Spatially explicit WEF modelling in transboundary river basins

Abstract

Water, energy and food resources are fundamental for human survival and are critical for supporting economic development. However, ensuring adequate supply is a major concern for the entire world, specifically in some countries and regions. Under the pressure of population growth, economic development, international trade, urbanization, diversifying diets, cultural and technological changes, global projections indicate a significant further increase in demand for water, energy and food over the next decades. Moreover, the development of these water, energy and food resources are intertwined. As a result, when demand grows, but resources are no longer abundant, competition between sectors increases. Especially in regions with upstream-downstream water connectivity, a national sectoral approach may result in friction, a decrease in mutual trust and international conflicts. On the other hand, the synergistic effects associated with regional resource coordination can contribute to improved resource availability and downstream livelihoods.

To overcome the shortcomings of the current generation of hydro-economic and WEF-nexus models in describing resource cooperation at regional level, in this study a new WEF-framework has been developed in which the heterogeneity in agro-climatic, socio-economic and resource availability, as well as the description of water and electric conveyance infrastructure is included spatially and temporally explicitly. The aim of this research is to create an integrated WEF-framework and to investigate the possibilities for, the relevance of and the challenges and difficulties associated with the implementation of such an integrated model.

The proposed framework includes the river conveyance infrastructure in a multi state river basin by means of a dynamic network model. In addition, a novel approach is used to describe both irrigated and rainfed agriculture in great detail on a regional scale, enabling a good representation of crops with multiple growing cycles per season, a distinction between annual and perennial crop management and the inclusion of agricultural losses. The framework is implemented as a model predictive control (MPC) problem. In this control problem, the reservoir operations and agricultural planning resulting in maximum economic value creation with the available resources are determined with the help of a non-linear problem solver. Receding horizon control accomplishes as part of the MPC framework feedback against uncertain disturbances (e.g. deviations in climate forcing) by applying only the optimal outputs in the first instance of the horizon in simulation and then updating the system states using new information. In addition, this control technique enables information exchange between riparian states within each MPC iteration. This allows us to add two new cooperation scenarios between the often studied scenarios of unilateralism and full coordination, with which the value of information exchange of river flows and trade flows can be studied.

Once developed, the framework is applied in the Eastern Nile basin. The Eastern Nile Basin is home to a large and rapidly growing population. Along with future population growth, changes in socio-economic conditions are expected, which will improve the coverage of the electricity grid and alter diets and water consumption. To meet the growing demand for food and energy, the Nile riparian countries have developed, and intend to further develop, their water resources. However, currently this development takes place unilaterally and can thereby threaten the livelihood in the downstream countries that are highly dependent on these water resources. The application of the proposed model framework aims to describe the qualitative and quantitative benefits and impacts of further collaboration in resource management within a predefined structural environment.

Simulation experiments with a monthly time step are conducted over a historical period between 1990-2010 and a future period between 2020-2040, by screening and incorporation of data on structural, socio-economic and climate constraints and demands. In addition to the four named cooperation scenarios, experiments are compiled to determine trade-offs between hydropower and agricultural water demand, the robustness of the solutions to imperfect climate foresights and the economic trade-offs related to different levels of agricultural self-sufficiency.

Comparative research with trade data from the FAO database indicates that historically every riparian state in the Eastern Nile basin could have benefited from the proposed integrated resource management, even in the unilateral national cases. Despite the variability in extent, all proposed and included forms of cooperation would have been beneficial for all individual states. Regionally, the flow-information, trade-information and regional coordination scenarios could have provided additional benefits of $32, $37 and $50 billion respectively throughout the period. Sharing information about the expected border flows would have generated, relatively in Sudan and absolutely in Egypt, by far the largest additional benefits. However, these benefits appear to correlate strongly with perfect climate foresight information. Because of its upstream location, Ethiopia could not have benefited economically from this flow information sharing. Overall, the benefits of resource optimization would have been relatively small in Ethiopia due to the limited infrastructure present during this period. In addition to the quantitative benefits mentioned, regional coordination also would have enabled the states to increase their resilience against long-lasting droughts and price fluctuations in the external market.

Results of the future model experiments suggest that every state will be disadvantaged in a regional coordinative scenario. To correct these physically and mathematically incorrect results, the current soft constrained implementation of the non-smooth complementarity relations for the reservoir filling process will have to be reconsidered. These non-smooth functions are the first out of three major difficulties encountered when implementing the proposed framework. Other difficulties arise when describing processes on different time scales (e.g. annual crop seasons) and when keeping the problem robust (in case predictions deviate from real events).

Overall, the case study illustrates that the proposed framework can account for spatial and temporal multisectoral trade-offs while finding non-trivial solutions for varying forms of national and regional cooperative resource management. Moreover, the operational resource reallocation choices proposed by this new framework and the spatial diversity in productivity that were discovered indicate that inadequate inclusion of these heterogeneities in WEF-nexus studies results in incomplete and potentially incorrect conclusions