Environ. Sci. Technol. 2010, 44, 3136–3141
Application of an Uncertainty Analysis Approach to Strategic Environmental Assessment for Urban Planning Y I L I U , * ,† J I N I N G C H E N , † W E I Q I H E , † QINGYUAN TONG,‡ AND WANGFENG LI‡ Department of Environmental Science and Engineering, and Urban Planning and Design Institute, Tsinghua University, Beijing 100084, China
Received February 4, 2010. Revised manuscript received February 4, 2010. Accepted February 17, 2010.
Urban planning has been widely applied as a regulatory measure to guide a city’s construction and management. It represents official expectations on future population and economic growth and land use over the urban area. No doubt, significant variations often occur between planning schemes and actual development; in particular in China, the world’s largest developing country experiencing rapid urbanization and industrialization. This in turn leads to difficulty in estimating the environmental consequences of the urban plan. Aiming to quantitatively analyze the uncertain environmental impacts of the urban plan’s implementation, this article developed an integrated methodology combining a scenario analysis approach and a stochastic simulation technique for strategic environmental assessment (SEA). Based on industrial development scenarios, Monte Carlo sampling is applied to generate all possibilities of the spatial distribution of newly emerged industries. All related environmental consequences can be further estimated given the industrial distributions as input to environmental quality models. By applying a HSY algorithm, environmentally unacceptable urban growth, regarding both economic development and land use spatial layout, can be systematically identified, providing valuable information to urban planners and decision makers. A case study in Dalian Municipality, Northeast China, is used to illustrate applicability of this methodology. The impacts of Urban Development Plan for Dalian Municipality (2003-2020) (UDP) on atmospheric environment are also discussed in this article.
1. Introduction SEA is a key approach to identifying potential solutions in harmonizing both the urban economy and environment. It is a tool intended to be used to evaluate the environmental impacts of development policies, plans and programs at an early stage in the overall planning process (1-3). SEA has gained widespread acceptability as a support tool for decision making in planning policy frameworks (4). Many countries, including China, have established laws and regulations to ensure that some regional and sectoral development plans should be combined with or examined by SEA (5, 6). * Corresponding author:
[email protected]. † Department of Environmental Science and Engineering. ‡ Urban Planning and Design Institute. 3136
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 8, 2010
Urban planning guide long-term urban expansion, economic growth and land use spatial layout. It causally relates to substantial environmental impacts over the urban region, therefore; it should take the environment into consideration when planning for a more sustainable pattern of urban development. Many studies have been carried out for SEA on urban planning and regional planning (6-9). These studies involve many different methodologies but, as Verheem and Tonk (10) have pointed out, there is still no general SEA approach. This is mainly due to the inherent complexity of SEA. Hilden et al. (11) stated that the major methodological problem for the prediction of impacts in SEA is from the high level of abstraction of policies and plans. The higher level of abstraction of the plan is, the greater likelihood of uncertainty occurs (12). The proposed actions are often not well-defined in detail and lack sufficient information in guiding development and construction. This in turn results in significant uncertainties in evaluating environmental changes during the implementation. Unfortunately, these problems are still rarely considered in current SEA researches. In fact, most research does not consider these uncertainties at all or if they do, they only do so for several alternative planning scenarios. Many existing SEA approaches are still based on qualitative judgments or lack adequate quantitative analysis (6, 13). This article aims to develop an integrated SEA methodology for the urban master plan (UMP) using uncertainty analysis techniques. This methodology, as discussed in Section 2, enables systematical examination of the uncertainties inherent in the UMP, and identification of key environmental changes with regard to structural and spatial variations. Section 3 is devoted to examining the applicability of this methodology to a case study area, i.e. Dalian Municipality. The Urban Development Plan for Dalian Municipality (2003-2020) (UDP), a comprehensive urban plan issued by the local government in 2003, is selected for this discussion. The plan-induced potential environmental impacts on the atmosphere are quantitatively analyzed. Some further discussions on methodological issues are also presented in the final section.
