COMPARISON OF PROCESS‐BASED AND ARTIFICIAL NEURAL NETWORK APPROACHES FOR STREAMFLOW MODELING IN AN AGRICULTURAL WATERSHED1

ABSTRACT: The performance of the Soil and Water Assessment Tool (SWAT) and artificial neural network (ANN) models in simulating hydrologic response was assessed in an agricultural watershed in southeastern Pennsylvania. All of the performance evaluation measures including Nash‐Sutcliffe coefficient of efficiency (E) and coefficient of determination (R²) suggest that the ANN monthly predictions were closer to the observed flows than the monthly predictions from the SWAT model. More specifically, monthly streamflow E and R² were 0.54 and 0.57, respectively, for the SWAT model calibration period, and 0.71 and 0.75, respectively, for the ANN model training period. For the validation period, these values were ?0.17 and 0.34 for the SWAT and 0.43 and 0.45 for the ANN model. SWAT model performance was affected by snowmelt events during winter months and by the model's inability to adequately simulate base flows. Even though this and other studies using ANN models suggest that these models provide a viable alternative approach for hydrologic and water quality modeling, ANN models in their current form are not spatially distributed watershed modeling systems. However, considering the promising performance of the simple ANN model, this study suggests that the ANN approach warrants further development to explicitly address the spatial distribution of hydrologic/water quality processes within watersheds.     

EFFECTS OF STATSGO AND SSURGO AS INPUTS ON SWAT MODEL'S SNOWMELT SIMULATION1

ABSTRACT: Soil data comprise a basic input of SWAT (Soil and Water Assessment Tool) for a watershed application. For watersheds where site specific soil data are unavailable, the two U.S. Department of Agriculture (USDA) soil databases, the State Soil Geographic (STATSGO) and Soil Survey Geographic (SSURGO) databases, may be the best alternatives. Although it has been noted that SWAT models using the STATSGO and SSURGO data may give different simulation results for water, sediment, and agricultural chemical yields, information is scarce on the effects of using these two databases in predicting streamflows that are predominantly generated from melting snow in spring. The objective of this study was to assess the effects of using STATSGO versus SSURGO as an input for the SWAT model's simulation of the streamflows in the upper 45 percent of the Elm River watershed in eastern North Dakota. Designating the model as SWAT‐STATSGO when the STATSGO data were used and SWAT‐SSURGO when the SSURGO data were used, SWAT‐STATSGO and SWAT‐SSURGO were separately calibrated and validated using the observed daily streamflows. The results indicated that SWAT‐SSURGO provided an overall better prediction of the discharges than SWAT‐STATSGO, although both did a good and comparable job of predicting the high streamflows. However, SWAT‐STATSGO predicted the low streamflows more accurately and had a slightly better performance during the validation period. In addition, the discrepancies between the discharges predicted by these two SWAT models tended to be larger at upstream locations than at those farther downstream within the study area.     

Estimation of Cumulative Frequency Distribution for Carbon Monoxide Concentration from Wind-Speed Data, in Buenos Aires (Argentina)

In this article we apply and test a methodology to estimate cumulative frequency distribution for air pollutant concentration from wind-speed data. We use the inverse relationship after Simpson et al. (Atmospheric Environment, 19, 75–82, 1985) between the opposing percentile values in the statistical distributions for air pollutant concentrations and wind-speed data. This relationship is valid, irrespective of the statistical distributions of both variables, if an inverse relationship between them is also applicable. The available data are five years of 8-h average carbon monoxide concentration and 8-h mean wind-speed, observed in Buenos Aires (Argentina). The performance of the obtained empirical expressions in estimating cumulative frequency distributions for 8-h CO is statistically evaluated. The results show that it is possible to obtain an acceptable cumulative frequency distribution for 8-h CO concentration at the site if the cumulative frequency distribution for wind-speed is known. Q–Q plots show a good agreement between estimated and observed values. From our data, the mean relative error of the estimations was found to be as much as 8.0%.     

Potential and Shortcomings of Numerical Weather Prediction Models in Providing Meteorological Data for Urban Air Pollution Forecasting

The last decade progress in numericalweather prediction (NWP) modelling and studies of urbanatmospheric processes for providing meteorological data forurban air pollution forecasting is analysed on examples ofseveral European meteorological centres. Modern nested NWP models are utilising land-use databasesdown to 1 km resolution or finer, and are approaching thenecessary horizontal and vertical resolution suitable forcity scale. The recent scientific developments in the fieldof urban atmospheric physics and the growing availabilityof high-resolution urban surface characteristics datapromise further improvements of the capability of NWPmodels for this aim. A strategy to improve NWP data forthe urban air pollution forecasting is suggested.     

2002年7-10月国内安全事故数据

统计了2002年7-10月国内发生的各种安全事故1231起,包括矿业事故、交通事故、爆炸事故、火灾、毒物泄露和中毒及其他事故.统计表明,在这些事故中,矿业事故最多,占63.1%,平均每天6.5起事故,其次是交通事故(18.6%)、其他事故(6.0%)、爆炸事故(4.8%)、毒物泄露和中毒(4.1%)、火灾(3.3%).1231起事故共死亡2 773人,伤3 206人,死亡人数的百分比分别为矿业事故48.9%、交通事故34.5%、其他事故8.3%、爆炸事故3.8%、泄露中毒2.8%、火灾1.7%;受伤人数的百分比分别为交通事故51.7%、泄露中毒19.1%、矿业事故12.2%、其他事故8.1%、爆炸事故7.0%、火灾1.9%.     

