*4.3. Comparison between Basin and Non-Basin Countries*

To compare findings from the within-basin assessment, and particularly to understand the degree to which outcomes strongly favored national interests, the same analysis was repeated to examine the level of agreement between basin and non-basin states. Contrary to the within basin states analysis, clusters between groupings of basin and non-basin states did not indicate a distinct separation (Figure 9). Clusters based on experts' professions between basin and non-basin states was also relatively indistinct at the factor scale.

**Figure 9.** Clusters based on membership in basin (red circles) or non-basin (blue diamonds) states. Where factor—#1 is geography, hydrology, ecology, and natural features, factor-#2 is socio-economic needs of basin states, factor-#3 is the population dependent on the watercourse, factor-#4 is the effects of water use, factor-#5 is existing and potential uses, factor-#6 is costs of conservation and protection, and factor-#7 is availability of alternative and comparable values and uses.

To address possible differences between basin and non-basin states on individual indicators outcomes from a t-test were examined (Table S2). This resulted in basin states experts having a significant difference with non-basin experts on 14 indicators. Three of these indicators were also identified as differences among basin states. These indicators included the average drought-affected people per year (I-6), red list index (I-11), total greenhouse gas emissions (I-12), environmental performance index (I-13), transport index (I-26), ICT index (I-27), education index (I-32), infant mortality rate (I-33), gender development index (I-35), areas exposed for severe soil erosion (I-52), protected forest area (I-53), cost to conserve erosion hot spot areas (I-58), employment in service provision sector (I-73), and revenue and job opportunity from ports (I-74),

To investigate the potential sources of these differences, we compared experts' responses by continent and the Middle East region. However, this did not indicate any potential geographic influence on the result. Similarly, the hydrologic position of non-basin countries did not appear to illustrate any distinct differences.

Comparing individual basin states with aggregated non-basin countries, for the 19 indicators (mean value), one or more basin states exhibited a statistically significant difference with non-basin countries (Figure 10), except for red list index value in each country (I-11). Predominantly, the non-basin experts' opinions aligned with upstream states, especially with Ethiopia. The largest gap for most of the indicators was observed between Egypt and non-basin experts. Sudan experts' opinions aligned closely with Egypt for indicators such as I-58 and I-74. However, for the remaining indicators, Sudan responses aligned more closely with upstream states, particularly with equatorial countries.

**Figure 10.** Mean value of all indicators for which individual basin states and non-basin states were statistically significantly different within, among, and between each other. | drought affected people per year (I-6), Red List Index (I-11), greenhouse gas emissions (I-12), environmental performance index (I-13), transport Index (I-26), ICT index (I-27), population living below poverty line (I-30), education index (I-32), infant mortality rate (I-33), gender development index (I-35), population growth rate (I-38), areas exposed for severe soil erosion (I-52), protected forest area (I-53), wetland area (I-56), estimated cost to conserve erosion areas (I-58), virtual water export (I-61), employment in service provision sector (I-73), revenue and job opportunity from ports (I-74), and water saving by crop pattern modification (I-75).

#### *4.4. Important Indicators within Nile Basin*

To categorize indicator consensus levels based on expert responses, both the statistical test results and the majority rules detailed in the methods section were applied. Accordingly, for 56 indicators, expert responses exhibited no significant difference, irrespective of their profession and country, and were classified as a high consensus level (Table S2). In contrast, for three indicators, there was no statistically significant agreement either among basin states or between basin and non-basin groups. Thus, they were classified at a level of low consensus. These three indicators included the average drought affected people per year (I-6), estimated cost to conserve erosion hot spot areas (I-58), and revenue and job opportunity from ports (I-74). Nine indicators with a statistically significant disagreement either among basin states or between the basin and non-basin states, yet experts rated

the indicators as very important more than50% of the time were classified at a level of moderate consensus. Finally, there were seven indicators for which there was no statistically significant agreement, either among basin states or between basin and non-basin groups. Experts rated the indicators as very important <50% of the time, and they were classified at a level of low consensus. As a result, 10 of the 75 indicators fell into the low consensus category.

## **5. Discussion**

To the best of our knowledge, this study is the first to identify indicators that define factors listed under the 1997 UN Watercourse Convention by involving multidisciplinary experts from five areas of expertise around the globe. Overall, though the majority of 215 experts tended to rate most of the indictors as important, there were also notable differences between indicators. The investigations into possible sources of these differences by clustering analysis indicate neither professional background nor experts' geographic home in terms of basin and non-basin states play a significant role. A clear grouping in data was observed, however, considering experts in upstream versus downstream countries. Still, this does not necessarily imply there are no difference between basin and non-basin or individual countries. Rather, the statistical test outcomes confirmed the existence of a significant difference for 8 (among basin countries) and 11 (between basin and non-basin countries) indicators. In addition, Egyptian experts had a significant difference on 2 of these 19 indicators among themselves.