2. Materials and Methods 2.1. Integrating Uncertainty Analysis into SEA. 2.1.1. Uncertainty in the UMP and Related Environmental Impacts. As primary guidance for city construction and management, the UMP determines the direction of future urban growth focusing on three key aspects: (1) The city’s functions: defining the city’s role regarding regional city aggregation/ cluster and specifying population size and urbanization level; (2) Economic development: providing economic objectives consisting of economic scale and pillar industrial sectors; (3) Spatial allocation: laying out land use patterns and types over urban areas in categories such as residential, industrial and infrastructure sections. Because socioeconomic and political factors are always changing, a city’s actual growth often strays away from the official expectation presented in the UMP (12, 14). From an environmental perspective, the variations of population development, economic growth and land use change inevitably lead to significant uncertainties in future resource consumption and pollution emission. In an UMP, the total amount of population and economy (usually defined by GDP) is well stipulated for future growth by planning year. However, descriptions on the developments of individual industrial sectors are often vague. A sophisticated projection of industrial structure is not an essential 10.1021/es902850q
2010 American Chemical Society
Published on Web 03/03/2010
requirement for urban planners in market-based economiess or even in transitioning China. From a SEA perspective, however, the absence of a clear definition of the industrial structure leads to considerable difficulty in estimating the plan’s environmental impacts, as pollution emissions from different industries vary significantly. The most significant uncertainty comes from the industrial layout scheme. The classification of industrial lands in UMPs is comparatively rough regardless the variation of industrial environmental performances. Therefore, it is unclear to SEA practitioners where individual industrial sectors are located, respectively, and where the conflicts between urban expansion and the environment would occur. To summarize, the structural and spatial uncertainties inherent in urban plans inevitably lead to fluctuations of the intensity and distribution of environmental impacts. 2.1.2. Uncertainty Analysis Techniques. Uncertainty is universally inherent to science, which should be recognized and addressed in projections (15, 16). While the future cannot be precisely predicted, identification of the trend, extent and range of the uncertainty becomes essential to decisionmaking. Massive efforts have been invested in the development of uncertainty analysis methods. Two of the most frequently applied methods to outlook future are scenario analysis and the Monte Carlo simulation. Scenario analysis has been widely used in SEA to describe a few typical trends which may lead to significant environmental impacts (17-19). Scenario analysis proposes plausible future scenarios that contain the major uncertainties (20, 21), and describes some key directions of future developments based on the present situation and logical chains of plausible events and their interactions (22). Regarding the uncertainties inherent in UMPs, however, conventional scenario analysis based on enumerative method and “subjective” judgment cannot be served as a systematically and quantitatively analytical tool. It can be applied to address some key changes of industrial structure, but not the variation of industrial land layout by sector. Therefore, a grid-based analytical approach is needed to handle the spatial uncertainty of industrial distribution. Monte Carlo approaches have been widely used in environment modeling (23-29). The Monte Carlo methods provide a computable way to approach the reality by generating a large number of samples according to given probability distribution. What a Monte Carlo approach can help is to generate a large enough and grid-to-grid sample space which represents all possibilities of various combinations of industrial lands over urban areas. Each single sample, containing full information on spatial distribution of industries, represents one possible development scenario. The aggregation of a sufficiently large amount of samples, therefore, theoretically contains all possibilities of the future industrial land use change. Compared to scenarios analysis, Monte Carlo approaches establish a continuum which contains all possible scenarios and statistical profiles of future development. Monte Carlo type approaches address uncertainty that can be expressed quantitatively, for which the simple quantitative estimates of uncertainty, that is, conventional scenario analysis, are not sufficient (30). Sensitive analysis is often operating for distinguishing the differentiation of the samples against certain criteria. In doing so, the HSY algorithm has been widely accepted, which was originally developed for application to poorly defined systems with insufficient information (31-33). The fundamental idea of the HSY algorithm is to divide the whole sample space into two subspaces according to given criterion. Then analysis is carried out upon those two subspaces to identify the sensitive parameters and critical unknown processes in the internal description of system behavior (23, 28, 34).
