Dam Management in the Era of Climate Change
Cristina Di Girolami, M’hamed Gaïgi, Vathana Ly Vath, Simone Scotti · 2025
Read Dam Management in the Era of Climate Change by Cristina Di Girolami, M’hamed Gaïgi, Vathana Ly Vath, Simone Scotti (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2508-02636.txt.
Introduction
Cristina Di Girolami and co-authors model optimal dam management under climate change in a 2025 paper. Climate change concentrates rain into short bursts split by long dry spells, which makes dam control harder. The water level must stay below a safety threshold yet above a minimum that keeps the dam and the downstream river working.
Rain bursts and dry spells are modeled with a self exciting Hawkes process, which reproduces event clustering. The control task becomes an optimal switching problem with constraints, and the authors prove existence results and give numerical methods. In their examples, the optimal level falls as the self exciting strength rises, a response driven more by overtopping risk than by dry spell risk.
Catalog record
| CREATOR | Cristina Di Girolami, M’hamed Gaïgi, Vathana Ly Vath, Simone Scotti |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 69 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2508-02636
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Dam Management in the Era of Climate Change Cristina Di Girolami, M’hamed Gaïgi, Vathana Ly Vath, Simone Scotti Abstract Abstract Climate change has a dramatic impact, particularly by concentrating rainfall into a few short periods, interspersed by long dry spells. In this context, the role of dams is crucial. We consider the optimal control of a dam, where the water level must not exceed a designated safety threshold, nor fall below a minimum level to ensure functionality and sustainability for for the outgoing river. To model dry spells and intense rainfall events, commonly referred to as water bombs, we introduce a Hawkes process, a well-known example of a self-exciting process characterised by time-correlated intensity, which endogenously reproduces the concentration of events. The problem is formulated as an optimal switching problem with constraints. We establish existence results and propose numerical methods for approximating the solution. Finally, we illustrate the main achievements of this approach through numerical examples. The main and counterintuitive result of our numerical analysis is that the optimal water level inside the dam increases with the self-exciting parameter. This result shows that, when facing the dilemma of managing the opposing risks of dam overtopping and dry spells, the former ultimately dominates the latter. In conclusion, dams will increasingly lose their role as water reserves and take on a greater role in flood protection. Keywords : Climate change, Dry spells, Water bombs, Hawkes processes, Dam management, Optimal switching, Viscosity solution. JEL Classification : C61, Q54, O32. AMS Classification : 49L25, 49J40, 60G55, 60G57, 91G99. 1 Introduction Climate change has dramatic impacts on people’s lives and the global economy. How these impacts manifest is, however, very peculiar since, rather than a simple rise in global temperatures, they are characterised by an exponential increase in extreme events. In this paper, we focus mainly on freshwater, which is the keystone of life and the economy, as it is used for drinking, agriculture, industry, etc. Freshwater is essentially generated by rainfall, and its management was the cornerstone of the most magnificent historical civilisations in the world, such as the Romans, with the construction of aqueducts, the Chinese, Babylonians, Maya, and Khmer, with the construction of channels, reservoirs and lakes, see for instance [ 35 , Chapters 26 and 35] . More recently, the role of dams in storing freshwater and producing electricity has helped populate vast desert areas around the world, such as the Colorado Plateau with the construction of the Hoover Dam, or to harness monsoonal rains in tropical areas. Local populations have adapted their lives to the specific situation and periodicity of rainfall. However, the current climate disruption has placed populations in a position where they must find new solutions to manage water flows. In particular, a new significant risk arises: rainfall was previously relatively homogeneous across years, neglecting natural seasonality, with certain extreme situations clearly identified by large fluctuations, such as the El Nino/La Nina phenomena, see for instance [ 42 ] . In the new century, this relative stability has completely broken down, with the occurrence of long droughts, such as the 2011-17 drought in California, followed by catastrophic rainfall events, such as the winter of 2017, which was the wettest on record. The resulting floodwaters caused severe damage to the Oroville Dam in Northern California. The importance of modelling rainfall dynamics has grown in recent years to capture droughts, even in temperate climate zones such as Western Europe, North America, and East Asia, see [ 21 , 33 ] , as well as the dramatic increase in so-called “water bombs”, that is episodes of intense rainfall over a short period. Another important yet often overlooked aspect of dam inflow is its direct relationship with climate change. Specifically, the inflow to a mountain dam depends on both rainfall and snowmelt, see [ 33 ] . Historically, in Europe, rainfall was more prominent in spring, especially in April, while snowmelt peaked in early June. This seasonal pattern ensures a prolonged period of significant inflows, enabling reservoirs to replenish before summer, which is typically characterised by minimal or no rainfall. However, global warming has advanced the timing of snowmelt to April or May, compelling dam managers to release more water into downstream rivers during the spring. This shift increases the risks of flooding in spring and drought in summer. One consequence of this phenomenon is that Switzerland has decided to increase the height of certain dams by approximately 20 metres, see [ 31 ] , incurring significant maintenance and risk costs. Suitable mathematical models for dam management are of critical importance for technological, economic and ecological reasons, see [ 1 , 4 , 5 , 13 , 34 , 40 , 45 ] . Moreover, the significant environmental impact of water reservoirs underscores the need for reliable models for managing and controlling dams and water flows, see [ 40 , 46 ] . Strict regulations govern water flow management through dams to ensure both safety and efficiency. These regulations are often expressed via the so-called ”Rule Curves”, see [ 38 ] . A substantial body of literature addresses dam management as an optimal control problem; see [ 16 ] for a comprehensive survey. In this paper, we focus on the problem of dam management as the cornerstone of water management and electricity production in temperated areas. Due to its scarcity, the freshwater inside artificial lakes need to be controlled in order to guarantee multiple objectives. First, a minimal outflow needs to be ensured during drought. A second but not negligible role of dam is that hydroelectric plants represent the only perfectly green, renewable and not intermittent, since wind and solar are intermittent [ 3 ] and nuclear plants produces radioactive waste. Hydroelectric plants play then a crucial role on both decarbonation and electric grid stabilization. Mitigation of carbon emissions can be achieved by replacing fuel power plants with renewable power installations and has a huge impact on the CO2 market, see Krach et al. [ 25 ] , and contributes to the growth of dams construction worldwide with projects increasing in number and size. However, the non-intermittency of hydropower plants is increasingly challenged by extreme events such as droughts and heavy rainfalls, which force operators to manage water flows that are less and less uniform and often extreme. In mathematical point of view, we study the dam control problem as an optimal mixed switching problem. The dam manager must optimise revenues from power production while maintaining the water level within safe limits. The main challenge lies in managing water inflows, which we model using a marked Hawkes process. This process describes inflow dynamics as a marked jump process, where the intensity itself experiences sudden changes. This framework captures extended periods of dry spells and clustered rain bombs endogenously, rather than assuming rainfall to be homogeneously distributed over time. Hawkes processes [ 18 ] are well-suited for modeling such behaviours and are increasingly applied in various domains, including mathematical finance (see [ 2 , 6 , 15 , 24 , 36 ] ), water management (see [ 47 , 48 ] ), and, more recently, cyber risk (see [ 7 , 19 , 20 ] ). The main advantage of Hawkes processes, compared