Geology topics

Zhe Jiang

Publications and source records attributed to Zhe Jiang.

9 recordsLinked to original sources

USGS2025

Physics-guided fair graph sampling for water temperature prediction in river networks

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.

Proceedings of the AAAI Conference on Artificial I↗

USGS2024

Transfer learning with convolutional neural networks for hydrological streamline delineation

Hydrological streamline delineation is critical for effective environmental management, influencing agriculture sustainability, river dynamics, watershed planning, and more. This study develops a novel approach to combining transfer learning with convolutional neural networks that capitalize on image-based pre-trained models to improve the accuracy and transferability of streamline delineation. We evaluate the performance of eleven image-based pre-trained models and a baseline model using datasets from Rowan County, North Carolina, and Covington River, Virginia in the USA. Our results demonstrate that when models are adapted to a new area, the fine-tuned ImageNet pre-trained model exhibits superior predictive accuracy, markedly higher than the models trained from scratch or those only fine-tuned on the same area. Moreover, the pre-trained model achieves better smoothness and connectivity between classified streamline channels. These findings underline the effectiveness of transfer learning in enhancing the delineation of hydrological streamlines across varied geographies, offering a scalable solution for accurate and efficient environmental modelling.

North Carolina, Virginia↗

USGS2024

Adaptive fine-tuning for transferring a U-net hydrography extraction model using K-means

The United States Geological Survey (USGS) coordinates the collection of hydrographic features derived from remotely sensed interferometric synthetic aperture radar (IfSAR) elevation and intensity data in Alaska. Hydrographic features are cartographic representations of surface water features such as stream, rivers, lakes, ponds, canals, etc. Collection and validation procedures involve complex automated and manual techniques that furnish snapshots of hydrographic vector data that exist during the IfSAR surveys. The dynamic nature of fluvial conditions warrants monitoring and updating hydrographic data, but extraction procedures for updates can be cost prohibitive. This paper overviews progress on automated workflows to extract hydrography from IfSAR data using deep learning methods trained and tested with USGS collected hydrography data. This research tests transfer learning methods on a well-performing U-net model trained on a 4600-square kilometer (sq km) base model area in northcentral Alaska. The base model is transferred and fine-tuned to regions in the target domain covering roughly 127,000 sq km. The target domain is subdivided into areas with similar hydrogeomorphic conditions using principal components and k-means clustering, and the base model is adaptively fine-tuned to each hydrogeomorphic class by selecting training watersheds from each cluster within the target domain. Results are compared with transfer learning that is fine-tuned with a random sample of watersheds in the target domain.

Conference Paper↗

USGS2023

Transferring deep learning models for hydrographic feature extraction from IfSAR data in Alaska

The National Hydrography Dataset (NHD) managed by the U.S. Geological Survey (USGS) is being updated with higher-quality feature representations through efforts that derive hydrography from 3DEP HR elevation datasets. Deriving hydrography from elevation through traditional flow routing and interactive methods is a complex, time-consuming process that must be tailored for different hydrogeomorphic conditions. The large volume of surface water features and HR remote sensing data make manual annotation of the water features over the entire nation infeasible. Furthermore, annual and seasonal variations of surface waters warrant some level of periodic updates to hydrography. Advances in deep learning technologies provide an opportunity to automate hydrography extraction and scale up the process to a nationwide level. One major challenge, however, is the effect of spatial heterogeneity due to the wide variety of hydrogeomorphic conditions in the United States. In other words, it is unclear how a deep learning model pre-trained in one set of hydrogeomorphic conditions can be effectively applied to other conditions for hydrographic feature extraction. This paper aims to provide some clarity in this regard by testing automated deep learning and its transferability to the extraction of hydrography from digital elevation model (DEM) data spanning a range of hydrogeomorphic conditions in Alaska. In transfer learning, the knowledge (e.g., neural network weights) from one domain is transferred to other domains and thereby decrease training requirements in the target domain.

Alaska↗

USGS2022

Scaling-up deep learning predictions of hydrography from IfSAR data in Alaska

The United States National Hydrography Dataset (NHD) is a database of vector features representing the surface water features for the country. The NHD was originally compiled from hydrographic content on U.S. Geological Survey topographic maps but is being updated with higher quality feature representations through flow-routing techniques that derive hydrography from high-resolution elevation data. However, deriving hydrography through flow-routing methods is a complex process that needs to be tailored to different geographic conditions, which can lead to varying solutions. To address this problem, this paper evaluates automated deep learning and its transferability to extract hydrography from interferometric synthetic aperture radar (IfSAR) elevation data spanning a range of geographic conditions in Alaska.

