Advertisement

Water Resources Management

, Volume 32, Issue 15, pp 4953–4970 | Cite as

Robust Decision-Making Technique for Strategic Environment Assessment with Deficient Information

  • Daeryong Park
  • Myoung-Jin Um
Article
  • 79 Downloads

Abstract

This study developed a framework for an evaluative decision–making system that compensates for information deficiencies by considering the sensitivity of weight factor criteria. The developed decision–making system was applied to the strategic environmental assessment (SEA) for dam planning in South Korea. The SEA investigation included ten potential dam sites (four sites in the Han River, four in the Nakdong River, and two in the Geum River), performing an evaluative comparison of dam construction effects using environmental adequacy criteria and ranking the ten potential sites appropriately. Six different deficient information filling methods were applied: the normal and uniform distribution generations, Maximax, Maximin, Hurwicz, and the equal likelihood criterion. Results indicated sites GM2, HN3, ND4, and GM2 demonstrated the highest environmental adequacies across the combined effected rivers, the Han River, the Nakdong River, and the Geum Rivers, respectively. The probabilistic distribution generations, Hurwicz and the equal likelihood criterion methods produced similar priority scores and rankings based on different river conditions; however, the Maximax and Maximin methods calculated different priority scores and rankings, indicating users should be careful when applying the latter two methods for use in a decision support system (DSS) with deficient information. In future works, it will be necessary to apply other deficient data conditions or SEA examples to perform a more robust verification of the proposed framework.

Keywords

Deficient information Environmental adequacy Equal likelihood criterion Hurwicz Maximax Maximin Normal distribution Strategic environmental assessment Uniform distribution 

1 Introduction

Since the 1980s, Environmental Impact Assessments (EIAs) have been performed during water resource development projects to predict environmental effects and prepare mitigation methods. However, EIAs are typically implemented after some important planning decisions have been made at a high strategic level, and thus suffer from inherent limitations such as difficulties in rejecting projects causing serious environmental damage or altering project boundaries (Song et al. 2002). Strategic approaches are thus required to organize “environmentally sound and sustainable development” (ESSD) by evaluating environmental impacts from the initial stages of the decision–making process, such as during policy plan programs. As a result, the Strategic Environmental Assessment (SEA) was developed (Therivel et al. 1992; Sadler and Verheem 1996) and introduced to Korea between 2004 and 2006. Numerous water resources development plans have been categorized as targets for SEA implementation since this time. The long–term dam construction master plan, one of the major plans in the water resources development field, was designed to establish basic political strategies for dam construction on a national level and requires SEA for its establishment (Song et al. 2010).

SEAs typically evaluate natural resource sharing considering various objectives and different stakeholder views, a difficult task that has drawn extensive interest and various resolution attempts, many applying multi-criteria decision analysis (MCDA) methods (Ananda and Herath 2003; Sheppard and Meitner 2005; Herath and Prato 2006; Zardari and Cordery 2007; Hajkowicz 2008; Higgins et al. 2008; Lai et al. 2008; Ananda and Herath 2009; Chen et al. 2009). Unfortunately, decision–making processes using MCDA methods contain various uncertainties, including method and data uncertainties. These methods are, therefore, required to consider both their uncertainty factors and data, resulting in suggested probabilistic priority scores rather than deterministic scores.