2.2. Uncertainty Analysis Based SEA. Combining a scenario analysis method and a Monte Carlo-process, this article develops an integrated SEA methodology to cope with the inherent uncertainties of the UMP, as shown in the Supporting Information (SI). A scenario analysis approach is applied to describe future possible development of urban size and industrial structure. On this basis, a stochastic simulation model, that is, SIMULAND (abbr. SIMulation of Urban LAND use change) is developed to generate a large number of spatial distributions of the industries (cf., SI). Environmental models are subsequently introduced to estimate the related environmental impacts of each spatial distribution under the given industrial structure. Consequently, environmentally sensitive industrial sectors and areas can be identified by using an HSY algorithm. This fosters a comprehensive assessment of the UMP and a proposal on alternative schemes and mitigating and complementary measures for the plan’s implementation. In principal, a finite number of scenarios can be designed to contain the uncertain range of the potential variations of future economic scale and industrial structure. The baseline scenario of GDP growth can be defined in accordance with the planning scheme. At lease a faster growth and a slower growth need to be taken into account according to the general development trajectory, socioeconomic conditions and unexpected factors. Given the certain amount of GDP, there could be many different combinations of industrial sectors. However, it is not necessary to enumerate all these possibilities in SEA. Likewise, the average trend and the best and worst situation are sufficient to estimate the full range of potential environmental impacts. The scenarios for industrial structures can be developed via various specific scenarios analysis techniques. Nevertheless, the historical economic growth, economic policies, global/regional competition and collaboration, market condition, and official perspectives presented in the UMP and other specific plans (e.g., plan for industrial development, etc.), should be taken into consideration. As long as these scenarios are constructed, a SEA will be prepared to calculate the future resource consumption and pollution emission over the implementation of the UMP with regard to the uncertainties, and to identify the environmental impacts of the uncertainties inherent to the development of economic scale and industrial structure. It is essentially important for SEA to quantify the spatial distribution of environmental pressure caused by the plan’s implementation. Given the significant uncertainty in context of regional land use (mainly regional industrial (re)allocation, as discussed in the previous section), the actual urban development can represent quite different environmental consequences at certain local areas. Therefore, a spatial uncertainty analysis method is introduced into our SEA framework, which is discussed in the following section. 2.3. The SIMULAND Model. A stochastic process based model, that is, SIMULAND, is developed, aiming to simulate, analyze, and assess the uncertainty of industrial distribution and the related environmental impacts. SIMULAND generates possible layouts of the newly emerged industries (keeping existing industries’ distribution constant in the future) under the given industrial structure, based on sampling strategies and principles. If the number of the sample is large enough, then these artificially generated layouts can represent all possibilities of industrial distribution. Then environmental models are operated to calculate the related environmental impacts of each sample. Finally, all samples should be subject to statistical tests based on given environmental criteria, such as environmental standard or environmental capacity. By doing so, irrational industrial distributions can be distinguished, involving both environmentally sensitive industrial sectors and sensitive areas, which may result in unacceptable environmental impacts at specific locations. This enables an VOL. 44, NO. 8, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3137
integrated assessment for the plan and a proposal on necessary amendments of the plan at a precautionary stage. For more sophisticated discussion on the SIMULAND model, see the SI.
