoon. The GSA revealed that interactions between model parameters accounted for over two-thirds of the model variance, a result that cannot be achieved using traditional RSA. Also, the GSA showed that parameter interactions involving liquid pH had more impact on the model output variance than the single parameters: (i) temperature, (ii) wind speed, or (iii) total ammoniacal nitrogen. This study demonstrates that GSA provides a more complete analysis of model input parameters and their interactions on the model output compared to RSA. A comprehensive tutorial regarding the application of GSA to a process model is presented.     

Sleepiness among night-time drivers: relationship to blood alcohol concentration and other factors

OBJECTIVE: The study aimed to explore the distribution and correlates of subjective sleepiness among the general night-time driving population. METHODS: The survey took place in three British Columbia communities in June 2003 between 21:00 hours to 03:00 hours. Sites and vehicles were selected randomly. Surveyors obtained information on several demographic and situational variables including self-assessed degree of sleepiness and self-reported hours asleep and awake, as well as an objective measure of blood alcohol concentration obtained from a hand-held breath-testing device. RESULTS: The total compliance rate among intercepted drivers was 85%. Among the 2335 drivers responding to the questionnaire, 68.4% indicated that they were wide awake, 27.6% were somewhat sleepy, and 4.1% were very sleepy. Logistic regression quantified the independent contributions of the various factors to subjective sleepiness. Male drivers with positive blood alcohol concentrations under 50 mg% were more likely to report feeling sleepy than those with either higher or with zero blood alcohol concentration. Greater relative risk of sleepiness was also associated with being female, being under age 55, and advanced hour of night. Driving with passengers of the same gender was associated with lower reported sleepiness. CONCLUSIONS: A substantial proportion of night-time drivers are driving while sleepy, especially at late night and early morning hours. The combination of alcohol and sleepiness compounds impairment in experimental studies and deserves greater attention in crash risk studies and as a topic for public education and awareness.     

Control chart for monitoring occupational asthma

INTRODUCTION: Work-related asthma has become the most prevalent occupational respiratory disease in the developed world. Occupational asthma is thought to affect 5%-10% of people worldwide. The first step in the diagnosis of occupational asthma is to establish work-relatedness. Although considerable research has been conducted in the area of occupational asthma, no simple, effective, and statistically sound method has been developed that can be used as an initial step to effectively identify the workers at risk for occupational asthma. The purpose of this research was to investigate whether Shewhart control chart method can be used as an effective method to detect occupational asthma. METHOD: Forty-five workers who completed the study and provided usable peak expiratory flow (a lung function marker) recordings while at work and away from work were included in this study. Control charts were developed using Shewhart's Method. The lower control limit of at work control chart (LCL(W)) was compared to each subject's Personal Best (PB) value. RESULTS: Reviewing the results of this comparison showed LCL(W)<60% PB to have a sensitivity of 85.71%, specificity of 87.50%, and an error rate of 13.33%. When the subjects suspected for false positive and false negative diagnoses were identified, the test produced a sensitivity of 95.24%, a specificity of 95.83% and an error rate of 4.44%. CONCLUSIONS: Our results were as good as, and in some cases better than, published clinical guidelines. IMPACT ON INDUSTRY: Our research showed that the control chart method is an effective, simple, and inexpensive tool for early intervention in workers suspected for occupational asthma.     

Mode choice model for vulnerable motorcyclists in Malaysia

OBJECTIVE: In developing countries, motorcycle use has grown in popularity in the past decades. Commensurate with this growth is the increase in death and casualties among motorcyclists in these countries. One of the strategic programs to minimize this problem is to reduce motorcyclists exposure by shifting them into safer modes of transport. This study aims to explore the differences in the characteristics of bus and motorcycle users. It identifies the factors contributing to their choice of transport mode and estimates the probability that motorcyclists might change their travel mode to a safer alternative; namely, bus travel. METHODS: In this article, a survey of 535 motorcycle and bus users was conducted in seven districts of Selangor state, Malaysia. A binary logit model was developed for the two alternative modes, bus and motorcycle. RESULTS: It was found that travel time, travel cost, gender, age, and income level are significant in influencing motorcyclists' mode choice behavior. The probability of motorcycle riders shifting to public transport was also examined based on a scenario of a reduction in bus travel time and travel cost. CONCLUSIONS: Reduction of total travel time for the bus mode emerges as the most important element in a program aimed at attracting motorcyclists towards public transport and away from the motorcycle mode.     

Characteristics and risks of drivers with low annual distance driven

OBJECTIVES: It has been noted by several authors that risk (defined only in terms of total expected numbers of crash involvements per total distance driven) paints a misleading picture of crash liability, particularly for the young and the old, as their high risk is associated with risky driving patterns typical of people who drive low annual kms. This article sets out to analyze these driving patterns of low-km drivers and to evaluate the risk of these patterns. As licensing programs tend to focus on young and old drivers, who tend to drive lower annual distances, income and employment data are also analyzed for low-km drivers. This is to provide a better picture for policy makers of the sort of people and the sorts of transportation requirements that their policies may affect. METHODS: Crash data and travel data were disaggregated by driver characteristics and by driving conditions (road type, day and night, weekend and weekday) and combined to form estimates of risk for typical driving patterns of driver groups. Characteristics of driving patterns and of the drivers themselves were derived for groups defined by age and by the amount of annual driving undertaken. RESULTS: Older drivers who drive less tend to have higher risk per km mainly due to their predominantly urban trips. Nevertheless, because older drivers on average manage to reduce their risk per distance driven by choosing driving patterns that are safer than the driving patterns of other age groups, the risk of older drivers as a group is not overestimated. CONCLUSION: Despite being quite different from one another, the low- and high-km driving patterns of younger drivers were found to impose identical risks.     

Young driver risk in relation to parents' retrospective driving record

PROBLEM: Parents are an important potential influence on the driving safety of their children. This study examined the relationship of parental driving record on male and female offspring's at-fault collision risk. METHOD: Drivers aged 16-21 on the date of full licensure were selected from driver records and a matching process was used to identify putative parents in two-parent households. Poisson regression models were developed to predict at-fault collisions of male and female youth in the three years following full licensure from parents' at-fault collisions, speeding offenses, and other moving offenses in the four years prior to children's licensure. One set of models examined the relative risk associated with increasing numbers of maternal and paternal at-fault collisions and offenses. Other models examined the joint versus separate maternal and parental contributions. RESULTS: Controlling for region of residence, both mothers' and fathers' at-fault collisions were associated with an increased risk in both male and female youth at-fault collisions. Mothers' and fathers' speeding offenses were also associated with increased relative risk of at-fault collisions for both sons and daughters, while fathers' other moving offenses increased collision risk for sons but not daughters. DISCUSSION: Further research is required to identify how parental driving risk is transmitted to children. IMPACT ON INDUSTRY: (a) Parents of young children should be informed of their role in influencing their children's future driving risk; (b) The results identify risk factors that could be of interest to licensing authorities and the insurance industry.