Maxmin, Maximax, Hurwicz, and the equal likelihood criterion have been used for studying decision–making under complete uncertainty conditions (Kahraman and Tolga 1998; Wilson and Rosen 2005; Jun et al. 2013; Chung and Kim 2014). Under such conditions, these methods are simple and easy to apply; however, they must be combined with an MCDA method to quantify the resulting priority scores, and their results are therefore limited to showing the uncertainty of alternative priority rankings. For example, Jun et al. (2013) investigated streamflow depletion in an urban watershed in South Korea utilizing the Fuzzy TOPSIS method to consider climate change scenario uncertainties and criteria weight factors with robust approaches such as the Maximax, Maximin, Hurwicz, and the equal likelihood criterion applied to mitigate uncertainty and ambiguity levels in each alternative. Chung and Kim (2014) analyzed the robustness of prioritizations depending on various scenarios for the treated wastewater use locations using the fuzzy TOPSIS technique in the Anyangcheon watershed, South Korea. Using MCDA approaches to combine specialized data and stakeholder objectives to assist decision–making in various areas has been particularly beneficial in environmental decision–making (Huang et al. 2011). MCDA methods can provide cogent frameworks for SEAs, using a logical approach to support accuracy and compromise in environmental issues (Ramanathan 2001). Saaty (1980, 1988) suggested analytical hierarchy priority (AHP), which is an MCDA method that has been widely used in multidisciplinary areas. AHP evaluates decision–making based on knowledge, instinct, and heuristics using suitable mathematical principles. AHP consists of a standard methodology, effectively saving time in the decision-making procedure, and improves the quality of solutions to complex problems. AHP has been used to successfully structure management and planning decision–making in business, sociology, science, and engineering areas. In particular, AHP has been used to evaluate the prioritization of civil engineering structure planning such as transportation structure planning (Javid et al. 2014; Quadros and Nassi 2015), water planning in semi-arid areas of Brazil (Garfì et al. 2011), environmentally conscious purchasing systems (Handfield et al. 2002), landfill site selections in Iran (Moeinaddini et al. 2010), landslide susceptibility mapping (Yalcin 2008; Kayastha et al. 2013), EIAs (Ramanathan 2001), flood risk index analysis in India (Sinha et al. 2008), and supplier selection in the supply chain model (Haq and Kannan 2006). Uncertainty analyses of MCDA have been performed to improve the accuracy and sensitivity of the methods as well as to provide more integrated comprehensive information in environmental and socio-economic sciences (Hyde et al. 2005; Herath and Prato 2006; Bryan and Crossman 2008; Lai et al. 2008; Hajkowicz 2008; Mosadeghi et al. 2009; Kim et al. 2015; Park et al. 2015). Notably, Park et al. (2015) and Kim et al. (2015) investigated SEA priority scores in the Long-term Plan for Dam Construction (LPDC) with incomplete information. Park et al. (215) generated the incomplete information using the normal and uniform distributions and applied AHP, PROMETHEE II, ELECTRE III, and compromise programming for the MCDA. Furthermore, they compared the priority rankings of five different river basin conditions depending on both incomplete data generation methods and MCDA methods. Kim et al. (2015) applied different classifications to incomplete SEA data applying the VIKOR method (which is an efficient method for optimizing the potential negative and positive impacts of alternatives). Incomplete information was filled in using a uniform distribution generation, and their work focused on the priority scores of entire river basin conditions.

The objective of this study was to propose a framework for performing SEA evaluations on plans with incomplete data in LPDC process as shown in Fig. 1. The investigation of different results from probabilistic distribution and robust priority approaches for filling in missing data and analyzing the similarities and differences of results was of particular interest. The uncertainties of criteria weight factors were also considered, and all possible weight factors were applied to the AHP criteria matrix alongside six different gap–filling approaches. Priority scores and rankings were represented as selection ratio or percent, as the uncertainty of incomplete information was evaluated as probabilistic rather than deterministic results. This study also investigated high–priority sites depending on river basin conditions. The results of this study provide decision–makers with a simple method of selecting environmentally adequate sites for dam construction depending on their river basin settings.
Fig. 1

Robust decision-making process with deficient information for strategic environment assessment in Long–term dam construction plan (modified from Song et al. 2010). Note: bold font indicates focus in this study

2 Materials and Methods

2.1 SEA for Dam Planning in South Korea

The LPDC is a national level plan in South Korea for water resources planning. It is implemented every 10 years and reviewed every 5 years based on altered social, economic, and environmental factors. The LPDC is now required to incorporate SEA because the South Korean government is focused on environmental as well as economic perspectives. Figure 2 shows the ten potential dam sites requiring SEA in the LPDC, further described in Table 1: four potential sites in the Han River basin (HN1–HN4), four in the Nakdong River basin (ND1–ND4), and two in the Geum River basin (GM1 and GM2). The LPDC, including SEA, investigated and identified the most satisfactory dam site, having minimum effects on the basin from an environmental perspective. All potential sites were characterized using the upstream river and mountainous regions, encompassing numerous legally protected species and landscape preservations as well as abandoned mines which may create environmental toxicity. These abandoned mines require investigation and assessment to determine their expected environmental toxicity effects and protect potential dam environments.
Fig. 2

Locations of possible dam construction sites in SEA

Table 1

Potential dam sites for SEA in the LPDC

Watershed

Sites

Site ID

Han River

Sooip stream 1

HN1

Sooip stream 2

HN2

Dal stream 1

HN3

Dal stream 2

HN4

Nakdong River

Im stream,

ND1

Mansoo stream

ND2

Gohyun stream

ND3

Hoenggye stream

ND4

Geum River

Bocheong 1

GM1

Bocheong 2

GM2

The SEA process for potential dam sites collected data in four categories: landscape (LA), ecology (EC), water quality (WQ), and toxicity (TC). This study selected data regarding the preservation areas and cultural scenic preservation for landscape assessment; animal and plant fauna for ecology; BOD, COD, TN, and TP for water quality; and the number of abandoned mines for toxicity criteria (see Table 2).
Table 2