PRECIPITATION CHANGES FROM 1956 TO 1996 ON THE WALNUT GULCH EXPERIMENTAL WATERSHED1

ABSTRACT: The climate of Southern Arizona is dominated by summer precipitation, which accounts for over 60 percent of the annual total. Summer and non‐summer precipitation data from the USDA‐ARS Walnut Gulch Experimental Watershed are analyzed to identify trends in precipitation characteristics from 1956 to 1996. During this period, annual precipitation increased. The annual precipitation increase can be attributed to an increase in precipitation during non‐summer months, and is paralleled by an increase in the proportion of annual precipitation contributed during non‐summer months. This finding is consistent with previously reported increases in non‐summer precipitation in the southwestern United States. Detailed event data were analyzed to provide insight into the characteristics of precipitation events during this time period. Precipitation event data were characterized based on the number of events, event precipitation amount, 30‐minute event intensity, and event duration. The trend in non‐summer precipitation appears to be a result of increased event frequency since the number of events increased during nonsummer months, although the average amount per event, average event intensity, and average event duration did not. During the summer “monsoon” season, the frequency of recorded precipitation events increased but the average precipitation amount per event decreased. Knowledge of precipitation trends and the characteristics of events that make up a precipitation time series is a critical first step in understanding and managing water resources in semiarid ecosystems.     

Modeling the relationships between land use and land cover on private lands in the Upper Midwest, USA

This paper presents an approach to modeling land-cover change as a function of land-use change. We argue that, in order to model the link between socio-economic change and changes in forest cover in a region that is experiencing residential and recreational development and agricultural abandonment, land-use and land-cover change need to be represented as separate processes. Forest-cover change is represented here using two transition probabilities that were calculated from Landsat imagery and that, taken together, describe a Markov transition matrix between forest and non-forest over a 10-year period. Using a three-date land-use data set, compiled and interpreted from digitized parcel boundaries, and scanned aerial photography for 136 sites (c. 2500 ha) sampled from the Upper Midwest, USA, we test functional relationships between forest-cover transition probabilities, standardized to represent changes over a decade, and land-use conditions and changes within sample sites. Regression models indicated that about 60% of the variation in the average forest-cover transition probabilities (i.e. from forest to non-forest and vice versa) can be predicted using three variables: amount of agricultural land use in a site; amount of developed land use; and the amount of area increasing in development. In further analysis, time lags were evaluated, showing that agricultural abandonment had a relatively strong time-lag effect but development did not. We demonstrate an approach to using forest-cover transition probabilities to develop spatially-constrained simulations of forest-cover change. Because the simulations are based on transition probabilities that are indexed to a particular time and place, the simulations are improved over previous applications of Markov transition models. This modeling approach can be used to predict forest-cover changes as a result of socio-economic change, by linking to models that predict land-use change on the basis of exogenous human-induced drivers.     

COUPLING OF NONPARAMETRIC FREQUENCY AND L-MOMENT ANALYSES FOR MIXED DISTRIBUTION IDENTIFICATION 1

ABSTRACT: Both L-moment and nonparametric frequency analyses were performed on a series of annual maximum floods from New Brunswick, Canada. The L-moment analysis concluded that the data were generated from a unimodal Generalized Extreme Value (GEV) distribution. However, the nonparametric frequency analysis indicated that a majority of stations followed nonunimodal mixed distributions since peak flows occur during different seasons and are the result of different generating mechanisms. The coupling of L-moment and nonparametric analyses facilitates mixed distribution identification. Thus, the nonparametric method helps in identifying underlying probability distribution, especially when samples arise from mixed distributions.     

A MATHEMATICAL MODEL OF SALMONID SPAWNING HABITAT1

ABSTRACT: A simulation model [Salmonid Spawning Analysis Model (SSAM)] was developed as a management tool to evaluate the relative impacts of stream sediment load and water temperature on salmonid egg survival. The model is useful for estimating acceptable sediment loads to spawning habitat that may result from upland development, such as logging and agriculture. Software in common use in the USA were adapted for use in gravel bedded rivers and linked to simulate water temperature (the USFWS Instream Water Temperature, SNTEMP model) and water and sediment routing (the USAE Scour and Deposition in Rivers and Reservoirs, HEC-6 model, version 3.2). These models drive the redd (spawning nest) model (the USDA-ABS Sediment Intrusion Dissolved Oxygen SIDO model) which simulates sediment intrusion and dissolved oxygen concentration in the redd environment. The SSAM model predictions of dissolved oxygen and water temperature compared favorably with field data from artificial redds containing hatchery chinook salmon eggs.     

ECOLOGICAL MODELS AND ENVIRONMENTAL STUDIES1

The recent world-wide trend towards centralization of all environmental management functions into one regulatory agency has illustrated the necessity for resources management agencies to adopt a total systems viewpoint. Environmental systems are typically complex and multi-dimensional in nature. Mathematical models for the management of air, water and land resources have found wide acceptance among planners and decision-makers. Ecological models of life processes have not reached the same state of development or acceptance. A general review of ecological systems theory and examples of the types of ecological models that have been developed to date arc presented in this paper. With this material as a background and given the vast literature on engineering and economic models, a conceptual framework for an approach to environmental studies and the analysis of polluted environmental systems is presented.