Alaska↗

USGS2022

Modeling reservoir release using pseudo-prospective learning and physical simulations to predict water temperature

This paper proposes a new data-driven method for predicting water temperature in stream networks with reservoirs. The water flows released from reservoirs greatly affect the water temperature of downstream river segments. However, the information of released water flow is often not available for many reservoirs, which makes it difficult for data-driven models to capture the impact to downstream river segments. In this paper, we first build a state-aware graph model to represent the interactions amongst streams and reservoirs, and then propose a parallel learning structure to extract the reservoir release information and use it to improve the prediction. In particular, for reservoirs with no available release information, we mimic the water managers' release decision process through a pseudo-prospective learning method, which infers the release information from anticipated water temperature dynamics. For reservoirs with the release information, we leverage a physics-based model to simulate the water release temperature and transfer such information to guide the learning process for other reservoirs. The evaluation for the Delaware River Basin shows that the proposed method brings over 10% accuracy improvement over existing data-driven models for stream temperature prediction when the release data is not available for any reservoirs. The performance is further improved after we incorporate the release data and physical simulations for a subset of reservoirs.

Conference Paper↗

USGS2022

Weakly supervised spatial deep learning for Earth image segmentation based on imperfect polyline labels

In recent years, deep learning has achieved tremendous success in image segmentation for computer vision applications. The performance of these models heavily relies on the availability of large-scale high-quality training labels (e.g., PASCAL VOC 2012). Unfortunately, such large-scale high-quality training data are often unavailable in many real-world spatial or spatiotemporal problems in earth science and remote sensing (e.g., mapping the nationwide river streams for water resource management). Although extensive efforts have been made to reduce the reliance on labeled data (e.g., semi-supervised or unsupervised learning, few-shot learning), the complex nature of geographic data such as spatial heterogeneity still requires sufficient training labels when transferring a pre-trained model from one region to another. On the other hand, it is often much easier to collect lower-quality training labels with imperfect alignment with earth imagery pixels (e.g., through interpreting coarse imagery by non-expert volunteers). However, directly training a deep neural network on imperfect labels with geometric annotation errors could significantly impact model performance. Existing research that overcomes imperfect training labels either focuses on errors in label class semantics or characterizes label location errors at the pixel level. These methods do not fully incorporate the geometric properties of label location errors in the vector representation. To fill the gap, this article proposes a weakly supervised learning framework to simultaneously update deep learning model parameters and infer hidden true vector label locations. Specifically, we model label location errors in the vector representation to partially reserve geometric properties (e.g., spatial contiguity within line segments). Evaluations on real-world datasets in the National Hydrography Dataset (NHD) refinement application illustrate that the proposed framework outperforms baseline methods in classification accuracy.

ACM Transactions on Intelligent Systems and Techno↗

USGS2021

Extensibility of U-net neural network model for hydrographic feature extraction and implications for hydrologic modeling

Accurate maps of regional surface water features are integral for advancing ecologic, atmospheric and land development studies. The only comprehensive surface water feature map of Alaska is the National Hydrography Dataset (NHD). NHD features are often digitized representations of historic topographic map blue lines and may be outdated. Here we test deep learning methods to automatically extract surface water features from airborne interferometric synthetic aperture radar (IfSAR) data to update and validate Alaska hydrographic databases. U-net artificial neural networks (ANN) and high-performance computing (HPC) are used for supervised hydrographic feature extraction within a study area comprised of 50 contiguous watersheds in Alaska. Surface water features derived from elevation through automated flow-routing and manual editing are used as training data. Model extensibility is tested with a series of 16 U-net models trained with increasing percentages of the study area, from about 3 to 35 percent. Hydrography is predicted by each of the models for all watersheds not used in training. Input raster layers are derived from digital terrain models, digital surface models, and intensity images from the IfSAR data. Results indicate about 15 percent of the study area is required to optimally train the ANN to extract hydrography when F1-scores for tested watersheds average between 66 and 68. Little benefit is gained by training beyond 15 percent of the study area. Fully connected hydrographic networks are generated for the U-net predictions using a novel approach that constrains a D-8 flow-routing approach to follow U-net predictions. This work demonstrates the ability of deep learning to derive surface water feature maps from complex terrain over a broad area.

Alaska↗

USGS2021

An attention U-Net model for detection of fine-scale hydrologic streamlines

Surface water is an irreplaceable resource for human survival and environmental sustainability. Accurate, finely detailed cartographic representations of hydrologic streamlines are critically important in various scientific domains, such as assessing the quantity and quality of present and future water resources, modeling climate changes, evaluating agricultural suitability, mapping flood inundation, and monitoring environmental changes. Conventional approaches to detecting such streamlines cannot adequately incorporate information from the complex three-dimensional (3D) environment of streams and land surface features. Such information is vital to accurately delineate streamlines. In recent years, high accuracy lidar data has become increasingly available for deriving both 3D information and terrestrial surface reflectance. This study develops an attention U-net model to take advantage of high-accuracy lidar data for finely detailed streamline detection and evaluates model results against a baseline of multiple traditional machine learning methods. The evaluation shows that the attention U-net model outperforms the best baseline machine learning method by an average F1 score of 11.25% and achieves significantly better smoothness and connectivity between classified streamline channels. These findings suggest that our deep learning approach can harness high-accuracy lidar data for fine-scale hydrologic streamline detection, and in turn produce desirable benefits for many scientific domains.

North Carolina↗