Search for: All records

Groundwater depletion threatens global freshwater resources, necessitating urgent water management and policies to meet current and future needs. However, existing data-intensive approaches to assessments do not fully account for the complex human, climate, and water interactions within transboundary groundwater systems. Here, we present the design of and findings from a pilot participatory modeling workshop aiming to advance understanding of the hydrologic–human–climate feedback loops underpinning groundwater systems. Using participatory modeling tools and methods from the system dynamics tradition, we captured the mental models of researchers from water, social, data, and systems sciences. A total of 54 feedback loops were identified, demonstrating the potential of this methodology to adequately capture the complexity of groundwater systems. Based on the workshop outcomes, as an illustrative example, we discuss the value of participatory system modeling as a conceptualization tool, bridging perspectives across disciplinary silos. We further discuss how outcomes may inform future research on existing knowledge gaps around groundwater issues, and in doing so, advance interdisciplinary, use-inspired research for water decision-making more broadly. 

Neural methods are state-of-the-art for urban prediction problems such as transportation resource demand, accident risk, crowd mobility, and public safety. Model performance can be improved by integrating exogenous features from open data repositories (e.g., weather, housing prices, traffic, etc.), but these uncurated sources are often too noisy, incomplete, and biased to use directly. We propose to learn integrated representations, called EquiTensors, from heterogeneous datasets that can be reused across a variety of tasks. We align datasets to a consistent spatio-temporal domain, then describe an unsupervised model based on convolutional denoising autoencoders to learn shared representations. We extend this core integrative model with adaptive weighting to prevent certain datasets from dominating the signal. To combat discriminatory bias, we use adversarial learning to remove correlations with a sensitive attribute (e.g., race or income). Experiments with 23 input datasets and 4 real applications show that EquiTensors could help mitigate the effects of the sensitive information embodied in the biased data. Meanwhile, applications using EquiTensors outperform models that ignore exogenous features and are competitive with "oracle" models that use hand-selected datasets.