Assessment criteria of potential dam sites

Criteria

Observed variables

Landscape Assessment (LA)

• Preservation area, cultural and scenic preservation

Ecology (EC)

• Land: plants, birds, mammals, insects, amphibians, and reptiles

• Water: fish, benthic macro-invertebrates, phytoplankton, and zooplankton

Water quality (WQ)

• BOD, COD, TN, TP

• Stream water quality assessment for investigation results

Toxicity (TC)

• Mines (including abandoned mines) in dam watersheds (assessment of potential soil pollution)

All survey data categories (Tables 3, 4, 5 and 6) were assigned scores from 1 to 9 depending on their assessment criteria. A higher score indicates greater environmental adequacy for dam construction in Tables 3, 4 and 5, whereas a lower score indicates greater environmental adequacy for dam construction in Table 6.
Table 3

Landscape assessment of potential dam sites

Sites

Preservation area

Cultural and scenic preservation

Worth

Evaluations

HN1

DMZ

Jigyeon Falls

Important

1

HN2

DMZ

Dutayeon Falls

Important

1

HN3

0

0

Not Important

7

HN4

Songnisan National Park

Yongchu Falls

Important

1

ND1

Jirisan National Park

Yongyudam Pond

Important

1

ND2

Jirisan National Park

Silsangsa Temple

Important

1

ND3

No data

No data

No data

No data

ND4

No data

No data

No data

No data

GM1

0

0

Not important

7

GM2

0

0

Not Important

7

Table 4

Number of endangered species, including level 2 legally protected species

Species

HN1

HN2

HN3

HN4

ND1

ND2

ND3

ND4

GM1

GM2

Plant

11

11

2

9

0

5

1

1

0

0

Mammals

2

2

1

1

1

2

0

1

0

0

Birds

2

1

2

1

2

1

0

0

1

0

Amphibians and reptiles

0

0

0

0

0

0

0

0

0

0

Fish

3

3

1

2

1

1

0

0

0

0

Insects

0

0

0

0

0

0

0

0

0

0

Invertebrates

0

0

0

0

0

0

0

0

0

0

Total

18

17

6

13

4

9

1

2

1

0

Evaluation

1

1

7

3

7

5

9

9

9

9

Evaluation: 1:17-20, 3:13-16, 5:9-12, 7:5-8, 9:0-4

Table 5

Water quality parameters

Parameters (mg/L)

HN1

HN2

HN3

HN4

ND1

ND2

ND3

ND4

GM1

GM2

BOD

0.32

0.30

0.81

0.76

1.58

1.09

0.71

0.45

0.52

0.80

COD

1.48

1.50

2.81

2.13

2.58

2.29

3.69

3.30

1.22

2.34

TN

1.324

1.245

2.484

2.546

1.119

0.901

4.459

2.911

0.863

2.581

TP

0.001

0.004

0.030

0.027

0.016

0.058

0.044

0.000

0.008

0.057

Evaluated BOD

9

9

9

9

7

7

9

9

9

9

Evaluated COD

7

7

7

7

7

7

5

5

7

7

Evaluated TN

1

1

1

1

1

3

1

1

3

1

Evaluated TP

9

9

5

7

7

3

5

9

9

3

Evaluation Mean

6.5

6.5

5.5

6

5.5

5

5

6

7

5

Table 6

Number of abandoned mining sites for potential toxicity

Potential toxicity

HN1

HN2

HN3

HN4

ND1

ND2

ND3

ND4

GM1

GM2

The number of mines

3

1

20

78

4

2

No Data

No Data

No Data

2

Evaluation

9

9

1

1

7

9

No Data

No Data

No Data

9

Table 3 shows the preservation area and cultural scenic preservation data for LA in the potential dam sites. Per the national reserve system, demilitarized zones (DMZs) and national parks (intended to preserve environments especially with famous waterfalls, reservoirs, or temples for cultural scenic preservation) were considered to contain valuable natural landscapes in this landscape assessment. No landscape assessment surveys were obtained from the two Geum river sites (GM1 and GM2), and only two scores (1 and 7) were assigned to the obtained surveys because of the difficulties inherent in evaluating comprehensive grades from qualitative surveys. The EC criterion, or surveyed numbers of legally protected species for six different animal and plant classifications in potential sites, are shown in Table 4. The EC criterion score was classified by the number of legally protected species: 0–4 species received a score of 1, 5–8 of 3, 9–12 of 5, 13–16 of 7, and 17–20 of 9. The BOD, COD, TN, and TP parameters obtained for WQ criterion are shown in Table 5. Each WQ criterion parameter was scored between 1 and 9, then averaged to calculate a simple evaluation number (similar to the EC criterion approach). Table 6 presents the number of upstream abandoned mines in the potential dam sites. These numbers were scored such that 0–3 mines received a score of 9, 4–7 of 7, 8–11 of 5, 12–15 of 3, and > 16 of 1 because fewer mines corresponded with more environmentally friendly conditions.