3. Case Study 3.1. Case Study Area. Dalian is a semiprovincial level municipality, located in the southern part of the Liaodong Peninsula in northeast China, surrounded by the Huanghai and Bohai Sea (cf. Dalian Municipal Map in the SI). The municipality has gained its reputation as an important international shipping center in northeast Asia, a regional center of finance, commerce, tourism and information, and a production base for petrochemical, electronic, and manufacturing industries. The territory of Dalian Municipality spans 13 538 km2, consisting of seven districts and four counties. In 2004, the municipal’s total population reached 5.62 million, and GDP achieved RMB ¥196.2 billion, roughly U.S. $4600 per capita. The Urban Development Plan for Dalian Municipality (2003-2020) (UDP) is selected as the target plan for SEA. According to the UDP, the total GDP, population and urbanization rate would increase by 440%, 42%, and 70% by 2020, respectively, compared to 2000. It is not surprising that there are major concerns from both local government and the public regarding the potential environmental impacts that may be caused by the rapid growth of the population, and especially the economy. In 2005, the local authority commissioned Tsinghua University to implement a SEA program for the UDP, which was supported by the State Environmental Protection Administration (SEPA) as a national pilot project. The main objective of this program was to systematically integrate environmental considerations into urban development by identifying, preventing, and minimizing potential environmental impacts. This section will focus on the simulation of air quality over the whole of Dalian Municipality by applying the methodology discussed in previous sections. Specific attention is given to spatial uncertainty of environmental impacts, caused largely by future industrial distribution. The base year is 2003 and the projection year is 2020 in accordance with the planning year. 3.2. Operationalization of the SIMULAND. According to historical environmental monitoring data, three key air pollutants were identified for Dalian Municipality: total suspended particles (TSP), sulfur dioxide (SO2), and nitrogen dioxide (NO2). The major sources responsible for air pollutants include power plants, industrial and residential coalburning boilers, and vehicle emissions. The most significant spatial uncertainties come from industrial emission, while the layout of the road system and residential areas is clearly stipulated in the UDP. Accordingly, the SEA concentrated on the spatial variation of newly emerged industries and the related environmental impacts. The UDP provided sketchy descriptions for a few key industries and official expectations on their growth. However, this information is not sufficient to determine the details of the overall industrial structure and thus the environmental consequences. Combining the UDP and historical data of industrial growth in Dalian Municipality, a baseline growth scenario for various industrial sectors was constructed as shown in the SI. The scenario, representing the most likely trends on the GDP and the land use intensity of each industrial sector, provides a starting point for simulating all possible distributions of the increment of industrial growth over the municipal jurisdiction. The SIMULAND model requires generalization of the study area (the model was running on a personal computer, CPU 3.2 G, RAM 1.0 G, hard disk 80 GB.). The whole municipal territory was divided into 6135 grids. The size of each grid is 1500 × 1500 m. Obviously, not all grids will be changed 3138
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 8, 2010
