Intervention analysis in environmental engineering

Intervention analysis techniques are described for identifying and statistically modelling trends which may be present in water quality time series. At the exploratory data analysis stage, simple graphical and modelling methods can be employed for visually detecting and examining trends in a time series caused by one or more external interventions. For instance, a plot of a robust locally weighted regression smooth through a graph of the observations over time may reveal trends and other interesting statistical properties contained in the time series. In addition, statistical tests, such as different versions of the nonparametric Mann-Kendall test, can be used to detect the presence of trends caused by unknown or known external interventions. To characterize rigorously and estimate trends which may be known in advance or else detected using exploratory data analysis studies, different parametric methods can be utilized at the confirmatory data analysis stage. Specifically, the time series modelling approach to intervention analysis can be employed to estimate the magnitudes of the changes in the mean level of the series due to the interventions. Particular types of regression models can also be used for estimating trends, especially when there are many missing observations. To demonstrate how intervention analysis methods can be effectively used in environmental impact assessment, representative applications to water quality time series are presented.Invited Paper for Presentation at The Workshop on Statistical Methods for the Assessment of Point Source Pollution, The Canada Centre for Inland Waters, Burlington, Ontario, Canada, L7R 4A6, September 12–14, 1988.     

Statistical models for evaluating suspected artefacts in long-term environmental monitoring data

Long-term water quality monitoring is of high value for environmental management as well as for research. Artificial level shifts in time series due to method improvements, flaws in laboratory practices or changes in laboratory are a common limitation for analysis, which, however, are often ignored. Statistical estimation of such artefacts is complicated by the simultaneous existence of trends, seasonal variation and effects of other influencing factors, such as weather conditions. Here, we investigate the performance of generalised additive mixed models (GAMM) to simultaneously identify one or more artefacts associated with artificial level shifts, longitudinal effects related to temporal trends and seasonal variation, as well as to model the serial correlation structure of the data. In the same model, it is possible to estimate separate residual variances for different periods so as to identify if artefacts not only influence the mean level but also the dispersion of a series. Even with an appropriate statistical methodology, it is difficult to quantify artificial level shifts and make appropriate adjustments to the time series. The underlying temporal structure of the series is especially important. As long as there is no prominent underlying trend in the series, the shift estimates are rather stable and show less variation. If an artificial shift occurs during a slower downward or upward tendency, it is difficult to separate these two effects and shift estimates can be both biased and have large variation. In the case of a change in method or laboratory, we show that conducting the analyses with both methods in parallel strongly improves estimates of artefact effects on the time series, even if certain problems remain. Due to the difficulties of estimating artificial level shifts, posterior adjustment is problematic and can lead to time series that no longer can be used for trend analysis or other analysis based on the longitudinal structure of the series. Before carrying out a change in analytic method or laboratory, it should be considered if this is absolutely necessary. If changes cannot be avoided, the analysis of the two methods considered, or the two laboratories contracted, should be run in parallel for a considerable period of time so as to enable a good assessment of changes introduced to the data series.     

A novel high-frequency groundwater quality monitoring system

High-frequency, long-term monitoring of water quality has revolutionized the study of surface waters in recent years. However, application of these techniques to groundwater has been limited by the ability to remotely pump and analyze groundwater. This paper describes a novel autonomous groundwater quality monitoring system which samples multiple wells to evaluate temporal changes and identify trends in groundwater chemistry. The system, deployed near Fresno, California, USA, collects and transmits high-frequency data, including water temperature, specific conductance, pH, dissolved oxygen, and nitrate, from supply and monitoring wells, in real-time. The system consists of a water quality sonde and optical nitrate sensor, manifold, submersible three-phase pump, variable frequency drive, data collection platform, solar panels, and rechargeable battery bank. The manifold directs water from three wells to a single set of sensors, thereby reducing setup and operation costs associated with multi-sensor networks. Sampling multiple wells at high frequency for several years provided a means of monitoring the vertical distribution and transport of solutes in the aquifer. Initial results show short period variability of nitrate, specific conductivity, and dissolved oxygen in the shallow aquifer, while the deeper portion of the aquifer remains unchanged—observations that may be missed with traditional discrete sampling approaches. In this aquifer system, nitrate and specific conductance are increasing in the shallow aquifer, while invariant changes in deep groundwater chemistry likely reflect relatively slow groundwater flow. In contrast, systems with high groundwater velocity, such as karst aquifers, have been shown to exhibit higher-frequency groundwater chemistry changes. The stability of the deeper aquifer over the monitoring period was leveraged to develop estimates of measurement system uncertainty, which were typically lower than the manufacturer’s stated specifications, enabling the identification of subtle variability in water chemistry that may have otherwise been missed.     

