THE EFFECT OF DATA INDEPENDENCE IN MODEL CALIBRATION AND MODEL TESTING1

ABSTRACT: With the increased use of models in hydrologic design, there is an immediate need for a comprehensive comparison of hydrologic models, especially those intended for use at ungaged locations (i.e., where measured data are either not available or inadequate for model calibration). But some past comparisons of hydrologic models have used the same data base for both calibration and testing of the different models or implied that the results of model calibration are indicative of the accuracy at ungaged locations. This practice was examined using both the regression equation approach to peak discharge estimation and a unit hydrograph model that was intended for use in urban areas. The results suggested that the lack of data independence in the calibration and testing of regression equations may lead to both biased results and misleading statements about prediction accuracy. Additionally, although split-sample testing is recognized as desirable, the split-samples should be selected using a systematic-random sampling scheme, rather than random sampling, because random sampling with small samples may lead to a testing sample that is not representative of the population. A systematic-random sampling technique should lead to more valid conclusions about model reliability. For models like a unit hydrograph model, which are more complex and for which calibration is a more involved process, data independence is not as critical because the data fitting error variation is not as dominant as the error variation due to the calibration process and the inability of the model structure to conform with data variability.     

RIDGE REGRESSION TECHNIQUES APPLIED TO LANDSAT INVESTIGATION OF WATER QUALITY IN LAKE OKEECHOBEE1

ABSTRACT: Landsat radiance values were processed at two different (single and double) levels of accuracy to estimate chlorophyll a, turbidity, and suspended sediment in Lake Okeechobee, Florida. Both ordinary least square and ridge regression analyses were used to establish a relationship between water quality parameters and Landsat radiance. Radiance measurements made at greater precision (double level) gave a better solution in this application. The ridge regression analysis for double level not only can reduce the total mean square error about 13–20 percent and confidence interval about 6–28 percent as compared to ordinary least square analysis, but it can also change the interpretation of analysis results.     

CAPITAL COST OF RURAL WATER DISTRIBUTION SYSTEMS1

ABSTRACT: The cost of water service to rural residents is very high compared to urban areas. This is true even after subsidization by Farmers Home Administration (FmHA) loans and grants. Capital cost data on 44 projects financed by the Ohio office of the FmHA during the period August 1968 to January 1977 are used to derive cost equations for 26 components of rural water distribution systems. These components represent 92 percent of the capital cost of the pipeline distribution systems studied. The data can be used to economically design rural water supply systems from a capital cost viewpoint. More data are needed on operation and maintenance costs as well as central and cluster well costs before totally economic system designs can be undertaken.     

A NONLINEAR MULTILEVEL MODEL FOR REGIONAL WATER RESOURCES PLANNING1

ABSTRACT: A nonlinear multilevel transportation model is developed to study large-scale allocations in a water resources system. The model uses a modified transportation matrix formulated with nonlinear cost functions as the basic subregional model and the goal coordination method for multilevel decomposition and optimization of the overall regional system. The model is applied to projected water requirements for Salt Lake County in 1985. Sources of water supply - surface water, ground water, import water, and reuse of reclaimed wastewater on a restricted basis - are available to satisfy water requirements for municipal, industrial, and agricultural sectors in four subregions. The conjugate gradient projection method is used to optimize the first level subregional models having cost functions of the form of C = aX^b, and the second level problem is solved using the conjugate gradient method.     

COMPARING METHODS OF HYDROLOGIC REGIONALIZATION1

ABSTRACT: Data splitting is used to compare methods of determining “homogeneous” hydrologic regions. The methods compared use cluster analysis based on similarity of hydrologic characteristics or similarity of characteristics of a stream's drainage basin. Data for 221 stations in Arizona are used to show that the methods, which are a modification of DeCoursey's scheme for defining regions, improve the fit of estimation data to the model, but that is is necessary to have an independent measure of predictive accuracy, such as that provided by data splitting, to demonstrate improved predictive accuracy. The methods used the complete linkage algorithm for cluster analysis and computed weighted average estimates of hydrologic characteristics at ungaged sites.     

Risk factors for driving into flooded roads

Motor vehicle-related deaths account for more than half of all flood fatalities in the United States, but to date, very little is known about the risk factors associated with why people drive into flooded roads. Using data from survey questionnaires administered in Denver, CO, and Austin, TX, this paper suggests that people who do not take warnings seriously are more likely to drive through flooded roads, as are people aged 18–35, and those that do not know that motor vehicles are involved in more than half of all flood fatalities. In Denver, people who have not experienced a flood previously and those who do not know they live in flood-prone areas are also more likely to drive into flooded roads.     

Estimated Injury Risk for Specific Injuries and Body Regions in Frontal Motor Vehicle Crashes

Objective: Injury risk curves estimate motor vehicle crash (MVC) occupant injury risk from vehicle, crash, and/or occupant factors. Many vehicles are equipped with event data recorders (EDRs) that collect data including the crash speed and restraint status during a MVC. This study's goal was to use regulation-required data elements for EDRs to compute occupant injury risk for (1) specific injuries and (2) specific body regions in frontal MVCs from weighted NASS-CDS data.