into industrial land. The valid grids that are suitable for allocating new industries are defined by current land use and ecologically feasible space. The present land uses, including transportation land, urban and rural residential land, agriculture land, and industrial land, are supposed to remain unchanged throughout the planning period. The ecologically feasible space for urban development is a geographical boundary where physical and ecological feasibilities, including topography, geomorphology, hydrology, meteorology, and ecology are considered (35, 36). Eventually, a total of 1095 grids were identified as valid sampling space for new industrial growth. Subsequently, a spatial Monte Carlo sampling process was carried out over the valid grids to generate possible scenarios of future industrial distribution. Given certain convergence criteria, a total of 50 000 scenarios were stochastically generated. The related air quality of each scenario was simulated by the ISCST3 model, one of the most widely applied environmental air quality models developed by the U.S. Environmental Protection Agency (37-39). The emissions from industries, transportation, and urban and rural households were all considered. The pollutant concentration of SO2 was simulated separately for the heating and nonheating period. The environmental constraints for the air quality were determined by the Ambient Air Quality Function Regionalization of Dalian Municipality. This regulation defined three levels of environmental air quality, namely Class I, Class II, and Class III and the related environmental functioning zones over the whole municipality. The quality standards of SO2 for Class I to Class III are 0.02 mg/m3, 0.06 mg/m3, and 0.10 mg/m3, respectively. Those of TSP are 0.08t/km2/month, 0.20t/km2/month and 0.30t/km2/month, respectively. And those of NO2 are 0.04 mg/m3, 0.08 mg/m3, and 0.08 mg/m3, respectively. In principle, Class I was applied for important nature reserves and parks, Class III for special industrial areas, and Class II for other areas. By applying a HSY algorithm, the scenarios were further divided into two groups according to their environmental performance/compliance. The environmentally sensitive areas and industrial sectors were then identified, which we will discuss in the following section. 3.3. Results and Discussion. The average pollutant concentrations of all scenarios represent the general trends of environmental impacts of the UDP, as shown in Figure 1. Comparing the projection results with monitored data in 2004, it is revealed that the average concentrations of TSP and SO2 in 2020 would become lower than the current level by some 35-88%, as shown in Table 1. This is mainly due to an environmentally sound shifting of energy consumption stipulated in the UDP. Great effort would be made to replace coal by much cleaner gas, leading to less emission, especially TSP and SO2. The reduction of NO2 would not be significant. On the contrary, the NO2 concentration in some districts would increase by about 88% due to the increase in road transportation. Furthermore, the overload areas and environmental risk probability for each pollutant can be identified, as shown in Figure 1, the SI, and Table 2. The results show that different pollutants have distinct profiles. The potential overload areas of TSP and SO2 in the heating period would be extensive, roughly reaching 16-17% of the whole municipal area; whereas those of NO2 and SO2 in the nonheating period would be less significant. However, the environmental risk probability, defined as the percentage of the number of “overload” grids to the total (cf. SI), would vary considerably among pollutants. The highest environmental risk would take place in the NO2 overload locations, although the absolute area is the least. Comparatively, most overload scenarios of TSP
FIGURE 1. Spatial distribution of average pollutant concentrations over Dalian Municipality: (a) TSP, (b) NO2, (c) SO2 in the nonheating period, and (d) SO2 in the heating period.
TABLE 1. Comparison of Average Air Quality between 2004 and 2020 2004
2020
reduction ratio
district
TSP (t/km2/month)
SO2 (mg/m3)
NO2 (mg/m3)
TSP (t/km2/month)
SO2 (mg/m3)
NO2 (mg/m3)
TSP
SO2
NO2
Jingang Jinzhou Lvshun Pulandian Wafangdian Zhuanghe
0.156 0.205 0.143 0.169 0.190 0.230
0.032 0.037 0.024 0.021 0.028 0.011
0.039 0.028 0.018 0.020 0.033 0.008
0.101 0.100 0.027 0.040 0.039 0.027
0.013 0.010 0.004 0.004 0.004 0.003
0.048 0.047 0.016 0.029 0.028 0.015
35% 51% 81% 76% 79% 88%
59% 73% 83% 81% 86% 73%
-23% -68% 11% -45% 15% -88%
TABLE 2. Overload Areas and Distributions of Environmental Risk Probability of Each Pollutant overload area pollutant TSP NO2 SO2a SO2b
km2
%
2463.75 17.8 60.75 0.4 627.75 4.5 2270.25 16.4
distribution of different environmental risk probabilities 0-20% 20-40% 40-60% 60-80% 80-100% 98% 15% 87% 62%
1% 0% 6% 7%
a SO2 in the nonheating period. period.