Trends in Nitrogen Transport in Swedish Rivers

Concern about nitrogen loads in marine environments has drawn attention to the existence and possible causes of long-term trends in nitrogen transport in rivers. The present study was based on data from the Swedish environmental monitoring programme for surface water quality; the continuity of these data is internationally unique. A recently developed semiparametric method was employed to study the development of relationships between runoff and river transport of nitrogen since 1971; the observed relationships were then used to produce time series of flow-normalised transports for 66 sites in 39 river basins. Subsequent statistical analyses of flow-normalised data revealed only few significant downward trends (p 0.05) during the time period 1971–1994, and the most pronounced of these downward trends were caused by reduced point emissions of nitrogen. The number of significant upward trends was substantially larger (15 for total-N and 18 for NO₃-N). Closer examination of obtained results revealed the following: (i) the most pronounced upward trends were present downstream of lakes, and (ii) observed increases in nitrogen transport coincided in time and space with reduced point emissions of phosphorus or organic matter. This indicated that changes in the retention of nitrogen in lakes were responsible for the upward nitrogen trends. The hypothesis that nitrogen saturation of forest soils has caused a general increase in the riverine export of nitrogen from forested catchments in Sweden was not confirmed. Neither did the results indicate that improved agricultural practices have reduced the export of nitrogen from agricultural catchments.     

Robust graphical methods for diagnosing trend in irregularly spaced water quality time series

Graphical methods can play an important role in the reliable assessment of trends in typically ill behaved river quality data series both as diagnostic tools and as visual corroborative evidence when assumptions required for formal statistical tests are not met. Robust, graphically-oriented trend diagnosis procedures are presented for data series characterized by nonnormal populations, uneven time spacing, nonmonotonic trend and other factors which can create serious problems for standard parametric time series methods. Cleveland's robust locally weighted regression (RLWR) developed for investigating nonlinearity in x-y scatterplots is adapted as a robust/resistant smoothing filter for the analysis of irregular time series comprising quantitative observations. Low powered RLWR trend lines reveal temporally local phenomena, e.g. abrupt jumps (often associated with point source impacts) and periodicities, while higher powered RLWR yields smooth lines characterizing medium and longer term trends. Simple variants of Tukey smoothing concepts are developed for series with censored observations. Applications to Ontario river quality series reveal that graphical evidence is frequently sufficient to obviate the need for formal trend testing. The methods are generally applicable to most time series.     

Principal response curves technique for the analysis of multivariate biomonitoring time series

Although chemical and biological monitoring is often used to evaluate the quality of surface waters for regulatory purposes and/or to evaluate environmental status and trends, the resulting biological and chemical data sets are large and difficult to evaluate. Multivariate techniques have long been used to analyse complex data sets. This paper discusses the methods currently in use and introduces the principal response curves method, which overcomes the problem of cluttered graphical results representation that is a great drawback of most conventional methods. To illustrate this, two example data sets are analysed using two ordination techniques, principal component analysis and principal response curves. Whereas PCA results in a difficult-to-interpret diagram, principal response curves related methods are able to show changes in community composition in a diagram that is easy to read. The principal response curves method is used to show trends over time with an internal reference (overall mean or reference year) or external reference (e.g. preferred water quality or reference site). Advantages and disadvantages of both methods are discussed and illustrated.     

