More Like this

1. Emerging investigator series: disaggregating residential sector high-resolution smart water meter data into appliance end-uses with unsupervised machine learning

   https://doi.org/10.1039/D0EW00724B  

   Bethke, Gabrielle M.; Cohen, Abigail R.; Stillwell, Ashlynn S. (March 2021, Environmental Science: Water Research & Technology)

   null (Ed.)

   The residential sector accounts for a significant amount of water consumption in the United States. Understanding this water consumption behavior provides an opportunity for water savings, which are important for sustaining freshwater resources. In this study, we analyzed 1-second resolution smart water meter data from a 4-person household over one year as a demonstration. We disaggregated the data using derivative signals of the influent water flow rate at the water supply point of the home to identify start and end times of water events. k -means clustering, an unsupervised machine learning method, then categorized these water events based on information collected from the appliance/fixture end uses. The use of unsupervised learning reduces the training data requirements and lowers the barrier of implementation for the model. Using the water use profiles, we determined peak demand times and identified seasonal, weekly, and daily trends. These results provide insight into specific water conservation and efficiency opportunities within the household ( e.g. , reduced shower durations), including the reduction of water consumption during peak demand hours. The widespread implementation of this type of smart water metering and disaggregation system could improve water conservation and efficiency on a larger scale and reduce stress on local infrastructure systems and water resources. 

   more » « less
2. High resolution synthetic residential energy use profiles for the United States

   https://doi.org/10.1038/s41597-022-01914-1  

   Thorve, Swapna; Baek, Young Yun; Swarup, Samarth; Mortveit, Henning; Marathe, Achla; Vullikanti, Anil; Marathe, Madhav (December 2023, Scientific Data)

   Abstract Efficient energy consumption is crucial for achieving sustainable energy goals in the era of climate change and grid modernization. Thus, it is vital to understand how energy is consumed at finer resolutions such as household in order to plan demand-response events or analyze impacts of weather, electricity prices, electric vehicles, solar, and occupancy schedules on energy consumption. However, availability and access to detailed energy-use data, which would enable detailed studies, has been rare. In this paper, we release a unique, large-scale, digital-twin of residential energy-use dataset for the residential sector across the contiguous United States covering millions of households. The data comprise of hourly energy use profiles for synthetic households, disaggregated into Thermostatically Controlled Loads (TCL) and appliance use. The underlying framework is constructed using a bottom-up approach. Diverse open-source surveys and first principles models are used for end-use modeling. Extensive validation of the synthetic dataset has been conducted through comparisons with reported energy-use data. We present a detailed, open, high resolution, residential energy-use dataset for the United States. 

   more » « less
3. Comparative Analysis of Supervised Classification Algorithms for Residential Water End Uses

   https://doi.org/10.1029/2023WR036690  

   Heydari, Zahra; Stillwell, Ashlynn_S (June 2024, Water Resources Research)

   Abstract Water sustainability in the built environment requires an accurate estimation of residential water end uses (e.g., showers, toilets, faucets, etc.). In this study, we evaluate the performance of four models (Random Forest, RF; Support Vector Machines, SVM; Logistic Regression, Log‐reg; and Neural Networks, NN) for residential water end‐use classification using actual (measured) and synthetic labeled data sets. We generated synthetic labeled data using Conditional Tabular Generative Adversarial Networks. We then utilized grid search to train each model on their respective optimized hyperparameters. The RF model exhibited the best model performance overall, while the Log‐reg model had the shortest execution times under different balanced and imbalanced (based on number of events per class) synthetic data scenarios, demonstrating a computationally efficient alternative for RF for specific end uses. The NN model exhibited high performance with the tradeoff of longer execution times compared to the other classification models. In the balanced data set scenario, all models achieved closely aligned F1‐scores, ranging from 0.83 to 0.90. However, when faced with imbalanced data reflective of actual conditions, both the SVM and Log‐reg models showed inferior performance compared to the RF and NN models. Overall, we concluded that decision tree‐based models emerge as the optimal choice for classification tasks in the context of water end‐use data. Our study advances residential smart water metering systems through creating synthetic labeled end‐use data and providing insight into the strengths and weaknesses of various supervised machine learning classifiers for end‐use identification. 

