Search USGS⌕ Search

SEARCH · Search USGS

Results for “Statistical Methodology”

Search indexed USGS publications on groundwater, aquifers, geologic maps, mineral resources and earthquakes. Explore source records by subject and place.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 127 records · Page 7Linked to original sources
USGS1996

Origins and philosophy of building a temporal database to examine human transformation processes

The U.S. Geological Survey (USGS), in partnership with the University of Maryland Baltimore County, is using historical maps and satellite images to map human-induced land transformations for the Baltimore-Washington metropolitan area. This work builds on an earlier effort that documented the historical urban development for the San Francisco Bay area. That effort used a geographic information system to compile a database that provided a visual and historical perspective of the urban growth experienced in the Bay area between 1850 and 1990. Historical overviews of urban development can be used to provide insights into the future. The Bay area work was inspired by the desire to draw upon the USGS's rich 100-year topographic map, and 20-year Landsat satellite data archives. A methodology was developed to combine the information from a variety of sources into an integrated, multi-scale, and multi-resolution dataset. Temporal urban mapping is used to examine past landscapes by incorporating historic maps, census statistics, and commerce records to generate a progressive geo-referenced representation of the past changes in the region. Contemporary analysis focuses on the use of remotely sensed data, existing digital land use data, digital census information, and a variety of Earth science infrastructure data, such as Digital Line Graphs, Digital Elevation Models, and key ancillary demographic information. The resulting database of temporal urban demographic changes provides an ideal source of test data and information for both urban geographers and global change research scientists.

San Francisco Bay↗
USGS2019

Sampling and analysis frameworks for inference in ecology

1. Reliable statistical inference is central to ecological research, much of which seeks to estimate population attributes and their interactions. The issue of sampling design and its relationship to inference has become increasingly important due to rapid proliferation of modeling methodology (line transect modeling, capture-recapture, estimation of occurrence, model selection procedures, hierarchical modeling) and new sampling approaches (adaptive sampling, other specialized designs). It is important for ecologists using these advanced methods to be aware of how the linkages between sample selection and data analysis can potentially affect inference. 2. We examine design-based and model-based inference frameworks for ecological data collected randomly, purposively, or opportunistically. We elucidate differences in the probability structures for data arising from these frameworks, clarify the assumptions that underlie them, and demonstrate their differences. 3. Design-based inference builds on a probability structure inherited from randomized data collection, whereas model-based inference relies on an assumed stochastic model of the data. By itself, a design-based approach is of limited value for inferences about causal hypotheses. In contrast, model-based inference is dependent on a conditionality principle that can seldom be shown to be met for an ecological system. We describe the conditions under which one can safely ignore sampling design in model-based analysis, along with inferential implications if these conditions are not met. The special case of opportunistic sampling is discussed. 4. We present a combined framework that takes advantage of both approaches to inference, and provides a robust methodology that can deal with the modeling of sampling problems such as nondetection and misclassification, as well as the exploration of causal hypotheses. The combined framework can be useful for identifying optimal sampling strategies. 5. Each approach to inference has its strengths and weaknesses, and practitioners should be aware of these in order to tailor designs and analyses to specific questions. We use the approaches and their underlying rationales to provide guidelines for choosing designs and estimators for reliable inference.

Methods in Ecology and Evolution↗
USGS1986

Vegetation mapping of Nowitna National Wildlife Refuge, Alaska using Landsat MSS digital data

A Landsat-derived vegetation map was prepared for Nowitna National Wildlife Refuge. The refuge lies within the middle boreal subzone of north central Alaska. Seven major vegetation classes and sixteen subclasses were recognized: forest (closed needleleaf, open needleleaf, needleleaf woodland, mixed, and broadleaf); broadleaf scrub (lowland, alluvial, subalpine); dwarf scrub (prostrate dwarf shrub tundra, dwarf shrub-graminoid tussock peatland); herbaceous (graminoid bog, marsh and meadow); scarcely vegetated areas (scarcely vegetated scree and floodplain); water (clear, turbid); and other areas (mountain shadow). The methodology employed a cluster-block technique. Sample areas were described based on a combination of helicopter-ground survey, aerial photointerpretation, and digital Landsat data. Major steps in the Landsat analysis involved preprocessing (geometric correction), derivation of statistical parameters for spectral classes, spectral class labeling of sample areas, preliminary classification of the entire study area using a maximum-likelihood algorithm, and final classification utilizing ancillary information such as digital elevation data. The final product is a 1:250,000-scale vegetation map representative of distinctive regional patterns and suitable for use in comprehensive conservation planning.