0% 0% 5% 5% b
0% 0% 1% 6%
1% 85% 0% 20%
SO2 in the heating
and SO2 in the nonheating period concentrate in a low-risk range (0-20%). Some environmentally sensitive areas characterized by high environmental risk can be further identified. In general, TSP pollution in the three towns are not significant, whereas the maximum pollution load ratios fall into a range of 58-85%, or 48-72% at a given confidence level of 95%, as shown in Figure 2 and Table 3. The NO2 sensitive areas are some of the Class I functioning zones. The probability distribution of the pollution load ratio is very sharp, indicating that this contaminant is extremely sensitive for these areas. This is mainly because these areas are very close to urban main roads and residential areas without necessary buffer zones (cf. Table 3). The result shows that SO2 would be a serious pollutant during the heating period in the identified two sensitive areas. Comparatively, SO2 pollution in Xiangying town would be
less significant because the average pollution load ratio is 95% at a given confidence level of 95%, although its maximum figure could reach 127%. Jingang District is identified as the most sensitive area throughout the whole region, as the pollution load ratio would reach 135% at a confidence level of 95% and the maximum figure could even be as high as 164%. The analysis of the contribution of emission sources reveals that industrial sectors are the major responsible actors in Jingang District (cf. Table 3). Identification of the environmentally sensitive industrial sector via the HSY algorithm and sensitivity analysis approach (cf. SI) enables precautionary prevention of industrial pollution in specific areas. Taking the SO2 pollution in Jingang District as example, it reveals that the industry as a whole is not sensitive to this area (cf. SI). This implies that there are possibilities to satisfy the air quality by appropriate adjustment of the industrial structure and layout, without limiting the overall industrial growth scale. The electricity is identified as the sensitive industrial sector (IS ) 0.85). According to the joint distribution of pollution load ratio and electricity scale within this area, given the confidence level of 85%, the appropriate scale of electricity in Jingang District should be limited within 500 000 kW by 2020. To simplify the application, only one structural scenario was constructed, although the methodology itself can easily deal with multi- growth scenarios. But it can still be concluded that the methodology shows its applicability and flexibility by providing a rich amount of information for decision making, in dealing with the significant uncertain environmental impacts of the plan. Furthermore, by identifying the VOL. 44, NO. 8, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3139
FIGURE 2. Probability distribution of average pollution load ratio for selected environmentally sensitive areas.
TABLE 3. Pollution Load Ratios and Pollution Sources of Selected Environmentally Sensitive Areas PLa pollutant TSP NO2 SO2 a
2
area (km )
at 95%
max
industry
transportation
Huaershan Town Paoya Town Xuling Town 3 Class I areas Jingang District Xiangying Town
121.5 238.5 186.75 56.25 366.75 94.5
49% 48% 72% 115% 135% 95%
74% 58% 85% 118% 164% 127%
77% 90% 87% 20% 88% 75%
23% 10% 13% 38%
resident
42% 12% 25%
Pollution load ratio.
environmentally sensitive areas and industrial sectors, it enables a rational division of the whole region into several zones where different developing priorities can be set up, and different economic and environmental measures can be accepted. This methodology enables a systematical examination of the environmental consequences of the urban plan, both structurally and spatially. The underlying rationale of this methodology is that the future cannot be precisely predicted; therefore, smart decision making should be adopted, and the most environmentally acceptable solution should be implemented instead of the most optimized one. The methodology shows its value as it can simulate all possibilities of urban growth and the related environmental impacts under the universal uncertainty embedded in the urban plan. The methodology can be further generalized and applied to screening potential environmental impacts for all land use kinds planning. The systematical identification of environmental spatial constraints could offer a way for planners to take environmental considerations into account at an early stage. This also provides an opportunity for SEA practitioners and planners to work together toward an integration of socioeconomic, political and environmental considerations. In this regards, the ideas of this methodology are important for SEA’s practices worldwide and especially 3140
contribution of source
critical area
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 8, 2010
for China’s current movement toward an environmental early intervention into decision making.
Acknowledgments This article is supported by National Natural Science Foundation of China (No. 0701057). We thank the three anonymous reviewers for their instructive and detailed comments. We also thank for Professor John Crittenden for his suggestions.
Supporting Information Available A sound description of the SIMULAND model and extensive figures and tables which are related to background information of the case study, or to extensive discussions of interest mainly to specialists. This material is available free of charge via the Internet at http://pubs.acs.org.