Evaluation of groundwater quality in and around Peenya industrial area of Bangalore, South India using GIS techniques

Groundwater resource forms a significant component of the urban water supply. Declining groundwater levels in Bangalore Urban District is generally due to continuous overexploitation during the last two decades or more. There is a tremendous increase in demand in the city for good quality groundwater resource. The present study monitors the groundwater quality using geographic information system (GIS) techniques for a part of Bangalore metropolis. Thematic maps for the study area are prepared by visual interpretation of SOI toposheets on 1:50,000 scale using MapInfo software. Physicochemical analysis data of the groundwater samples collected at predetermined locations form the attribute database for the study, based on which spatial distribution maps of major water quality parameters are prepared using MapInfo GIS software. Water quality index was then calculated by considering the following water quality parameters--pH, total dissolved solids, total hardness, calcium hardness, magnesium hardness, alkalinity, chloride, nitrate and sulphate to find the suitability of water for drinking purpose. The water quality index for these samples ranged from 49 to 502. The high value of water quality index reveals that most of the study area is highly contaminated due to excessive concentration of one or more water quality parameters and that the groundwater needs pretreatment before consumption.     

Estimating contaminant attenuation half-lives in alluvial groundwater systems

One aspect of describing contamination in an alluvial aquifer is estimating changes in concentrations over time. A variety of statistical methods are available for assessing trends in contaminant concentrations. We present a method that extends trend analysis to include estimating the coefficients for the exponential decay equation and calculating contaminant attenuation half-lives. The conceptual model for this approach assumes that the rate of decline is proportional to the contaminant concentration in an aquifer. Consequently, the amount of time to remove a unit quantity of the contaminant inventory from an aquifer lengthens as the concentration decreases. Support for this conceptual model is demonstrated empirically with log-transformed time series of contaminant data. Equations are provided for calculating system attenuation half-lives for non-radioactive contaminants. For radioactive contaminants, the system attenuation half-life is partitioned into the intrinsic radioactive decay and the concentration reduction caused by aquifer processes. Examples are presented that provide the details of this approach. In addition to gaining an understanding of aquifer characteristics and changes in constituent concentrations, this method can be used to assess compliance with regulatory standards and to estimate the time to compliance when natural attenuation is being considered as a remediation strategy. A special application of this method is also provided that estimates the half-life of the residence time for groundwater in the aquifer by estimating the half life for a conservative contaminant that is no longer being released into the aquifer. Finally, the ratio of the half-life for groundwater residence time to the attenuation half-life for a contaminant is discussed as a system-scale retardation factor which can be used in analytical and numerical modeling.     

Time Series Analysis to Monitor and Assess Water Resources: A Moving Average Approach

An understanding of the behavior of the groundwater body and its long-term trends are essential for making any management decision in a given watershed. Geostatistical methods can effectively be used to derive the long-term trends of the groundwater body. Here an attempt has been made to find out the long-term trends of the water table fluctuations of a river basin through a time series approach. The method was found to be useful for demarcating the zones of discharge and of recharge of an aquifer. The recharge of the aquifer is attributed to the return flow from applied irrigation. In the study area, farmers mainly depend on borewells for water and water is pumped from the deep aquifer indiscriminately. The recharge of the shallow aquifer implies excessive pumping of the deep aquifer. Necessary steps have to be taken immediately at appropriate levels to control the irrational pumping of deep aquifer groundwater, which is needed as a future water source. The study emphasizes the use of geostatistics for the better management of water resources and sustainable development of the area.     

