Geology topics

Larry V. Stanislawski

Publications and source records attributed to Larry V. Stanislawski.

15 recordsLinked to original sources

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↗

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 NHD features are often digitized representations of historic topographic map blue lines and may be outdated.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↗

USGS2018

Streams do work: Measuring the work of low-order streams on the landscape using point clouds

The mutable nature of low-order streams makes regular updating of surface water maps necessary for accurate representation. Low-order streams make up roughly half the streams in the conterminous United States by length, and small inaccuracies in stream head location can result in significant error in stream reach, order, and density. Reliable maps of stream features are vital for hydrologic modeling, ecosystem research, and boundary monitoring. High resolution digital elevation models derived from lidar data have shown promise in low order stream modeling yet forested high relief landscapes and low relief agricultural areas remain challenging. Here we present early results from research analyzing lidar point clouds to identify features and patterns that may be used in low-order stream identification and classification in challenging geographic conditions. This work has identified characteristics derived from point clouds that correlate with the presence of streams and stream heads and show promise for mapping small streams. In low topographic relief agricultural areas, cross sections collected at regular intervals along drainage channels extracted as 3D lines show a significant jump in value and variance of profile curvature standard deviation at stream heads. In high relief areas, observations show potential for stream mapping by identifying trends in riparian zone structure. Lidar return point density from riparian vegetation under 30 feet tall dips in the vicinity of intermittent stream heads. Also seen is an increase in point density above 60 feet downstream of stream heads. The trends found here likely reflect a change in vegetation structure relative to the presence of streams.

Illinois, Iowa, North Carolina↗

USGS2018

Classifying physiographic regimes on terrain and hydrologic factors for adaptive generalization of stream networks

Automated generalization software must accommodate multi-scale representations of hydrographic networks across a variety of geographic landscapes, because scale-related hydrography differences are known to vary in different physical conditions. While generalization algorithms have been tailored to specific regions and landscape conditions by several researchers in recent years, the selection and characterization of regional conditions have not been formally defined nor statistically validated. This paper undertakes a systematic classification of landscape types in the conterminous United States to spatially subset the country into workable units, in preparation for systematic tailoring of generalization workflows that preserve hydrographic characteristics. The classification is based upon elevation, standard deviation of elevation, slope, runoff, drainage and bedrock density, soil and bedrock permeability, area of inland surface water, infiltration-excess of overland flow, and a base flow index. A seven class solution shows low misclassification rates except in areas of high landscape diversity such as the Appalachians, Rocky Mountains, and Western coastal regions.

International Journal of Cartography↗

USGS2018

Similarity assessment of linear hydrographic features using high performance computing

This work discusses a current open source implementation of a line similarity assessment workflow to compare elevation-derived drainage lines with the high-resolution National Hydrography Dataset (NHD) surface-water flow network. The process identifies matching and mismatching lines in each dataset to help focus subsequent validation procedures to areas of the NHD that more critically need updates.

Conference Paper↗

USGS2018

Area-preserving simplification of polygon features

Developing simplified representations of a two-dimensional polyline is an important problem in cartographic data analytics where datasets must be integrated across spatial resolutions. This problem is generally referred to as line simplification, and is increasingly driven by preservation of specific analytic properties such as positional accuracy and high-frequency detail. However, the distinction between linear features and polygon boundaries is rarely considered. Polygonal features differ fundamentally from linear features in that they represent areal regions covering a portion of the earth's surface. As such, they lend themselves naturally to the objective of preserving area across scale.

Conference Paper↗

USGS2018

Generalizing linear stream features to preserve sinuosity for analysis and display: A pilot study in multi-scale data science

Cartographic generalization can impact geometric properties of geospatial data and subsequent analyses. This study evaluates simplification methods with the goal of preserving geometric details, such as sinuosity. We evaluate two recently developed line simplification algorithms that introduce Steiner points: Raposo’s Spatial Means, and Kronenfeld’s new area-preserving segment collapse algorithm, and compare them with several well-known algorithms. Results indicate the area-preserving segment collapse algorithm optimally simplifies linear stream features with minimal horizontal displacement and the best retention of sinuosity.

Conference Paper↗

USGS2017

An open source high-performance solution to extract surface water drainage networks from diverse terrain conditions

This paper describes a workflow for automating the extraction of elevation-derived stream lines using open source tools with parallel computing support and testing the effectiveness of procedures in various terrain conditions within the conterminous United States. Drainage networks are extracted from the US Geological Survey 1/3 arc-second 3D Elevation Program elevation data having a nominal cell size of 10 m. This research demonstrates the utility of open source tools with parallel computing support for extracting connected drainage network patterns and handling depressions in 30 subbasins distributed across humid, dry, and transitional climate regions and in terrain conditions exhibiting a range of slopes. Special attention is given to low-slope terrain, where network connectivity is preserved by generating synthetic stream channels through lake and waterbody polygons. Conflation analysis compares the extracted streams with a 1:24,000-scale National Hydrography Dataset flowline network and shows that similarities are greatest for second- and higher-order tributaries.