Literature Cited (1) Therivel, R.; Wilson, E.; Thompson, S.; Heaney, D.; Pritchard, D. Strategic Environmental Assessment; Earthscan: London. 1992. (2) Fisher, B. Strategic Environmental Assessment in post-modern times. Environ Impact Asses Rev. 2003, 23, 155–170. (3) Noble, B. F. Strategic Environmental Assessment: what is it? And what makes it strategic. J. Environ. Assess. Policy Manag. 2000, 3, 203–224.
(4) Noble, B. F. The Canadian experience with SEA and sustainability. Environ. Impact Assess. Rev. 2002, 22, 3–16. (5) Partida´rio, M. R.; Clark, R. Perspectives on Strategic Environmental Assessment; CRC Press LLC: Boca Raton, 2000. (6) Dalal-Clayton, B.; Sadler, B. Strategic Environmental Assessment: A Sourcebook and Reference Guide to International Experience; Earthscan: London., 2005. (7) Diaz, M.; Illera, J. C.; Hedo, D. Strategic environmental assessment of plans and programs: A methodology for estimating effects on biodiversity. Environ. Manage. 2001, 28 (2), 267–279. (8) Mortberg, U. M.; Balfors, B.; Knol, W. C. Landscape ecological assessment: A tool for integrating biodiversity issues in strategic environmental assessment and planning. J. Environ. Manage. 2007, 82 (4), 457–470. (9) Willis, M. R.; Keller, A. A. A framework for assessing the impact of land use policy on community exposure to air toxics. Environ. Manage. 2007, 83 (2), 213–227. (10) Verheem, R. A. A.; Tonk, J. A. M. N. Strategic environmental assessment: One concept, multiple forms. Impact Assess. Project Appraisal 2000, 18, 117–182. (11) Hilden, M.; Valve, H.; Jonsdottir, S.; Balfors, B.; Faith-Ell, C.; Moen, B.; Peuhkuri, T.; Schmidtbauer, J.; Swensen, I.; Tesli, A. EIA and Its Application for Policies, Plans and Programmes in Sweden; Nordic Council of Ministers, TemaNordCopenhagen, Finland, Iceland, and Norway, 1998; p 567. (12) Christensen, K. S. Coping with uncertainty in planning. Am. Plann. Assoc. J. 1985, 51 (1), 63–73. (13) Implementing Strategic Environmental Assessment; Schmidt, M., Joa˜o, E., Albrecht, E., Eds.; Springer: Berlin, 2005. (14) Allmendinger, P. Planning in Postmodern Times; Routledge: New York, 2001. (15) Beck, M. B. ed. Environmental Foresight and Models: A Manifesto. Elsevier: Amsterdam. 2002. (16) Krayer von Krauss, M.; van Asselt, M. B. A.; Henze, M.; Ravetz, J.; Beck, M. B. Uncertainty and precaution in environmental management. Water Sci. Technol. 2005, 52 (6), 1–9. (17) Rhyne, R. Technological forecasting within alternative whole futures projections. Technol. Forecast. Social Change. 1974, 6, 133–162. (18) Godet, M.; Roubelat, F. Creating the future: The use and misuse of scenarios. Int. J. Strategic Manage. Long Range Plann. 1996, 29 (2), 164–171. (19) Van Notten, P.; Rotmans, J.; van Asselt, M.; Rothman, D. An updated scenario typology. Futures. 2003, 35, 423–443. (20) Rockafellar, R. T.; Wets, R. J. B. Scenario and policy aggregation in optimization under uncertainty. Math. Oper. Res. 1991, 16, 119–147. (21) Walz, A.; Lardelli, C.; Behrendt, H.; Gret-Regamey, A.; Lundstroma, C.; Kytzia, S.; Bebi, P. Participatory scenario analysis for integrated regional modeling. Landscape Urban Plann. 2007, 81, 114–131. (22) van den Berg, L. M.; Veeneklaas, F. R. Scenario building: art, craft or just a fashionable whim?; Schoute, J. F. T., Finke, P. A., Veeneklaas, F. R., Wolfert, H. P., Eds,; Scenario Studies for the Rural Environment; Kluwer Academic Publishers: Dordrecht, 1995.