2.2 Framework

This study provides a framework for robust SEA priority evaluation starting with incomplete information. The ten potential sites were selected for SEA to reduce environmental impacts after dam construction, and environmental impact criteria data were collected. The AHP method, probabilistic generation method, and sensitivity of weight factors were applied to quantify scores and ranking the sites based on their predicted environmental adequacy. For the probabilistic distribution approach, this study applied a normal (particularly integer numbers with a normal distribution), and uniform distributions. Each distribution had 1000 generated to fill in missing information. For the robust decision theory approach, the Maximax, Maximin, Hurwicz, and the equal likelihood criterion methods were utilized. The framework of this study contained four stages: the first stage was to build a performance matrix including missing data spaces based on AHP formulation, where development was performed via hierarchized grading of multiple environmental criteria based on environmental adequacy; the second was to fill missing information in the matrix, for which the numbers generated by probabilistic distributions and robust decision theory methods were used; the third was to evaluate alternatives by applying the AHP method to the completed matrix, dependent on river basin–combination conditions; and the final step was to evaluate the combination AHP scores depending on decision–making factors. The final resulting rankings were represented as percent ratios of selection based on uncertain missing information satisfaction. The percent depiction in the final results could provide more accurate information to decision–makers.

2.3 AHP Method

This study applied the AHP method, which has been shown to be a suitable approach in the past, to evaluate the environmental prospects of ten potential sites composed of four environmental factors. The AHP methodology process is as follows (Saaty 1980; Bhushan and Rai 2007; Saaty and Vargas 2012). First, AHP organizes a hierarchy consisting of the target, criteria, and alternatives required for this study. Second, AHP qualitatively compares matrices constructed using a pairwise comparison of various criteria from the retrieved information. Third, the principal eigenvalue and corresponding normalized right eigenvector for the comparison matrix provide the relative importance values of the various criteria. Fourth, to determine the inconsistency of the matrix using its level of redundancy, the coefficient of inconsistency (CI) is calculated as follows:
$$ CI=\frac{CR}{RI}=\frac{\uplambda_{\mathrm{max}}-n}{RI\left(n-1\right)}, $$
(1)
where n is the matrix size, RI is the random inconsistency index, λmax is the maximum eigenvalue, and CR is the consistency ratio. The value of CI should be less than 0.1 according to Saaty (1980). Finally, the rating score of each alternative is multiplied by the weights of the sub-criteria and integrated to estimate local rankings depending on each criterion.

2.4 Generation of Incomplete Information

To fill incomplete information in the environmental adequacy matrix, two different approaches were utilized: probabilistic distribution and robust decision–making. For the probabilistic distribution approach, 1000 generation sets were applied to the five missing information sets, creating a 5 × 1000 matrix with uniform and normal distributions. Normal distributions indicated an integer number whereas normal and uniform distributions implied equivalent selections. The generated information sets were applied to AHP for MCDA. For the robust decision–making approach, this study applied Maximax, Maximin, Hurwicz, and the equal likelihood criterion methods, all of which are well known and have been widely used as uncertainty conditions in MCDA. In the Maximax method, the largest evaluation value is selected as the optimal alternative after selecting the largest benefit among the evaluation values for each alternative criterion, whereas the Maximin method is a general decision–making technique that calculates and selects the smallest gain among the evaluation values for each alternative criterion, and the largest evaluation value is selected from these as the optimal alternative. This is because the criterion of importance differs between Maximax and Maximin. The Hurwicz method is to multiply by an α between 0 and 1 selected by the maximum evaluation value (which itself is selected by the decision maker) after choosing the evaluation value that takes the maximum benefit and the minimum benefit among the evaluation values for each alternative criterion. It is a decision technique that multiplies α and 1 − α by small and large evaluation values, respectively, and selects the alternative with the greatest sum of the two as the optimal alternative. The equal likelihood criterion method selects the largest value as an optimal alternative by averaging all the evaluation values of the alternatives.

2.5 Criteria Weight Factors

This study required the weight factors of the five criteria, which should have a total sum of one. It was assumed that the step of the weight factors was 0.1. All possible weight factor sets were then estimated as \( \left(\begin{array}{c}9\\ {}5\end{array}\right)=\frac{9!}{5!4!}=128 \). As such, 128 weight factor sets were applied for the completed AHP matrix using six different approaches: normal, uniform distributions, Maximax, Maximin, Hurwicz, and the equal likelihood criterion methods.