Alaska↗
USGS2015

Social Values for Ecosystem Services, version 3.0 (SolVES 3.0): documentation and user manual

The geographic information system (GIS) tool, S ocial V alues for E cosystem S ervices (SolVES), was developed to incorporate quantified and spatially explicit measures of social values into ecosystem service assessments. SolVES 3.0 continues to extend the functionality of SolVES, which was designed to assess, map, and quantify the social values of ecosystem services. Social values—the perceived, nonmarket values the public ascribes to ecosystem services, particularly cultural services, such as aesthetics and recreation—can be evaluated for various stakeholder groups. These groups are distinguishable by their attitudes and preferences regarding public uses, such as motorized recreation and logging. As with previous versions, SolVES 3.0 derives a quantitative 10-point, social-values metric—the value index—from a combination of spatial and nonspatial responses to public value and preference surveys. The tool also calculates metrics characterizing the underlying environment, such as average distance to water and dominant landcover. SolVES 3.0 is integrated with Maxent maximum entropy modeling software to generate more complete social-value maps and offer robust statistical models describing the relationship between the value index and explanatory environmental variables. A model’s goodness of fit to a primary study area and its potential performance in transferring social values to similar areas using value-transfer methodology can be evaluated. SolVES 3.0 provides an improved public-domain tool for decision makers and researchers to evaluate the social values of ecosystem services and to facilitate discussions among diverse stakeholders regarding the tradeoffs among ecosystem services in a variety of physical and social contexts ranging from forest and rangeland to coastal and marine.

Open-File Report↗
USGS2014

Using spatial resampling to assess redd count survey length requirements for Pacific Lamprey

Pacific Lamprey Entosphenus tridentatus has declined across its range along the West Coast of North America, and an understanding of all life history phases is needed to address population recovery. Spawning surveys (redd counts) are common tools currently used to monitor returning adult salmonids, but such methodologies are in their infancy for Pacific Lamprey. Our objective was to assess the minimum spawning survey distance required to detect the presence of Pacific Lamprey redds and obtain precise redd density estimates from these data. To do this, we statistically resampled existing spawning locations of Pacific Lamprey collected during spawning surveys in four streams of the Willamette River Basin, Oregon, during spring of 2013. We found that the minimum survey distance for Pacific Lamprey redd detection was inversely related to the observed redd density and was always less than 1.2 km. Survey distance requirements to obtain precise redd counts (±20% of observed redd densities) were also inversely related to redd density and habitat availability, and varied between 1.3 km and 13.7 km. Our results suggest that spawning surveys are a potential tool for monitoring adult Pacific Lamprey abundance, but the specific objectives of the monitoring programs and acknowledgment of unknowns must be considered prior to implementation into recovery plans.

Oregon↗
USGS2015

Structural equation modeling: Building and evaluating causal models

Scientists frequently wish to study hypotheses about causal relationships, rather than just statistical associations. This chapter addresses the question of how scientists might approach this ambitious task. Here we describe structural equation modeling (SEM), a general modeling framework for the study of causal hypotheses. Our goals are to (a) concisely describe the methodology, (b) illustrate its utility for investigating ecological systems, and (c) provide guidance for its application. Throughout our presentation, we rely on a study of the effects of human activities on wetland ecosystems to make our description of methodology more tangible. We begin by presenting the fundamental principles of SEM, including both its distinguishing characteristics and the requirements for modeling hypotheses about causal networks. We then illustrate SEM procedures and offer guidelines for conducting SEM analyses. Our focus in this presentation is on basic modeling objectives and core techniques. Pointers to additional modeling options are also given.

Book chapter↗
USGS1988

Intermediate-scale vegetation mapping of Innoko National Wildlife Refuge, Alaska using Landsat MSS digital data