(23) Beck, M. B. Water quality modeling: a review of the analysis of uncertainty. Water Resour. Res. 1987, 23 (8), 1393–1442. (24) Keesman, K. Software for identification of Ill-defined systems: a water quality example. Environ. Software 1989, 4 (1), 28–34. (25) Chen, J.; Wheater, H. S. Identification and uncertainty analysis of soil water retention models using lysimeter data. Water Resour. Res. 1999, 35 (8), 2401–2414. (26) ApSimon, H. M.; Warren, R. F.; Kayin, S. Addressing uncertainty in environmental modeling: A case study of integrated assessment of strategies to combat long-range transboundary air pollution. Atmos. Environ. 2002, 36 (35), 5417–5426. (27) Spear, R. C.; Grieb, T. M.; Shang, N. Parameter uncertainty and interaction in complex environmental models. Water Resour. Res. 1994, 30 (11), 3159–3169. (28) Wu, F. C.; Tsang, Y. P. Second-order Monte Carlo uncertainty/ variability analysis using correlated model parameters: application to salmonid embryo survival risk assessment. Ecol. Model. 1994, 72 (1), 1–20. (29) Chen, J.; Beck, M. B. Quality Assurance of Multi-Media Model for Predictive Screening Tasks, EPA/600/R-98/106; U. S. Environmental Protection Agency: Washington, DC, 1999. (30) Krayer von Krauss, M. P.; Janssen, P. H. M. Using the W&H integrated uncertainty analysis framework with non-initiated experts. Water Sci. Technol. 2005, 52 (6), 145–152. (31) Young, P. C.; Hornberger, G. M.; and Spear, R. C. Modeling badly defined systems-some further thoughts. Proceedings SIMSIG Simulation Conference; Australian National University: Canberra, Australia, 1978; p 24-32. (32) Hornberger, G. M.; Spear, R. C. Eutrophication in Peel InletsI. The problem-defining behavior and a mathematical model for the phosphorus scenario. Water Res. 1980, 14, 29–42. (33) Spear, R. C.; Hornberger, G. M. Eutrophication in Peel Inlet II. Identification of critical uncertainties via generalized sensitivity analysis. Water Res. 1980, 14, 43–49. (34) Osidele, O. O.; Beck, M. B. Identification of model structure for aquatic ecosystems using regionalized sensitivity analysis. Water Sci. Technol. 2001, 43 (7), 271–278. (35) Marull, J.; Pino, J.; Mallarach, J. M.; Cordobilla, M. J. A land suitability index for strategic environmental assessment in metropolitan areas. Landscape Urban Plann. 2007, 81 (3), 200– 212. (36) Liu, Y.; Li, T.; Chen, J.; Zhang, L.; Wang, W.; Tong, Q.; Lv, C. Ecologically feasible space for urban development: A methodology and case study on Dalian Municipality, Northeastern China. China Environ. Sci. 2007, 27 (1), 34–38. (37) Elbir, T. Application of an ISCST3 model for predicting urban air pollution in the Izmir metropolitan area. Int. J. Environ. Pollut. 2002, 18 (5), 498–507. (38) Hao, J. M.; Wang, L. T.; Li, L.; Hu, J. N.; Yu, X. C. Air pollutants contribution and control strategies of energy-use related sources in Beijing. Sci. China Ser. D: Earth Sci. 2005, 48, 138–146. (39) Venegas, L. E.; Mazzeo, N. A. Modeling of urban background pollution in Buenos Aires City (Argentina). Environ. Model. Software 2006, 21 (4), 577–586.
ES902850Q
VOL. 44, NO. 8, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3141