美国Surfer 8.0软件在地下水环境质量评价中的应用

通过Surfer 8.0软件在某地地下水环境评价中的应用实例,重点阐述了其在浅层地下水污染物浓度空间分布图绘制、地下水质量分区图绘制、地下水污染分区图绘制、区域地下水污染趋势分析、地下水污染原因分析等方面的应用。     

Water quality indices as indicators of ecosystem change

The operational management of water quality requires a methodology that can provide precise information on cycles and trends in water quality in an objective and reproducible manner. Such information can be provided by the adoption of a water quality indexing system. The continuous scale afforded by a water quality index allows changes in river water quality to be highlighted. At the same time the sub-division of this scale into a series of water quality and water use categories provides an easy means of relating information to government and public.The development of four independent water quality indices (WQIs) is outlined. These have been applied to data for a number of UK river reaches. The results of these applications indicate the utility of these indices in the classification of water quality and the monitoring of ecosystem change.     

Application of geographically weighted regression model to analysis of spatiotemporal varying relationships between groundwater quantity and land use changes (case study: Khanmirza Plain,Iran)

Understanding the spatiotemporal relationships between land use/cover changes (LUCC) and groundwater resources is necessary for effective and efficient land use management. In this paper, geographically weighted regression (GWR) and ordinary least squares (OLS) models have been expanded to analyze varying spatial relationships between groundwater quantity changes and LUCC for three periods: 1987–2000, 2000–2010, and 1987–2010 in the Khanmirza Plain of southwestern Iran. For this purpose, TM images were used to generate LUCC (rainfed, irrigated, meadow, and bare lands). Groundwater quantity variables, including groundwater level changes (GLC) and groundwater withdrawal differences (GWD), were gathered from piezometric and agricultural wells data. The analysis of spatial autocorrelation (Moran’s I and local indicators of spatial association ) demonstrated that GWR has a better ability to model spatially varying data with very minimal clustering of residuals. The results R ² and corrected Akaike’s Information Criterion parameters revealed that the GWR has the lowest similarity in space and time in neighboring situations and it has the high ability to explain more variance in the LUCC as a function of the groundwater quantity changes. All results of the distribution of local R ² values from GWR confirm our assertion that there is a spatiotemporal relationship between types of land use and each of groundwater quantity variables within the region. According to the t test results from GWR, there are significant differences between the GLC and GWD and the land use types in different places of region in each of the three time series. The GWR results can help decision-makers to make appropriate decisions for future planning.     

Water quality time series for Big Melen stream (Turkey): its decomposition analysis and comparison to upstream

Big Melen stream is one of the major water resources providing 268 km³ year^???1 of drinking and municipal water for Istanbul. Monthly time series data between 1991 and 2004 for 25 chemical, biological, and physical water properties of Big Melen stream were separated into linear trend, seasonality, and error components using additive decomposition models. Water quality index (WQI) derived from 17 water quality variables were used to compare Aksu upstream and Big Melen downstream water quality. Twenty-six additive decomposition models of water quality time series data including WQI had R ² values ranging from 88% for log(water temperature) (P?≤?0.001) to 3% for log(total dissolved solids) (P?≤?0.026). Linear trend models revealed that total hardness, calcium concentration, and log(nitrite concentration) had the highest rate of increase over time. Tukey’s multiple comparison pointed to significant decreases in 17 water quality variables including WQI of Big Melen downstream relative to those of Aksu upstream (P?≤?0.001). Monitoring changes in water quality on the basis of watersheds through WQI and decomposition analysis of time series data paves the way for an adaptive management process of water resources that can be tailored in response to effectiveness and dynamics of management practices.     

