Spatial extrapolation models

Experimental aquatic ecosystems have become widely used tools in ecotoxicology because they allow for a greater degree of control, replication, and repeatability than is achievable in natural ecosystems. The test systems in use vary from small indoor microcosms to large and complex outdoor experimental ecosystems. However, natural freshwater systems may also vary considerably in size and ecological complexity. Before addressing the spatial extrapolation of results of model ecosystem experiments that were conducted on different localities, the possible influence of the size and ecological complexity of test systems on responses to chemical stress will be discussed. 

At the regional scale, other means of parameterization are required. A typical example is the evaluation of large-scale non-point source pollution with spatially distributed modeling approaches, which very often rely on the availability of the soil s physico-chemical properties at the scale of each grid of a constructed soil information system. Unfortunately, only limited hard data are available in most soil information systems. Grid scale modeling parameters need to be generated by interpolation, extrapolation, geo-statistics, pedo-transfer functions and the like. Pedotransfer functions will play an important role in this context. 

Applying MD to systems of biochemical interest, such as proteins or DNA in solution, one has to deal with several thousands of atoms. Models for systems with long spatial correlations, such as liquid crystals, micelles, or any system near a phase transition or critical point, also must involve a large number of atoms. Some of these systems, including synthetic polymers, obey certain scaling laws that allow the estimation of the behaviour of a large system by extrapolation. Unfortunately, proteins are very precise structures that evade such simplifications. So let us take 10,000 atoms as a reasonable size for a realistic complex system. 

As probabilistic exposure and risk assessment methods are developed and become more frequently used for environmental fate and effects assessment, OPP increasingly needs distributions of environmental fate values rather than single point estimates, and quantitation of error and uncertainty in measurements. Probabilistic models currently being developed by the OPP require distributions of environmental fate and effects parameters either by measurement, extrapolation or a combination of the two. The models predictions will allow regulators to base decisions on the likelihood and magnitude of exposure and effects for a range of conditions which vary both spatially and temporally, rather than in a specific environment under static conditions. This increased need for basic data on environmental fate may increase data collection and drive development of less costly and more precise analytical methods. 

The model was forced with agricultural application data of the insecticide DDT compiled by Semeena and Lammel (2003). Statistical data of DDT consumption reported by member of the UN states to Food and Agriculture Organisation (FAO) were combined with other published data (details in Semeena and Lammel (2003)). The emission inventory assumed 100 % of p,p -DDT. After scaling the DDT consumption with crop land distribution, the data were extrapolated to the model grid. The result was a data set with spatially and temporally varying applications (accumulated application and temporal evolution shown in Figure 3.1). No seasonal or diurnal variation of the applications is considered. 

FIGURE 16.16 Peak ozone isopleths calculated for downtown Los Angeles (DTLA) and Rubidoux, approximately 100 km east and downwind of DTLA under typical meteorological conditions. Spatially uniform reductions of VOC and NOx were employed in an airshed model by Milford el at. (1989). The top shows isopleths in two dimensions as presented by Milford et at. (f989), and the bottom shows these data extrapolated to three dimensions (from Finlayson-Pitts and Pitts, f993). 

Besides meeting its assumptions, other problems in the application of SSD in risk assessment to extrapolate from the population level to the community level also exist. First, when use is made of databases (such as ECOTOX USEPA 2001) from which it is difficult to check the validity of the data, one does not know what is modeled. In practice, a combination of differences between laboratories, between endpoints, between test durations, between test conditions, between genotypes, between phenotypes, and eventually between species is modeled. Another issue is the ambiguous integration of SSD with exposure distribution to calculate risk (Verdonck et al. 2003). They showed that, in order to be able to set threshold levels using probabilistic risk assessment and interpret the risk associated with a given exposure concentration distribution and SSD, the spatial and temporal interpretations of the exposure concentration distribution must be known. 

Recently, metapopulation models have been successfully applied to assess the risks of contaminants to aquatic populations. A metapopulation model to extrapolate responses of the aquatic isopod Asellus aquaticus as observed in insecticide-stressed mesocosms to assess its recovery potential in drainage ditches, streams, and ponds is provided by van den Brink et al. (2007). They estimated realistic pyrethroid concentrations in these different types of aquatic ecosystems by means of exposure models used in the European legislation procedure for pesticides. It appeared that the rate of recovery of Asellus in pyrethroid-stressed drainage ditches was faster in the field than in the isolated mesocosms. However, the rate of recovery in drainage ditches was calculated to be lower than that in streams and ponds (van den Brink et al. 2007). In another study, the effects of flounder foraging behavior and habitat preferences on exposure to polychlorinated biphenyls in sediments were assessed by Linkov et al. (2002) using a tractable individual-based metapopulation model. In this study, the use of a spatially and temporally explicit model reduced the estimate of risk by an order of magnitude as compared with a nonspatial model (Linkov et al. 2002). 

Complex specific 2 Extrapolation by statistical model Spatial variation in community responses Comparing model ecosystem studies performed in different geographical regions Statistical comparison... 

Extrapolation by Spatial variation in community Comparing model ecosystem Food-web models, IFEM,... 

The tests cannot be extrapolated directly to field reservoir performance because the spatial geometry is different. The core tests have a linear, 1-dimensional flow geometry while the actual reservoir has radial, 3-dimensional flow. In 3-dimensional flow the displacement efficiency is typically less than that measured in linear displacement studies. Chilton (1987) showed in his computer simulation studies that, as compared with the linear flow case, the predicted oil produced was 10% less for the two-dimensional model and 27% less for the three-dimensional model. However, when mobility control was used with a tenfold decrease in carbon dioxide mobility, the calculated improvement in displacement efficiency was much less for the linear case than the three-dimensional case. This result indicates that the increase in displacement efficiency under field conditions should be greater than that recorded in these linear laboratory tests. 

A central issue for pesticide risk assessment is extrapolation from individual- to population-level effects and from small temporal and spatial scales to larger ones. Empirical methods to tackle these issues are limited. Models are thus the only way to explore the full range of ecological complexities that may be of relevance for ecological risk assessment. However, EMs are not a silver bullet. Transparency is key, and certain challenges exist, for example, translating model output to useful risk measures. To make full use of models and get them established for risk assessment, we need case studies that clearly demonstrate the added value of this approach (Chapter 10). 

Spatial variations in exposure This could, for instance, be relevant if some species failed to recover in a mesocosms study, whereas in the real world migration would enable recovery. In such situations ecological modeling may be used to extrapolate from mesocosm scale to landscape scale (e.g.,... 

Such simple models need validation and for this reason ETAD is conducting in a field study to investigate some representative dyes (at manufacturing sites and dyehouses) under a project termed Pathways of Colorants to the Environment. The environmental risk posed by a colorant is a function of both its inherent ecotoxicity and the concentrations attained in the environmental compartments. Unlike other substances eg, household detergents) which are emitted continuously, dyes releases result mainly from batch processes and result in spatial and temporal peak emissions. Obviously, short-time concentrations should be compared with acute data on ecotoxicity, whereas long-tom residual concentrations need to be cranpared with chronic effect levels. Because, data on chronic effects are not often available, empirical information serves as a basis for the effects assessment, ie, the extrapolation to a Predicted No Effect Concentration (PNEC). This PNEC value is to be compared with the so-called Predicted Environmental Concentration (PEC) in order to estimate safe levels of residual dye in the environment. Since it is the dissolved state in which a dyes may become biologically available, it is the aquatic environmental compartment which is primarily addressed here. Nonetheless, some consideration of the impact of dyes on sewage and soil is also included. 

---