More Like this

1. Physics-Guided Recurrent Graph Model for Predicting Flow and Temperature in River Networks

   Jia, X.; Zwart, J.; Sadler, J.: Appling; Oliver, S.; Markstrom, S.; Willard, J.; Xu, S.; Steinbach, M.; Read, J.; Kumar, V. (April 2021, Proceedings of the 2021 SIAM International Conference on Data Mining)

   Demeniconi, Carlotta; Davidson, Ian (Ed.)

   This paper proposes a physics-guided machine learning approach that combines machine learning models and physics-based models to improve the prediction of water flow and temperature in river networks. We first build a recurrent graph network model to capture the interactions among multiple segments in the river network. Then we transfer knowledge from physics-based models to guide the learning of the machine learning model. We also propose a new loss function that balances the performance over different river segments. We demonstrate the effectiveness of the proposed method in predicting temperature and streamflow in a subset of the Delaware River Basin. In particular, the proposed method has brought a 33%/14% accuracy improvement over the state-of-the-art physics-based model and 24%/14% over traditional machine learning models (e.g., LSTM) in temperature/streamflow prediction using very sparse (0.1%) training data. The proposed method has also been shown to produce better performance when generalized to different seasons or river segments with different streamflow ranges. 

   more » « less
2. Multi-Scale Graph Learning for Anti-Sparse Downscaling

   https://doi.org/10.1609/aaai.v39i27.35014  

   Fan, Yingda; Yu, Runlong; Barclay, Janet R; Appling, Alison P; Sun, Yiming; Xie, Yiqun; Jia, Xiaowei (April 2025, Proceedings of the AAAI Conference on Artificial Intelligence)

   Water temperature can vary substantially even across short distances within the same sub-watershed. Accurate prediction of stream water temperature at fine spatial resolutions (i.e., fine scales, ≤ 1 km) enables precise interventions to maintain water quality and protect aquatic habitats. Although spatiotemporal models have made substantial progress in spatially coarse time series modeling, challenges persist in predicting at fine spatial scales due to the lack of data at that scale. To address the problem of insufficient fine-scale data, we propose a Multi-Scale Graph Learning (MSGL) method. This method employs a multi-task learning framework where coarse-scale graph learning, bolstered by larger datasets, simultaneously enhances fine-scale graph learning. Although existing multi-scale or multi-resolution methods integrate data from different spatial scales, they often overlook the spatial correspondences across graph structures at various scales. To address this, our MSGL introduces an additional learning task, cross-scale interpolation learning, which leverages the hydrological connectedness of stream locations across coarse- and fine-scale graphs to establish cross-scale connections, thereby enhancing overall model performance. Furthermore, we have broken free from the mindset that multi-scale learning is limited to synchronous training by proposing an Asynchronous Multi-Scale Graph Learning method (ASYNC-MSGL). Extensive experiments demonstrate the state-of-the-art performance of our method for anti-sparse downscaling of daily stream temperatures in the Delaware River Basin, USA, highlighting its potential utility for water resources monitoring and management. 

   more » « less
3. Spatio-Temporal Graph Neural Networks for Streamflow Prediction in the Upper Colorado Basin

   https://doi.org/10.3390/hydrology12030060  

   Akkala, Akhila; Boubrahimi, Soukaina Filali; Hamdi, Shah Muhammad; Hosseinzadeh, Pouya; Nassar, Ayman (March 2025, Hydrology)

   Streamflow prediction is vital for effective water resource management, enabling a better understanding of hydrological variability and its response to environmental factors. This study presents a spatio-temporal graph neural network (STGNN) model for streamflow prediction in the Upper Colorado River Basin (UCRB), integrating graph convolutional networks (GCNs) to model spatial connectivity and long short-term memory (LSTM) networks to capture temporal dynamics. Using 30 years of monthly streamflow data from 20 monitoring stations, the STGNN predicted streamflow over a 36-month horizon and was evaluated against traditional models, including random forest regression (RFR), LSTM, gated recurrent units (GRU), and seasonal auto-regressive integrated moving average (SARIMA). The STGNN outperformed these models across multiple metrics, achieving an R2 of 0.78, an RMSE of 0.81 mm/month, and a KGE of 0.79 at critical locations like Lees Ferry. A sequential analysis of input–output configurations identified the (36, 36) setup as optimal for balancing historical context and forecasting accuracy. Additionally, the STGNN showed strong generalizability when applied to other locations within the UCRB. These results underscore the importance of integrating spatial dependencies and temporal dynamics in hydrological forecasting, offering a scalable and adaptable framework to improve predictive accuracy and support adaptive water resource management in river basins. 

   more » « less
4. Physics-Guided Fair Graph Sampling for Water Temperature Prediction in River Networks

   https://doi.org/10.1609/aaai.v39i27.35025  

   He, Erhu; Kutscher, Declan; Xie, Yiqun; Zwart, Jacob; Jiang, Zhe; Yao, Huaxiu; Jia, Xiaowei (April 2025, Proceedings of the AAAI Conference on Artificial Intelligence)

   This work introduces a novel graph neural networks (GNNs)-based method to predict stream water temperature and reduce model bias across locations of different income and education levels. Traditional physics-based models often have limited accuracy because they are necessarily approximations of reality. Recently, there has been an increasing interest of using GNNs in modeling complex water dynamics in stream networks. Despite their promise in improving the accuracy, GNNs can bring additional model bias through the aggregation process, where node features are updated by aggregating neighboring nodes. The bias can be especially pronounced when nodes with similar sensitive attributes are frequently connected. We introduce a new method that leverages physical knowledge to represent the node influence in GNNs, and then utilizes physics-based influence to refine the selection and weights over the neighbors. The objective is to facilitate equitable treatment over different sensitive groups in the graph aggregation, which helps reduce spatial bias over locations, especially for those in underprivileged groups. The results on the Delaware River Basin demonstrate the effectiveness of the proposed method in preserving equitable performance across locations in different sensitive groups. 

   more » « less
5. Large Graph Property Prediction via Graph Segment Training

   Cao, Kaidi; Phothilimthana, Phitchaya Mangpo; Abu-El-Haija, Sami; Zelle, Dustin; Zhou, Yanqi; Mendis, Charith; Leskovec, Jure; Perozzi, Bryan (December 2023, Advances in neural information processing systems)

   Learning to predict properties of a large graph is challenging because each prediction requires the knowledge of an entire graph, while the amount of memory available during training is bounded. Here we propose Graph Segment Training (GST), a general framework that utilizes a divide-and-conquer approach to allow learning large graph property prediction with a constant memory footprint. GST first divides a large graph into segments and then backpropagates through only a few segments sampled per training iteration. We refine the GST paradigm by introducing a historical embedding table to efficiently obtain embeddings for segments not sampled for backpropagation. To mitigate the staleness of historical embeddings, we design two novel techniques. First, we finetune the prediction head to fix the input distribution shift. Second, we introduce Stale Embedding Dropout to drop some stale embeddings during training to reduce bias. We evaluate our complete method GST+EFD (with all the techniques together) on two large graph property prediction benchmarks: MalNet and TpuGraphs. Our experiments show that GST+EFD is both memory-efficient and fast, while offering a slight boost on test accuracy over a typical full graph training regime. 

   more » « less