Cartography and Geographic Information Science↗

USGS2017

Partial polygon pruning of hydrographic features in automated generalization

This paper demonstrates a working method to automatically detect and prune portions of waterbody polygons to support creation of a multi-scale hydrographic database. Water features are known to be sensitive to scale change; and thus multiple representations are required to maintain visual and geographic logic at smaller scales. Partial pruning of polygonal features—such as long and sinuous reservoir arms, stream channels that are too narrow at the target scale, and islands that begin to coalesce—entails concurrent management of the length and width of polygonal features as well as integrating pruned polygons with other generalized point and linear hydrographic features to maintain stream network connectivity. The implementation follows data representation standards developed by the U.S. Geological Survey (USGS) for the National Hydrography Dataset (NHD). Portions of polygonal rivers, streams, and canals are automatically characterized for width, length, and connectivity. This paper describes an algorithm for automatic detection and subsequent processing, and shows results for a sample of NHD subbasins in different landscape conditions in the United States.

Transactions in GIS↗

USGS2016

High performance computing to support multiscale representation of hydrography for the conterminous United States

The National Hydrography Dataset (NHD) for the United States furnishes a comprehensive set of vector features representing the surface-waters in the country (U.S. Geological Survey 2000). The high-resolution (HR) layer of the NHD is largely comprised of hydrographic features originally derived from 1:24,000-scale (24K) U.S. Topographic maps. However, in recent years (2009 to present) densified hydrographic feature content, from sources as large as 1:2,400, have been incorporated into some watersheds of the HR NHD within the conterminous United States to better support the needs of various local and state organizations. As such, the HR NHD is a multiresolution dataset with obvious data density variations because of scale changes. In addition, data density variations exist within the HR NHD that are particularly evident in the surface-water flow network (NHD flowlines) because of natural variations of local geographic conditions; and also because of unintentional compilation inconsistencies due to variations in data collection standards and climate conditions over the many years of 24K hydrographic data collection (US Geological Survey 1955).

Conference Paper↗

USGS2015

An evaluation of unsupervised and supervised learning algorithms for clustering landscape types in the United States

Knowledge of landscape type can inform cartographic generalization of hydrographic features, because landscape characteristics provide an important geographic context that affects variation in channel geometry, flow pattern, and network configuration. Landscape types are characterized by expansive spatial gradients, lacking abrupt changes between adjacent classes; and as having a limited number of outliers that might confound classification. The US Geological Survey (USGS) is exploring methods to automate generalization of features in the National Hydrography Data set (NHD), to associate specific sequences of processing operations and parameters with specific landscape characteristics, thus obviating manual selection of a unique processing strategy for every NHD watershed unit. A chronology of methods to delineate physiographic regions for the United States is described, including a recent maximum likelihood classification based on seven input variables. This research compares unsupervised and supervised algorithms applied to these seven input variables, to evaluate and possibly refine the recent classification. Evaluation metrics for unsupervised methods include the Davies–Bouldin index, the Silhouette index, and the Dunn index as well as quantization and topographic error metrics. Cross validation and misclassification rate analysis are used to evaluate supervised classification methods. The paper reports the comparative analysis and its impact on the selection of landscape regions. The compared solutions show problems in areas of high landscape diversity. There is some indication that additional input variables, additional classes, or more sophisticated methods can refine the existing classification.

Cartography and Geographic Information Science↗

USGS2015

A rapid approach for automated comparison of independently derived stream networks

This paper presents an improved coefficient of line correspondence (CLC) metric for automatically assessing the similarity of two different sets of linear features. Elevation-derived channels at 1:24,000 scale (24K) are generated from a weighted flow-accumulation model and compared to 24K National Hydrography Dataset (NHD) flowlines. The CLC process conflates two vector datasets through a raster line-density differencing approach that is faster and more reliable than earlier methods. Methods are tested on 30 subbasins distributed across different terrain and climate conditions of the conterminous United States. CLC values for the 30 subbasins indicate 44–83% of the features match between the two datasets, with the majority of the mismatching features comprised of first-order features. Relatively lower CLC values result from subbasins with less than about 1.5 degrees of slope. The primary difference between the two datasets may be explained by different data capture criteria. First-order, headwater tributaries derived from the flow-accumulation model are captured more comprehensively through drainage area and terrain conditions, whereas capture of headwater features in the NHD is cartographically constrained by tributary length. The addition of missing headwaters to the NHD, as guided by the elevation-derived channels, can substantially improve the scientific value of the NHD.