Spatio-temporal analyses of cropland degradation in the irrigated lowlands of Uzbekistan using remote-sensing and logistic regression modeling

Advancing land degradation in the irrigated areas of Central Asia hinders sustainable development of this predominantly agricultural region. To support decisions on mitigating cropland degradation, this study combines linear trend analysis and spatial logistic regression modeling to expose a land degradation trend in the Khorezm region, Uzbekistan, and to analyze the causes. Time series of the 250-m MODIS NDVI, summed over the growing seasons of 2000–2010, were used to derive areas with an apparent negative vegetation trend; this was interpreted as an indicator of land degradation. About one third (161,000 ha) of the region’s area experienced negative trends of different magnitude. The vegetation decline was particularly evident on the low-fertility lands bordering on the natural sandy desert, suggesting that these areas should be prioritized in mitigation planning. The results of logistic modeling indicate that the spatial pattern of the observed trend is mainly associated with the level of the groundwater table (odds?=?330 %), land-use intensity (odds?=?103 %), low soil quality (odds?=?49 %), slope (odds?=?29 %), and salinity of the groundwater (odds?=?26 %). Areas, threatened by land degradation, were mapped by fitting the estimated model parameters to available data. The elaborated approach, combining remote-sensing and GIS, can form the basis for developing a common tool for monitoring land degradation trends in irrigated croplands of Central Asia.     

Assessment of the statistical significance of seasonal groundwater quality change in a karstic aquifer system near Izmir-Turkey

The main objective of this study was to statistically evaluate the significance of seasonal groundwater quality change and to provide an assessment on the spatial distribution of specific groundwater quality parameters. The studied area was the Mount Nif karstic aquifer system located in the southeast of the city of Izmir. Groundwater samples were collected at 57 sampling points in the rainy winter and dry summer seasons. Groundwater quality indicators of interest were electrical conductivity (EC), nitrate, chloride, sulfate, sodium, some heavy metals, and arsenic. Maps showing the spatial distributions and temporal changes of these parameters were created to further interpret spatial patterns and seasonal changes in groundwater quality. Furthermore, statistical tests were conducted to confirm whether the seasonal changes for each quality parameter were statistically significant. It was evident from the statistical tests that the seasonal changes in most groundwater quality parameters were statistically not significant. However, the increase in EC values and aluminum concentrations from winter to summer was found to be significant. Furthermore, a negative correlation between sampling elevation and groundwater quality was found. It was shown that with simple statistical testing, important conclusions can be drawn from limited monitoring data. It was concluded that less groundwater recharge in the dry period of the year does not always imply higher concentrations for all groundwater quality parameters because water circulation times, lithology, quality and extent of recharge, and land use patterns also play an important role on the alteration of groundwater quality.     

The Swedish System for Quality Assessment of Agricultural Soils

Soil quality assessment is based on the concept of soilfunctions. The performances of three soil functions, cropproduction, biological decomposition and matter exchange withthe atmosphere and groundwater, are used as quality criteria.Soil properties that can be used as indicators for the degreeof functional performance were identified. Each soil propertyselected was graded into five classes – from best (class 1) toworst (class 5). Grading was based biological on boundaryconditions as well as on statistical distribution. The systemoutlined can be used to interpret the quality state ofagricultural soils, provides for a relative comparison betweensoils, and may be helpful in an environmental monitoringprogram to assess trends in data.     

Spatiotemporal variation of groundwater quality using integrated multivariate statistical and geostatistical approaches in Amol–Babol Plain,Iran

In recent years, groundwater quality has become a global concern due to its effect on human life and natural ecosystems. To assess the groundwater quality in the Amol–Babol Plain, a total of 308 water samples were collected during wet and dry seasons in 2009. The samples were analysed for their physico-chemical and biological constituents. Multivariate statistical analysis and geostatistical techniques were applied to assess the spatial and temporal variabilities of groundwater quality and to identify the main factors and sources of contamination. Principal component analysis (PCA) revealed that seven factors explained around 75 % of the total variance, which highlighted salinity, hardness and biological pollution as the dominant factors affecting the groundwater quality in the Plain. Two-way analysis of variance (ANOVA) was conducted on the dataset to evaluate the spatio-temporal variation. The results showed that there were no significant temporal variations between the two seasons, which explained the similarity between six component factors in dry and wet seasons based on the PCA results. There are also significant spatial differences (p?>?0.05) of the parameters under study, including salinity, potassium, sulphate and dissolved oxygen in the plain. The least significant difference (LSD) test revealed that groundwater salinity in the eastern region is significantly different to the central and western side of the study area. Finally, multivariate analysis and geostatistical techniques were combined as an effective method for demonstrating the spatial structure of multivariate spatial data. It was concluded that multiple natural processes and anthropogenic activities were the main sources of groundwater salinization, hardness and microbiological contamination of the study area.     