The mean comparison of these 19 indicators suggests that Egypt experts' preferences are different from other countries. This also seems likely due to the influence of national interests than other scientific justification for water allocation. Considering the socioeconomic conditions of Egypt versus other basin states—with the largest per capita water storage, fewest impoverished people, fixed crop types, and largest port-based income [33,37–39]—the influence of nationalistic bias is not unexpected. Sudan exhibits a mix of characteristics. Given the high crop water requirements in both Egypt and Sudan as compared to other basin states [40], both are interested in virtual water trade (I-61). Similarly, experts from Sudan and Egypt valued the importance of considering costs to conserve severe erosion (I-58) and opportunities from ports (I-74) less. However, the two countries were distinctly different on the remaining indicators, with Sudan generally aligning more with other basin states (Figure A3–Appendix A). Comparing countries, these indicator preferences were generally aligned with the positions of basin states regarding the Cooperative Framework Agreement, the most recent legal instrument based on the principle of equitable and reasonable utilization still waiting for the required number of ratifications to go into force [31]. Nevertheless, none of the experts from the non-basin countries were observed to side with a particular basin country (Figure A4–Appendix A).

.Therefore, in addition to 24 unique indicators previously applied in different studies to inform fair share of basin states [8–11], this study introduced 51 additional indicators. Of these, 56 out of the 75 indicators were categorized as highly important, and of the remaining 19 indicators for which significant differences were observed, the level of consensus was labelled as moderate and low (9 and 10), respectively. Out of to the 24 indicators used in previous studies on different river basins, 16 of them were found highly relevant to the Nile basin.

#### **6. Conclusions**

The principle of equitable and reasonable utilization agreed to by Egypt, Sudan, and Ethiopia in their 2015 Declaration of Principles (DoP) does not readily allow for quantification of water sharing due to limitations of international laws in detailing measurable criteria for factors listed in this document and the UNWC guidelines. This study identified and evaluated basin-specific indicators by engaging professionals from basin and non-basin states. As observed from the summary of survey result, although experts from different geographic locations have divided responses on the importance of 75 proposed indicators, about 60% and 80% of basin and non-basin professionals, respectively, had a tendency to consider the majority of indicators to be relevant. The statistical tests revealed that experts grouped as basin versus non-basin and downstream versus upstream states had a significant difference

among, between, and within countries on 19 indicators. The findings also illustrate that the major differences in assigning importance levels to proposed indicators mainly occurred between Egypt and other countries, even more so than between basin and non-basin states. Furthermore, a clustering analysis indicates that these differences were likely more a result of national interests rather than profession background.

Though the position of Sudan varies from indicator to indicator, the findings of this study are generally similar to the position that basin countries have on the Cooperative Framework Agreement and current disagreements over the Grand Ethiopian Renaissance Dam. In conclusion, out of 75 indicators, multidisciplinary experts identified 56 as highly relevant indicators, while only 9 (10) indicators were categorized as moderate (less) relevant. On this basis, given the number of highly important indicators, the degree of disagreement is not interpreted to be as wide as reported in press releases following talks and negotiations discussing political engagements over the use of the Nile River. Rather, the indicators for which there is a high level of consensus can potentially lead to a solid foundation for putting the UNWC into action and pave the way for utilization of the Nile River based on the equitable and reasonable principle. Moreover, in addition to facilitating the allocation of water between riparian countries, breaking down these broad factors into measurable indicators could help legal and water resource policy makers to resolve disputes and mitigate controversial issues in the river basin. However, as highlighted in the results, since competing interests of the basin countries potentially influence expert's judgment and these interests vary from basin to basin, all indicators evaluated in this study for the Nile River basin may not be appropriate for other basins, as local context must be considered. Even within the same basin, regularly gauging expert opinions about indicators may also be important as countries and conditions change.

As the number of experts surveyed increases from all locations, the categories of moderately important indicators may change. However, we deem the 56 highly important indicators as adequate to improve measurability of the UNWC factors, capture the unique features of the Nile basin, and accommodate conflicting interests of states. Beyond laying a foundation for operationality of the UNWC, the outcomes of this study pave the way for the quantification of water allocations to riparian countries. Future work should address investigating the priority or weight of individual indicators for use in water resources modelling.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2073-4441/12/9/2499/s1, Table S1: Factors, Potential indicators and their description, Table S2: Cumulative percentage vote of indicators and the level of consensus among experts.

**Author Contributions:** Conceptualization, Y.G. and S.A.T.; Data curation, Y.G.; Formal analysis, Y.G.; Funding acquisition, G.A. and S.A.T.; Investigation, Y.G.; Methodology, Y.G. and S.A.T.; Project administration, S.A.T.; Resources, P.B., G.A. and S.A.T.; Software, Y.G.; Supervision, P.B. and S.A.T.; Validation, P.B. and S.A.T.; Visualization, Y.G.; Writing—original draft, Y.G.; Writing—review & editing, P.B., G.A., M.M. and S.A.T. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was financially supported by a collaborative partnership between the University of Calgary, Canada and Bahir Dar University (BDU), Ethiopia.

**Acknowledgments:** The authors would like to thank University of Wisconsin-Madison for facilitating data collection and research visit.

**Conflicts of Interest:** The authors declare no conflict of interest.
