A drought is a slowly developing natural phenomenon that can occur in all climatic zones and can be defined as a temporary but significant decrease in water availability. Over the past three decades, the cost of droughts in Europe has amounted to over 100 billion euros and it is expected to considerably increase as future droughts are projected to be more severe and long-lasting. Although drought monitoring and management are largely studied in the literature, they often fail at yielding precise information on critical events occurrence and associated impacts such as reduction of electricity production or crop failures. This is due to the difficulty of capturing the complexity of drought dynamics, which evolve over diverse temporal (and spatial) scales, including short-term meteorological droughts, medium-term agricultural droughts, and long-term hydrological droughts, as well as the non-physical aspects related to droughts (water management, irrigation, etc.). In this work, we contribute a Machine Learning based framework named FRIDA (FRamework for Index-based Drought Analysis) for the identification of impact-based, site-specific drought indexes. FRIDA is a fully automated data-driven approach that relies on advanced feature extraction algorithms to identify relevant drought drivers from a pool of candidate hydro-meteorological variables. Selected predictors are combined into an index representing a surrogate of the drought conditions in the considered area, including either observed or simulated water deficits or remotely sensed information about the state of the crops. FRIDA is portable across different contexts for supporting the formulation of basin-specific indexes to better inform drought management strategies. Several real-world examples will be used to provide a synthesis of recent applications of FRIDA in case studies featuring diverse hydroclimatic conditions and variable levels of data availability.

Demand-side management strategies based on customized feedback have proved their worth in supporting water conservation efforts and behavior change programs. Several studies in both the water and energy sectors report of observed short-term savings deriving from feedback-based programs and awareness campaigns, often based on smart metered data and high levels of customization in presenting information on resource usage to users in the form of past consumption, real-time information, peer comparison, analogies, and resource saving tips. Yet, feedback-based programs are often run as part of experimental trials with a limited duration, and their effectiveness is therefore only evaluated for a short time span, potentially overlooking rebound effects. Assessing the long-term effect of feedback information on behavior change is still an open research question. In this work, we analyze the long-term impacts of a smart-meter fed gamified ICT platform providing customized feedback to water users, which was deployed starting in 2014 in a long experiment trial with over 200 users of the Global Omnium utility in Valencia (Spain). The platform core is a data-driven demand management pipeline that enables water utilities to foster consumer engagement and promote water conservation via customized feedbacks. It includes customized water saving tips, peer-comparison of water usage, and a reward program based on gamification tools and mechanisms. After three years of development and testing from 2014 to 2017, the platform has proven to be very effective in the short-term, when a user is engaged. A 5.7% volumetric water use reduction among Global Omnium users was achieved after the first year of full implementation, along with a 20% approximate water consumption difference with respect to non-platform users. Here, we analyze the smart meter data of the platform users, respectively after one and two years from the end of the funded platform trial period, to assess long-term behavior changes and rebound effects on different groups of platform adopters.

The value of streamflow forecasts to inform water infrastructure operations has been extensively studied. Yet, their value in informing infrastructure design is still unexplored. In this work, we investigate how dam design is shaped by information feedbacks. We demonstrate how flexible operating policies informed by streamflow forecasts enable the design of less costly reservoir relative to alternatives that do not rely on forecast information. Our approach initially establishes information bounds by selecting the most informative lead times of perfect streamflow forecasts to be included in the infrastructure design. We then analyze the design and operational sensitivities relative to realistic imperfect streamflow forecasts synthetically modeled to explicitly represent different biases. We demonstrate our approach through an ex-post analysis of the Kariba dam in the Zambezi river basin. Results show that informing dam design with perfect forecasts enable attaining the same hydropower production of the existing dam, while reducing infrastructure size and associated capital costs by 20%. The use of forecasts with lower skill reduces this gain to approximately 15%. Finally, the adoption of forecast information in the operation of the existing system facilitate an annual average increase of 60 GWh in hydropower production. This finding, extrapolated to the new planned dams in the basin, suggests that consideration of forecast informed policies could yield power production benefits equal to 75% of the current annual electricity consumption of the Zambian agricultural sector. Forecast information feedbacks have a strong potential to become a valuable asset for the ongoing hydropower expansion in the basin.