Groundwater levels time series sensitivity to pluviometry and air temperature: a geostatistical approach to Sfax region,Tunisia

In this paper, the pattern of groundwater level fluctuations is investigated by statistical techniques for 24 monitoring wells located in an unconfined coastal aquifer in Sfax (Tunisia) for a time period from 1997 to 2006. Firstly, a geostatistical study is performed to characterize the temporal behaviors of data sets in terms of variograms and to make predictions about the value of the groundwater level at unsampled times. Secondly, multivariate statistical methods, i.e., principal component analysis (PCA) and cluster analysis (CA) of time series of groundwater levels are used to classify groundwater hydrographs regard to identical fluctuation pattern. Three groundwater groups (A, B, and C) were identified. In group “A,” water level decreases continuously throughout the study periods with rapid annual cyclic variation, whereas in group “B,” the water level contains much less high-frequency variation. The wells of group “C” represents a steady and gradual increase of groundwater levels caused by the aquifer artificial recharge. Furthermore, a cross-correlation analysis is used to investigate the aquifer response to local rainfall and temperature records. The result revealed that the temperature is more affecting the variation of the groundwater level of group A wells than the rainfall. However, the second and the third groups are less affected by rainfall or temperature.     

模糊-主成分分析综合评价法在地下水水质评价中的应用

主成分分析法可以在保留数据原始信息的基础上有效地降低数据维度,从而将影响地下水质的多因子简化为几个综合因子,但对地下水的综合水质状况无法直接表达;模糊综合评价法能直接给出水体综合水质情况结论,但评价因子多为人为选取,存在较强的主观性。以辽宁思山岭矿区地下水水质现状评价为例,利用主成分分析法选取影响各断面水质的关键因子,将其作为模糊综合评价的评价因子,建立了模糊?主成分分析综合评价法的地下水水质耦合评价模型。研究结果表明,该模型能很好地体现分析因子对评价结果带来的影响,使得评价结果更科学、合理。     

Spatiotemporal analysis of the PM10 concentration over the Taranto area

In this paper, an analysis of air quality data is provided for the municipal area of Taranto (southern Italy) characterized by high environmental risks as formally decreed by the Italian government in the 1990s with two administrative measures. This is due to the massive presence of industrial sites with elevated environmental impact activities along the NW boundary of the city conurbation. The aforementioned activities have effects on the environment and on public health, as a number of epidemiological researches concerning this area reconfirm. The present study is focused on particulate matter as measured by PM10 concentrations at 13 monitoring stations, equipped with analogous instruments based on the Beta absorption technology, either reporting hourly, two-hourly, or daily measurements. Daily estimates of the PM10 concentration surfaces are obtained in order to identify areas of higher concentration (hot spots), possibly related to specific anthropic activities. Preliminary analysis involved addressing several data problems: (1) due to the use of two different validation techniques, a calibration procedure was devised to allow for data comparability; (2) imputation techniques were considered to cope with the large number of missing data, due to both different working periods and occasional malfunctions of PM10 sensors; and (3) reliable weather covariates (wind speed and direction, pressure, temperature, etc.) were obtained and considered within the analysis. Spatiotemporal modelling was addressed by a Bayesian kriging-based model proposed by Le and Zidek (2006) characterized by the use of time varying covariates and a semiparametric covariance structure. Advantages and disadvantages of the model are highlighted and assessed in terms of fit and performance. Estimated daily PM10 concentration surfaces are suitable for the interpretation of time trends and for identifying concentration peaks within the urban area.