A Landsat-derived vegetation map was prepared for lnnoko National Wildlife Refuge. The refuge lies within the northern boreal subzone of northwestern central Alaska. Six major vegetation classes and 21 subclasses were recognized: forest (closed needleleaf, open needleleaf, needleleaf woodland, mixed, and broadleaf); broadleaf scrub (lowland, upland burn regeneration, subalpine); dwarf scrub (prostrate dwarf shrub tundra, erect dwarf shrub heath, dwarf shrub-graminoid peatland, dwarf shrub-graminoid tussock peatland, dwarf shrub raised bog with scattered trees, dwarf shrub-graminoid marsh); herbaceous (graminoid bog, graminoid marsh, graminoid tussock-dwarf shrub peatland); scarcely vegetated areas (scarcely vegetated scree and floodplain); and water (clear, sedimented). The methodology employed a cluster-block technique. Sample areas were described based on a combination of helicopter-ground survey, aerial photo-interpretation, and digital Landsat data. Major steps in the Landsat analysis involved preprocessing (geometric correction), derivation of statistical parameters for spectral classes, spectral class labeling of sample areas, preliminary classification of the entire study area using a maximum-likelihood algorithm, and final classification utilizing ancillary information such as digital elevation data. The final product is 1:250,000-scale vegetation map representative of distinctive regional patterns and suitable for use in comprehensive conservation planning.

Alaska↗
USGS2012

Guidelines for a graph-theoretic implementation of structural equation modeling

Structural equation modeling (SEM) is increasingly being chosen by researchers as a framework for gaining scientific insights from the quantitative analyses of data. New ideas and methods emerging from the study of causality, influences from the field of graphical modeling, and advances in statistics are expanding the rigor, capability, and even purpose of SEM. Guidelines for implementing the expanded capabilities of SEM are currently lacking. In this paper we describe new developments in SEM that we believe constitute a third-generation of the methodology. Most characteristic of this new approach is the generalization of the structural equation model as a causal graph. In this generalization, analyses are based on graph theoretic principles rather than analyses of matrices. Also, new devices such as metamodels and causal diagrams, as well as an increased emphasis on queries and probabilistic reasoning, are now included. Estimation under a graph theory framework permits the use of Bayesian or likelihood methods. The guidelines presented start from a declaration of the goals of the analysis. We then discuss how theory frames the modeling process, requirements for causal interpretation, model specification choices, selection of estimation method, model evaluation options, and use of queries, both to summarize retrospective results and for prospective analyses. The illustrative example presented involves monitoring data from wetlands on Mount Desert Island, home of Acadia National Park. Our presentation walks through the decision process involved in developing and evaluating models, as well as drawing inferences from the resulting prediction equations. In addition to evaluating hypotheses about the connections between human activities and biotic responses, we illustrate how the structural equation (SE) model can be queried to understand how interventions might take advantage of an environmental threshold to limit Typha invasions. The guidelines presented provide for an updated definition of the SEM process that subsumes the historical matrix approach under a graph-theory implementation. The implementation is also designed to permit complex specifications and to be compatible with various estimation methods. Finally, they are meant to foster the use of probabilistic reasoning in both retrospective and prospective considerations of the quantitative implications of the results.

Ecosphere↗
USGS2020

Are the stress drops of small earthquakes good predictors of the stress drops of moderate-to-large earthquakes?

The stress drops of small earthquakes often exhibit spatial patterns of variability. If moderate and large earthquakes follow the same spatial patterns, the stress drops of possible future damaging earthquakes could be better predicted by considering the stress drops of nearby small events. Better stress drop predictability could reduce ground-motion uncertainty in Probabilistic Seismic Hazard Assessment (PSHA) and Earthquake Early Warning (EEW). I find that for an internally consistent stress drop catalog of M1.8-3.1 events in southern California, the stress drops of the bigger earthquakes are predictable from the nearby smaller events. However, this catalog only weakly spatially correlates with another catalog of M3.0-5.8 earthquakes, and is spatially uncorrelated with five other stress drop catalogs of M≥3.4 earthquakes. For southern California events M5.5-7.5, stress drops compiled from the literature are weakly spatially correlated with the stress drops of the M1.8-3.1 events, although the correlations are not statistically significant. The lack of strong spatial correlation may be due to actual differences in the controlling factors of stress drop, for example dynamic weakening in moderate-to-large earthquakes. Alternatively, a stronger spatial correlation may exist that is obscured by methodological heterogeneity and large errors in the stress drop estimates. Either way, the stress drops of small earthquakes do not appear to be good predictors of the stress drops of nearby moderate-to-large earthquakes, at least for current techniques of stress drop estimation. If these results are representative, small-earthquake stress drops are not currently useful for substantially reducing uncertainty in PSHA and EEW.