Cartography and Geographic Information Science↗

USGS2015

Automated extraction of natural drainage density patterns for the conterminous United States through high performance computing

Hydrographic networks form an important data foundation for cartographic base mapping and for hydrologic analysis. Drainage density patterns for these networks can be derived to characterize local landscape, bedrock and climate conditions, and further inform hydrologic and geomorphological analysis by indicating areas where too few headwater channels have been extracted. But natural drainage density patterns are not consistently available in existing hydrographic data for the United States because compilation and capture criteria historically varied, along with climate, during the period of data collection over the various terrain types throughout the country. This paper demonstrates an automated workflow that is being tested in a high-performance computing environment by the U.S. Geological Survey (USGS) to map natural drainage density patterns at the 1:24,000-scale (24K) for the conterminous United States. Hydrographic network drainage patterns may be extracted from elevation data to guide corrections for existing hydrographic network data. The paper describes three stages in this workflow including data pre-processing, natural channel extraction, and generation of drainage density patterns from extracted channels. The workflow is concurrently implemented by executing procedures on multiple subbasin watersheds within the U.S. National Hydrography Dataset (NHD). Pre-processing defines parameters that are needed for the extraction process. Extraction proceeds in standard fashion: filling sinks, developing flow direction and weighted flow accumulation rasters. Drainage channels with assigned Strahler stream order are extracted within a subbasin and simplified. Drainage density patterns are then estimated with 100-meter resolution and subsequently smoothed with a low-pass filter. The extraction process is found to be of better quality in higher slope terrains. Concurrent processing through the high performance computing environment is shown to facilitate and refine the choice of drainage density extraction parameters and more readily improve extraction procedures than conventional processing.

Conference Paper↗

USGS2015

Synoptic evaluation of scale-dependent metrics for hydrographic line feature geometry

Methods of acquisition and feature simplification for vector feature data impact cartographic representations and scientific investigations of these data, and are therefore important considerations for geographic information science (Haunert and Sester 2008). After initial collection, linear features may be simplified to reduce excessive detail or to furnish a reduced-scale version of the features through cartographic generalization (Regnauld and McMaster 2008, Stanislawski et al. 2014). A variety of algorithms exist to simplify linear cartographic features, and all of the methods affect the positional accuracy of the features (Shahriari and Tao 2002, Regnauld and McMaster 2008, Stanislawski et al. 2012). In general, simplification operations are controlled by one or more tolerance parameters that limit the amount of positional change the operation can make to features. Using a single tolerance value can have varying levels of positional change on features; depending on local shape, texture, or geometric characteristics of the original features (McMaster and Shea 1992, Shahriari and Tao 2002, Buttenfield et al. 2010). Consequently, numerous researchers have advocated calibration of simplification parameters to control quantifiable properties of resulting changes to the features (Li and Openshaw 1990, Raposo 2013, Tobler 1988, Veregin 2000, and Buttenfield, 1986, 1989). This research identifies relations between local topographic conditions and geometric characteristics of linear features that are available in the National Hydrography Dataset (NHD). The NHD is a comprehensive vector dataset of surface 18 th ICA Workshop on Generalisation and Multiple Representation, Rio de Janiero, Brazil 2015 2 water features within the United States that is maintained by the U.S. Geological Survey (USGS). In this paper, geometric characteristics of cartographic representations for natural stream and river features are summarized for subbasin watersheds within entire regions of the conterminous United States and compared to topographic metrics. A concurrent processing workflow is implemented using a Linux high-performance computing cluster to simultaneously process multiple subbasins, and thereby complete the work in a fraction of the time required for a single-process environment. In addition, similar metrics are generated for several levels of simplification of the hydrographic features to quantify the effects of simplification over the various landscape conditions. Objectives of this exploratory investigation are to quantify geometric characteristics of linear hydrographic features over the various terrain conditions within the conterminous United States and thereby illuminate relations between stream geomorphological conditions and cartographic representation. The synoptic view of these characteristics over regional watersheds that is afforded through concurrent processing, in conjunction with terrain conditions, may reveal patterns for classifying cartographic stream features into stream geomorphological classes. Furthermore, the synoptic measurement of the amount of change in geometric characteristics caused by the several levels of simplification can enable estimation of tolerance values that appropriately control simplification-induced geometric change of the cartographic features within the various geomorphological classes in the country. Hence, these empirically derived rules or relations could help generate multiscale-representations of features through automated generalization that adequately maintain surface drainage variations and patterns reflective of the natural stream geomorphological conditions across the country.

Conference Paper↗

USGS2014

Generalisation operators

This chapter summarises cartographic generalisation operators used to generalise geospatial data. It includes a review of recent approaches that have been tested or implemented to generalise networks, points, or groups. Emphasis is placed on recent advances that permit additional flexibility to tailor generalisation processing in particular geographic contexts, and to permit more advanced types of reasoning about spatial conflicts, preservation of specific feature characteristics, and local variations in geometry, content and enriched attribution. Rather than an exhaustive review of generalisation operators, the chapter devotes more attention to operators associated with network generalisation, which illustrates well the logic behind map generalisation developments. Three case studies demonstrate the application of operators to road thinning, to river network and braid pruning, and to hierarchical point elimination. The chapter closes with some summary comments and future directions.

Book chapter↗