|
ZEDD lab research is dedicated to advancing our understanding of water sustainability and developing innovative solutions to complex water resource challenges. Through interdisciplinary collaborations and the integration of cutting-edge data mining and control techniques, the lab aims to address pressing issues of water scarcity, equitable access of resources, and climate change impacts. |
1. Urban water planning under drought uncertainty.
|
Increasingly severe droughts are straining municipal water resources and jeopardizing urban water security.
The 20th century approach to ensuring robustness of an urban water system is through redundancy, i.e., oversizing traditional water infrastructure by a safety factor to buffer hydrological variability. A rapidly changing climate threatens this approach, exposing water planners to risks of either overbuilding costly water infrastructure or underbuilding and incurring supply failures. Through collaboration with industry experts at NAWI and the water planners in the city of Santa Barbara, this research develops new tools for long-term water supply planning that bring together expertise and scales that have mostly operated independently: climate, watershed, urban distribution, and technology models to optimize climate-robust urban water portfolios. Robust water supply portfolios for my case study in Santa Barbara, CA, were sensitive to the intensity, persistence, and severity of drought, underscoring the importance of accounting for climate change effects on drought characteristics. Water supply portfolios were also sensitive to source and technology attributes, suggesting value in a more diverse set of non-traditional water supplies and treatment technologies. [1] |
2. Decision-making in power-imbalanced, transboundary water systems.
|
Complex transboundary water systems are characterized by conflicting uses of water, upstream-downstream power imbalance, and risk of international tensions as shared resources become scarcer.
Among the potential sources of conflict, dam filling is a critical one. When a new mega dam is built, its reservoir is filled by withholding abundant river streamflow from downstream users. Filling megadams takes years, and their downstream impacts, occasionally magnified by ongoing droughts, have historically instigated international tensions in the Middle East and East Africa. For this research, part of a larger EU-funded project, I developed new methods for a more sustainable dam filling by collaborating with experts in social science, economy, ecology, agriculture, as well as local and national stakeholders and planners. I worked on the fast-developing Omo-Turkana basin shared between Ethiopia and Kenya, where Ethiopia, upstream, is expanding its hydropower production via the construction of two megadams. I contributed methods to predict climate oscillations by mining global datasets of sea surface temperatures, and used them to optimize timing and modulation of dam filling strategies that can reduce downstream impacts while preserving upstream interests. [2], [3], [4] |
3. Mitigating multisectoral conflicts with the smart use of information.
|
Designing operating rules for water systems is a challenging control problem that typically features a complex decision space and multiple conflicting operating objectives. I contributed the first multi-objective neuro-evolutionary algorithm for direct policy search (NEMODPS).
In NEMODPS, decision policies are parameterized as neural networks subject to automatic parameter and architectural optimization. Reservoir operation policies designed with NEMODPS prove more resistant to overfitting than the benchmarks, and perform better on extreme conditions outside the training dataset. Additionally, by combining feature extraction and neuro-evolution online, I developed an original methodology that optimizes decision policy architecture, parameters, and input set by selecting information that enables conflict mitigation between different operating objectives. More informed water reservoir operations typically require no investment, but hold great potential to mitigate critical conditions, from drought and flood impacts to conflicts for scarce resources. [5] [6] |
|
[1] Zaniolo, M., Fletcher, S., & Mauter, M. S. (2023). Multi-scale planning model for robust urban drought response. Environmental Research Letters, 18(5), 054014. [2] Zaniolo, M., Giuliani, M., Sinclair, S., Burlando, P., & Castelletti, A. (2021). When timing matters—misdesigned dam filling impacts hydropower sustainability. Nature Communications, 12(1), 3056. [3] Giuliani, M., Zaniolo, M., Castelletti, A., Davoli, G., & Block, P. (2019). Detecting the state of the climate system via artificial intelligence to improve seasonal forecasts and inform reservoir operations. Water Resources Research, 55(11), 9133-9147. [4] Zaniolo, M., Giuliani, M., Bantider, A., & Castelletti, A. (2021). Hydropower development: Economic and environmental benefits and risks. In The Omo-Turkana Basin (pp. 37-57). Routledge. [5] Zaniolo, M., Giuliani, M., & Castelletti, A. (2021). Neuro-evolutionary direct policy search for multiobjective optimal control. IEEE Transactions on Neural Networks and Learning Systems, 33(10), 5926-5938. [6] Zaniolo, M., Giuliani, M., & Castelletti, A. (2021). Policy Representation Learning for multiobjective reservoir policy design with different objective dynamics. Water Resources Research, 57(12), e2020WR029329. |
|
All are welcome.
|
|