3 Results and Discussion

The first portion of this study investigated and compared the ranking changes of potential dam construction sites across entire basins with six different gap filling methods (Fig. 3). The second portion investigated the probabilistic priority scores as percentages for potential dam sites based on different river basin conditions (Figs. 4, 5, 6, 7, 8 and 9). All types of weight factor sets were used for AHP evaluation criteria with six different gap filling methods applied to consider the sensitivity of weight factors.
Fig. 3

Changes in potential dam site environmental adequacy rankings from six different decision methods

Fig. 4

Selection rates for environmentally adequate sites (depending on rivers) calculated by normal distribution generation

Fig. 5

Selection rate for environmentally adequate sites (depending on rivers) calculated by uniform distribution generation

Fig. 6

Selection rate for environmentally adequate sites (depending on rivers) calculated by Maximax

Fig. 7

Selection rate for environmentally adequate sites (depending on rivers) calculated by Maximin

Fig. 8

Selection rate for environmentally adequate sites (depending on rivers) calculated by Hurwicz method

Fig. 9

Selection rate for environmentally adequate sites (depending on rivers) calculated by the equal likelihood criterion method

Figure 3 shows the resulting rankings of the ten potential dam sites dependent on the six gap filling methods. The result for the Hurwicz method was the averaged ranking for all different α value cases. Four of the decision–making methods showed similar raking changes for potential dam sites; however, the Maximax and Maximin methods differed. The ND3 and ND4 sites show particularly different ranking trends between the Maximax and Maximin methods, showing very high (2.4 and 1.2) and very low (8.1 and 6.7) priorities calculated by the Maximin and Maximax methods, respectively. Overall, NH4 demonstrated the least environmental adequacy, whereas the most environmentally adequate site was GM2 (except when the Maximin method was used).

The different decision–making methods used to investigate the environmental adequacy of the selected rivers and river basins are shown in Figs. 4, 5, 6, 7, 8 and 9, with the adequacy results represented as selection percent.

Figure 4 shows the environmental adequacy results with normal distribution generation for missing information. The most environmentally adequate sites in each river (for the entire Han, Nakdong, and Geum Rivers) were GM2 (60%), HN3 (80%), ND4 (58%), and GM2 (64%), respectively. For all rivers combined, GM2 and GM1 demonstrated the first and the second highest adequacy rates, respectively. Six sites (all excluding ND3, ND4, GM1, and GM2) received a 0 % selection rate. In the Han River basin, HN3 had greater adequacy than HN1, which was the second most selected in that river. As above, HN2 and HN4 received 0 % selection rates. Conversely, in the Nakdong River ND4 was the most adequate with a slightly greater selection rate than ND3 (the second most adequate), and only ND1 received a 0 % selection rate. In the Geum River, GM2 (64%) was somewhat more adequate than GM1 (38%).

Figure 5 shows the environmental adequacy results with uniform distribution generation for missing information. All results were similar to those in Fig. 4 with normal distribution; however, the differences in the selection rates of Fig. 5 are smaller than those of Fig. 4.

Figure 6 shows the results of Maximax method, demonstrating that, for selection sites across the combined rivers, the Han River, and the Geum River, the same results were calculated. The Nakdong River, however, showed higher selection rates for ND1 and ND2 but zero or negligible selection rates for ND3 and ND4.

Figure 7 shows the adequacy results the Maximin method calculated for the river basins. All river environmental adequacy results were the same as those in Figs. 4 and 5 excepting the combined rivers and the Geum River. Sites ND4 and GM1 received the highest selection rates for combined rivers and the Geum River, respectively. It should be noted that ND4 and GM1 probabilistically showed pure (100%) selection rates for the Nakdong and Geum Rivers, respectively.

Figures 8 and 9 show the adequacy results produced by the Hurwicz and equal likelihood criterion methods for river basin sites. The Hurwicz and equal likelihood criterion methods calculated similar selection ratios for GM2 in the combined river analysis and the Geum River, respectively. The Hurwicz method in Fig. 8 never selected ND3 in the Nakdong River, similar to the Maximax and Maximin methods, and partially selected for HN1 in the Han River. Conversely, the equal likelihood criterion method (Fig. 9) had a slightly greater selection rate for ND4 than ND3 in the Nakdong River, but calculated a 100% selection rate for HN3 in the Han River.

All river–dependent environmental adequacy results in Figs. 8 and 9 are the same as those in Figs. 4 and 5; however, the selection rates of various sites are quite different from Figs. 4 and 5.