   more » « less
4. Unsafe Events Detection in Smart Water Meter Infrastructure via Noise-Resilient Learning

   https://doi.org/10.1109/ICCPS61052.2024.00030  

   Oluyomi, Ayanfeoluwa; Abedzadeh, Sahar; Bhattacharjee, Shameek; Das, Sajal K (May 2024, IEEE)

   Residential smart water meters (SWMs) collect real-time water consumption data, enabling automated billing and peak period forecasting. The presence of unsafe events is typically detected via deviations from the benign profile of water usage. However, profiling the benign behavior is non-trivial for large-scale SWM networks because once deployed, the collected data already contain those events, biasing the benign profile. To address this challenge, we propose a real-time data-driven unsafe event detection framework for city-scale SWM networks that automatically learns the profile of benign behavior of water usage. Specifically, we first propose an optimal clustering of SWMs based on the recognition of residential similarity water usage to divide the SWM network infrastructure into clusters. Then we propose a mathematical invariant based on the absolute difference between two generalized means – one with positive and the other with negative order. Next, we propose a robust threshold learning approach utilizing a modified Hampel loss function that learns the robust detection thresholds even in the presence of unsafe events. Finally, we validated our proposed framework using a dataset of 1,099 SWMs over 2.5 years. Results show that our model detects unsafe events in the test set, even while learning from the training data with unlabeled unsafe events. 

   more » « less
5. Quantifying the effects of land use and model scale on water partitioning and water ages using tracer-aided ecohydrological models

   https://doi.org/10.5194/hess-25-2239-2021  

   Smith, Aaron; Tetzlaff, Doerthe; Kleine, Lukas; Maneta, Marco; Soulsby, Chris (January 2021, Hydrology and Earth System Sciences)

   null (Ed.)

   Abstract. Quantifying how vegetation mediates water partitioning at different spatialand temporal scales in complex, managed catchments is fundamental forlong-term sustainable land and water management. Estimations fromecohydrological models conceptualising how vegetation regulates theinterrelationships between evapotranspiration losses, catchment water storage dynamics, and recharge and runoff fluxes are needed to assess water availability for a range of ecosystem services and evaluate how these might change under increasing extreme events, such as droughts. Currently, the feedback mechanisms between water and mosaics of different vegetation and land cover are not well understood across spatial scales, and the effects of different scaleson the skill of ecohydrological models needs to be clarified. We used thetracer-aided ecohydrological model EcH2O-iso in an intensively monitored 66 km2 mixed land use catchment in northeastern Germany to quantify water flux–storage–age interactions at four model grid resolutions (250, 500, 750, and 1000 m). This used a fusion of field (including precipitation, soil water, groundwater, and stream isotopes) and remote sensing data in the calibration. Multicriteria calibration across the catchment at each resolution revealed some differences in the estimation of fluxes, storages, and water ages. In general, model sensitivity decreased and uncertainty increased with coarser model resolutions. Larger grids were unable to replicate observed streamflow and distributed isotope dynamics in the way smaller pixels could. However, using isotope data in the calibration still helped constrain the estimation of fluxes, storage, and water ages at coarserresolutions. Despite using the same data and parameterisation for calibration at different grid resolutions, the modelled proportion of fluxes differed slightly at each resolution, with coarse models simulating higher evapotranspiration, lower relative transpiration, increased overland flow, and slower groundwater movement. Although the coarser resolutions also revealed higher uncertainty and lower overall model performance, the overall results were broadly similar. The study shows that tracers provide effective calibration constraints on larger resolution ecohydrological modelling and help us understand the influence of grid resolution on the simulation of vegetation–soil interactions. This is essential in interpreting associated uncertainty in estimating land use influence on large-scale “blue” (ground and surface water) and “green” (vegetation and evaporated water) fluxes, particularly for future environmental change. 

   more » « less