Methods: Logistic regression analysis of NASS-CDS single-impact frontal MVCs involving front seat occupants with frontal airbag deployment was used to produce 23 risk curves for specific injuries and 17 risk curves for Abbreviated Injury Scale (AIS) 2+ to 5+ body region injuries. Risk curves were produced for the following body regions: head and thorax (AIS 2+, 3+, 4+, 5+), face (AIS 2+), abdomen, spine, upper extremity, and lower extremity (AIS 2+, 3+). Injury risk with 95% confidence intervals was estimated for 15–105 km/h longitudinal delta-Vs and belt status was adjusted for as a covariate.

Results: Overall, belted occupants had lower estimated risks compared to unbelted occupants and the risk of injury increased as longitudinal delta-V increased. Belt status was a significant predictor for 13 specific injuries and all body region injuries with the exception of AIS 2+ and 3+ spine injuries. Specific injuries and body region injuries that occurred more frequently in NASS-CDS also tended to carry higher risks when evaluated at a 56 km/h longitudinal delta-V. In the belted population, injury risks that ranked in the top 33% included 4 upper extremity fractures (ulna, radius, clavicle, carpus/metacarpus), 2 lower extremity fractures (fibula, metatarsal/tarsal), and a knee sprain (2.4–4.6% risk). Unbelted injury risks ranked in the top 33% included 4 lower extremity fractures (femur, fibula, metatarsal/tarsal, patella), 2 head injuries with less than one hour or unspecified prior unconsciousness, and a lung contusion (4.6–9.9% risk). The 6 body region curves with the highest risks were for AIS 2+ lower extremity, upper extremity, thorax, and head injury and AIS 3+ lower extremity and thorax injury (15.9–43.8% risk).

Conclusions: These injury risk curves can be implemented into advanced automatic crash notification (AACN) algorithms that utilize vehicle EDR measurements to predict occupant injury immediately following a MVC. Through integration with AACN, these injury risk curves can provide emergency medical services (EMS) and other patient care providers with information on suspected occupant injuries to improve injury detection and patient triage.     

On-Scene Injury Severity Prediction (OSISP) Algorithm for Truck Occupants

Objective: The aim of this study is to develop an on-scene injury severity prediction (OSISP) algorithm for truck occupants using only accident characteristics that are feasible to assess at the scene of the accident. The purpose of developing this algorithm is to use it as a basis for a field triage tool used in traffic accidents involving trucks. In addition, the model can be valuable for recognizing important factors for improving triage protocols used in Sweden and possibly in other countries with similar traffic environments and prehospital procedures.

Methods: The scope is adult truck occupants involved in traffic accidents on Swedish public roads registered in the Swedish Traffic Accident Data Acquisition (STRADA) database for calendar years 2003 to 2013. STRADA contains information reported by the police and medical data on injured road users treated at emergency hospitals. Using data from STRADA, 2 OSISP multivariate logistic regression models for deriving the probability of severe injury (defined here as having an Injury Severity Score [ISS] > 15) were implemented for light and heavy trucks; that is, trucks with weight up to 3,500 kg and ??16,500 kg, respectively. A 10-fold cross-validation procedure was used to estimate the performance of the OSISP algorithm in terms of the area under the receiver operating characteristic curve (AUC).

Results: The rate of belt use was low, especially for heavy truck occupants. The OSISP models developed for light and heavy trucks achieved cross-validation AUC of 0.81 and 0.74, respectively. The AUC values obtained when the models were evaluated on all data without cross-validation were 0.87 for both light and heavy trucks. The difference in the AUC values with and without use of cross-validation indicates overfitting of the model, which may be a consequence of relatively small data sets. Belt use stands out as the most valuable predictor in both types of trucks; accident type and age are important predictors for light trucks.

Conclusions: The OSISP models achieve good discriminating capability for light truck occupants and a reasonable performance for heavy truck occupants. The prediction accuracy may be increased by acquiring more data. Belt use was the strongest predictor of severe injury for both light and heavy truck occupants. There is a need for behavior-based safety programs and/or other means to encourage truck occupants to always wear a seat belt.     

回采工作面瓦斯涌出量遗传投影寻踪回归预测

为快速、准确预测回采工作面瓦斯涌出量,基于投影降维思想,建立一种遗传算法(GA)投影寻踪回归预测方法。选取煤层瓦斯原始含量、埋藏深度、煤层厚度、煤层倾角、工作面长度、推进速度、采出率、临近层瓦斯含量、临近层厚度、临近层层间距、岩层岩性、开采深度作为评价因子,对某矿15个学习样本进行训练,建立GA投影寻踪回归预测模型。利用该矿3个实测样本对模型进行检验,并与主成分分析和BP神经网络方法结果进行对比。研究表明:利用GA投影寻踪回归预测回采工作面瓦斯涌出量,平均误差为3.43%,最大误差为5.7%,精度优于其他2种方法。