These results demonstrated that four of six decision–making methods (normal distribution generation, uniform distribution generation, Hurwicz, and the equal likelihood criterion) calculated the same highest–adequacy sites depending on river basins. The results for the Maximax and Maximin methods, in contrast, calculated different site adequacies. Notably, the Maximax method calculated the most adequate site to be ND2 (in the Nakdong River), whereas the Maxmin method represented the most adequate sites to be ND4 (between all rivers combined) and GM1 (in the Geum River).

As these results indicate the suggested system, combining the effects of the weight and missing information in the AHP method, could provide probabilistic environmental adequacy rankings among alternative sites for SEAs, providing accurate information under various conditions to decision–makers.

The results of this study, which applied the AHP method and included specific missing data, show that the normal distribution generation, uniform distribution generation, Hurwicz, and the equal likelihood criterion methods are more reliable than the Maximax and Maximin methods for decision–making to determine the environmental adequacy of dam construction sites depending on river basins.

The selection or priority orders of the sites yielded by the Maximax and Maximin methods are different from those of the other four methods. For example, for the selection percent or the number of selected sites, the results for the entire Nakdong River obtained by Maximax method and the results obtained for the entire Nakdong and Geum Rivers by the Maximin method differ from the results of the other four methods.

In addition, focusing on potential sites in the Han River, the selected sites obtained by the Hurwicz method coincide with the results obtained by the normal and uniform distributions (NH3 then NH1) whereas the selected site for the equal likelihood criterion method is only NH3. However, site selection rates for the four Maximax, Maximin, Hurwicz and the equal likelihood criterion methods in the Nakdong River are different from the results of the normal and uniform distributions, although the site priorities for Maximin, Hurwicz and the equal likelihood criterion methods are consistent with the results for the normal and uniform distributions. This indicates that the four Nakdong River sites contain a large amount of uncertain information (four missing data in Table 7). We revealed that these incomplete data caused the inconsistent priority results for the Maximax, Maximin, Hurwicz, and the equal likelihood criterion methods for the Nakdong River. However, the normal and uniform distribution generation methods obtain consistent priority results for the Nakdong River because these two distributions are symmetric distributions.
Table 7

Site assessment results based on environmental aspects

Criteria

HN1

HN2

HN3

HN4

ND1

ND2

ND3

ND4

GM1

GM2

Landscape (LS)

1

1

7

1

1

1

No Data

No Data

7

7

Ecology (EC)

1

1

7

3

7

5

9

9

9

9

Water quality (WQ)

6.5

6.5

5.5

6

5.5

5

5

6

7

5

Potential toxicity (PT)

9

9

1

1

7

9

No Data

No Data

No Data

9

The results in this study coincide with those of Jun et al. (2013), Chung and Kim (2014), and Kassa (2017). Jun et al. (2013) and Chung and Kim (2014) found that the Maximin and Maximax methods yield different priorities because these two methods are based on different assumptions (conservative or optimistic) and the results for Hurwicz are more reliable than those of the other three methods: Maximax, Maximin, and the equal likelihood criterion. In addition, Kassa (2017) analyzed the robust ranking of maize production in Ethiopia and showed that the Maximax and Hurwicz methods provided more consistent results than other methods.

4 Summary and Conclusions

This study shows an accurate evaluation approach for prioritizing the plans using SEA with missing data. The suggested system was adopted to perform the SEA required for the LPDC in South Korea, specifically estimating the environmental adequacy priority of ten potential sites as selection percentage rates. To fill in missing information, this study applied the normal and uniform distribution generation methods to evaluate the probabilistic approach as well as the Maximax, Maximin, Hurwicz, and the equal likelihood criterion methods to evaluate the non-probabilistic approach. After the data sets were completed, the environmental adequacy priorities of the combinations were evaluated by the AHP method.

This study investigated the environmental adequacy of sites depending on their locations in different rivers, performing an evaluation accounting for the entire river areas of the Han, Nakdong, and Geum Rivers. The dominant adequate sites were represented in the combined overall rivers and the Han River conditions; however, slightly more environmentally adequate sites were represented in the Nakdong and Geum Rivers. Furthermore, the normal and uniform distribution generations, the Hurwicz, and the equal likelihood criterion methods estimated similar environmental adequacy scores, whereas the Maximax and Maximin methods produced different environmental adequacy scores and rankings relative to the former four methods. The results of the study indicate that the probabilistic generation methods like normal and uniform distribution generation provide more consistent results than those of the non-probabilistic Maximax, Maximin, Hurwicz, and the equal likelihood criterion methods. The probabilistic generation, Hurwicz, and equal likelihood criterion methods presented consistent scores and rankings based on changing criteria weight factors. Conversely, the Maximax and Maximin methods calculated different scores and rankings, demonstrating more optimistic and conservative approaches, respectively. Given these observations, users should exercise caution when applying these two methods for use in the DSS of missing information.

This study investigated how the ranges of final priorities will change depending on different conditions with regard to the data distributions, hierarchy criteria, and applied decision theory methods using one example. The proposed framework thus requires additional applications using other missing data conditions or other SEA examples for more thorough verification. Based on such technical investigations, the observed weight sensitivities for the AHP method could be applied to build a SEA–DSS in the LPDC. Future research will use the proposed DSS framework to consider the flexibility of environmental criteria in SEA evaluations for the LPDC in South Korea. The expected effects and conclusions suggested by this study for the proposed framework can be verified following its implementation in SEAs in South Korea.

Notes

Acknowledgements

This paper was supported by Konkuk University in 2016. An initial abbreviated version of the paper has been presented at the 10th World Congress of EWRA “Panta Rhei”, Athens, Greece, 5–9 July, 2017.

Compliance with Ethical Standards

Conflict of Interest

None.

References

  1. Ananda J, Herath G (2003) The use of analytic hierarchy process to incorporate stakeholder preferences into regional forest planning. Forest Policy Econ 5:13–26CrossRefGoogle Scholar
  2. Ananda J, Herath G (2009) A critical review of multi-criteria decision making methods with special reference to forest management and planning. Ecol Econ 68:2535–2548CrossRefGoogle Scholar
  3. Bhushan N, Rai K (2007) Strategic decision making: applying the analytic hierarchy process. Springer Science & Business MediaGoogle Scholar
  4. Bryan BA, Crossman ND (2008) Systematic regional planning for multiple objective natural resource management. J Environ Manag 88:1175–1189CrossRefGoogle Scholar
  5. Chen Y, Yu J, Shahbaz K, Xevi E (2009) A GIS-based sensitivity analysis of multi-criteria weights. In: Proceedings of the 18th World IMACS/MODSIM Congress, Cairns, Australia pp 13–17Google Scholar
  6. Chung E-S, Kim Y (2014) Development of fuzzy multi-criteria approach to prioritize locations of treated wastewater use considering climate change scenarios. J Environ Manag 146:505–516CrossRefGoogle Scholar
  7. Garfì M, Ferrer-Martí L, Bonoli A, Tondelli S (2011) Multi-criteria analysis for improving strategic environmental assessment of water programmes. A case study in semi-arid region of Brazil. J Environ Manag 92:665–675CrossRefGoogle Scholar
  8. Hajkowicz SA (2008) Supporting multi-stakeholder environmental decisions. J Environ Manag 88:607–614CrossRefGoogle Scholar
  9. Handfield R, Walton SV, Sroufe R, Melnyk SA (2002) Applying environmental criteria to supplier assessment: a study in the application of the analytical hierarchy process. Eur J Oper Res 141:70–87CrossRefGoogle Scholar
  10. Haq AN, Kannan G (2006) Fuzzy analytical hierarchy process for evaluating and selecting a vendor in a supply chain model. Int J Adv Manuf Technol 29:826–835CrossRefGoogle Scholar
  11. Herath G, Prato T (2006) Using multi-criteria decision analysis in natural resource management. Ashgate Publishing, Ltd.Google Scholar
  12. Higgins AJ, Hajkowicz S, Bui E (2008) A multi-objective model for environmental investment decision making. Comput Oper Res 35:253–266CrossRefGoogle Scholar
  13. Huang IB, Keisler J, Linkov I (2011) Multi-criteria decision analysis in environmental sciences: ten years of applications and trends. Sci Total Environ 409:3578–3594CrossRefGoogle Scholar
  14. Hyde K, Maier H, Colby C (2005) A distance-based uncertainty analysis approach to multi-criteria decision analysis for water resource decision making. J Environ Manag 77:278–290CrossRefGoogle Scholar
  15. Javid RJ, Nejat A, Hayhoe K (2014) Selection of CO 2 mitigation strategies for road transportation in the United States using a multi-criteria approach. Renew Sust Energ Rev 38:960–972CrossRefGoogle Scholar
  16. Jun S-M, Chung E-S, Lee S-H, Kim Y (2013) Development and application of robust decision making technique considering uncertainty of climatic change scenarios. J Korea Water Resour Assoc 46:897–907CrossRefGoogle Scholar
  17. Kahraman C, Tolga E (1998) Fuzzy decision making under uncertainty methodology and tools in knowledge-based systems 263–271Google Scholar
  18. Kassa AM (2017) Application of decision making with uncertainty techniques: a case of production volume of maize in Ethiopia. Int J Qual Res 11(2)Google Scholar
  19. Kayastha P, Dhital MR, De Smedt F (2013) Application of the analytical hierarchy process (AHP) for landslide susceptibility mapping: a case study from the Tinau watershed, West Nepal. Comput Geosci 52:398–408CrossRefGoogle Scholar
  20. Kim Y, Park D, Um M-J, Lee H (2015) Prioritizing alternatives in strategic environmental assessment (SEA) using VIKOR method with random sampling for data gaps. Expert Syst Appl 42(22):8550–8556CrossRefGoogle Scholar
  21. Lai E, Lundie S, Ashbolt N (2008) Review of multi-criteria decision aid for integrated sustainability assessment of urban water systems. Urban Water J 5:315–327CrossRefGoogle Scholar
  22. Moeinaddini M, Khorasani N, Danehkar A, Darvishsefat AA (2010) Siting MSW landfill using weighted linear combination and analytical hierarchy process (AHP) methodology in GIS environment (case study: Karaj). Waste Manag 30:912–920CrossRefGoogle Scholar
  23. Mosadeghi R, Tomlinson R, Mirfenderesk H, Warnken J (2009) Coastal management issues in Queensland and application of the multi-criteria decision making techniques. J Coast Res 1252–1256Google Scholar
  24. Park D, Kim Y, Um M-J, Choi S-U (2015) Robust priority for strategic environmental assessment with incomplete information using multi-criteria decision making analysis. Sustainability 7:10233–10249CrossRefGoogle Scholar
  25. Quadros SGR, Nassi CD (2015) An evaluation on the criteria to prioritize transportation infrastructure investments in Brazil. Transp Policy 40:8–16CrossRefGoogle Scholar
  26. Ramanathan R (2001) A note on the use of the analytic hierarchy process for environmental impact assessment. J Environ Manag 63:27–35CrossRefGoogle Scholar
  27. Saaty TL (1980) The analytic hierarchy process. NY McGraw-Hill, USAGoogle Scholar
  28. Saaty TL (1988) What is the analytic hierarchy process? In: Mathematical models for decision support. Springer pp 109–121Google Scholar
  29. Saaty TL, Vargas LG (2012) Models, methods, concepts & applications of the analytic hierarchy process vol 175. Springer Science & Business MediaGoogle Scholar
  30. Sadler B, Verheem R (1996) Strategic environmental assessment 53: status, challenges and future directions. Ministry of Housing, Spatial Planning and the Environment of the NetherlandsGoogle Scholar
  31. Sheppard SR, Meitner M (2005) Using multi-criteria analysis and visualisation for sustainable forest management planning with stakeholder groups. For Ecol Manag 207:171–187CrossRefGoogle Scholar
  32. Sinha R, Bapalu G, Singh L, Rath B (2008) Flood risk analysis in the Kosi river basin, North Bihar using multi-parametric approach of analytical hierarchy process (AHP). J Indian Soc Remote Sens 36:335–349CrossRefGoogle Scholar
  33. Song Y-I, Seo S, Kim S, Choi J, Han S (2002) Improving prior environmental review system Korea Environmental Institute: Sejong City, KoreaGoogle Scholar
  34. Song Y-I, Park D, Shin G, Kim C, Grigg NS (2010) Strategic environmental assessment for dam planning: a case study of South Korea’s experience. Water Int 35:397–408.  https://doi.org/10.1080/02508060.2010.506268
  35. Therivel R, Wilson E, Thompson S, Heaney D, Pritchard D (1992) Strategic environmental assessmentGoogle Scholar
  36. Wilson JO, Rosen D (2005) Selection for rapid manufacturing under epistemic uncertainty. In: ASME 2005 International Design Engineering Technical Conferences and Computers and Information in Engineering Conference, 2005. American Society of Mechanical Engineers, pp 451–460Google Scholar
  37. Yalcin A (2008) GIS-based landslide susceptibility mapping using analytical hierarchy process and bivariate statistics in Ardesen (Turkey): comparisons of results and confirmations. Catena 72:1–12CrossRefGoogle Scholar
  38. Zardari N, Cordery I (2007) The use of a multicriteria method in irrigation water allocations. Agric J 2:236–241Google Scholar

Copyright information

© Springer Nature B.V. 2018

Authors and Affiliations

  1. 1.Department of Civil and Environmental EngineeringKonkuk UniversitySeoulRepublic of Korea
  2. 2.Department of Civil and Environmental EngineeringYonsei UniversitySeoulRepublic of Korea

Personalised recommendations