California↗
USGS2011

2-D inner-shelf current observations from a single VHF WEllen RAdar (WERA) station

The majority of High Frequency (HF) radars used worldwide operate at medium to high frequencies (8 to 30 MHz) providing spatial resolutions ranging from 3 to 1.5 km and ranges from 150 to 50 km. This paper presents results from the deployment of a single Very High Frequency (VHF, 48 MHz) WEllen RAdar (WERA) radar with spatial resolution of 150 m and range 10-15 km, used in the nearshore off Cape Hatteras, NC, USA. It consisted of a linear array of 12 antennas operating in beam forming mode. Radial velocities were estimated from radar backscatter for a variety of wind and nearshore wave conditions. A methodology similar to that used for converting acoustically derived beam velocities to an orthogonal system is presented for obtaining 2-D current fields from a single station. The accuracy of the VHF radar-derived radial velocities is examined using a new statistical technique that evaluates the system over the range of measured velocities. The VHF radar velocities showed a bias of 3 to 7 cm/s over the experimental period explainable by the differences in radar penetration and in-situ measurement height. The 2-D current field shows good agreement with the in-situ measurements. Deviations and inaccuracies are well explained by the geometric dilution analysis.

North Carolina↗
USGS2018

Implementation of MOVE.1, censored MOVE.1, and piecewise MOVE.1 low-flow regressions with applications at partial-record streamgaging stations in New Jersey

The U.S. Geological Survey (USGS) uses Maintenance of Variance Extension Type 1 (MOVE.1) regression to transfer streamflows measured at long-term continuous-record streamgaging stations to partial-record (PR) streamgaging stations where intermittent base-flow measurements are available. MOVE.1 regression is used widely throughout the hydrologic community to extend historic low flows and low-flow statistics at continuous-record streamgaging stations to streamgaging stations that have access to only a partial record of low flows. The method correlates base-flow measurements at PR streamgaging stations with daily mean streamflows measured at index stations that exhibit similar streamflow characteristics. Following changes in the computing platform for storing, processing, retrieving, and publishing National Water Information System (NWIS) hydrologic data, legacy Statistical Analysis System (SAS) code developed by the USGS to implement the MOVE.1 regression was no longer suitable for reading and processing NWIS streamflow data. To migrate the MOVE.1 program so that it could continue to read streamflow data using the new hydrologic data platform, the SAS code was re-written in R, an open source programming language and software environment for statistical computing and graphics supported by the R Foundation for Statistical Computing. The work described in this report was performed in a study conducted by USGS in cooperation with the New Jersey Department of Environmental Protection. During migration from SAS to R, graphical and tabular output generated by the R script was compared to output produced by the legacy SAS code to ensure that equations used to perform the MOVE.1 regression remained the same. An option to perform censored MOVE.1 regression was added to extend the MOVE.1 methodology to cases where one or more measured continuous-record or PR streamgaging station flows are zero valued. In addition to permitting censored regression, the new R script includes an option to perform piecewise MOVE.1 regression when the relation between PR station and index station low flows varies significantly across the range of index station streamflows. Together with traditional MOVE.1 regression, censored, and piecewise MOVE.1 regression methods implemented by the R script offer less biased estimates than ordinary least squares regression for the annual 7-day 10-year and other low-flow statistics at PR stations for a range of base-flow conditions. The R script is used to implement the MOVE.1 regression methods across a variety of computing platforms.

New Jersey↗
USGS2016

Testing for variation in taxonomic extinction probabilities: A suggested methodology and some results

Several important questions in evolutionary biology and paleobiology involve sources of variation in extinction rates. In all cases of which we are aware, extinction rates have been estimated from data in which the probability that an observation (e.g., a fossil taxon) will occur is related both to extinction rates and to what we term encounter probabilities. Any statistical method for analyzing fossil data should at a minimum permit separate inferences on these two components. We develop a method for estimating taxonomic extinction rates from stratigraphic range data and for testing hypotheses about variability in these rates. We use this method to estimate extinction rates and to test the hypothesis of constant extinction rates for several sets of stratigraphic range data. The results of our tests support the hypothesis that extinction rates varied over the geologic time periods examined. We also present a test that can be used to identify periods of high or low extinction probabilities and provide an example using Phanerozoic invertebrate data. Extinction rates should be analyzed using stochastic models, in which it is recognized that stratigraphic samples are random variates and that sampling is imperfect.

Paleobiology↗
USGS1981

Broad area forest fuels and topography mapping using digital Landsat and terrain data

A spatially registered digital data base of fuels and terrain information was generated for a test site on the Lolo National Forest in Montana. The data base was developed specifically for providing spatially relevant data to a mathematical fire behavior model developed by the Forest Service which integrates this information along with current weather data to produce realistic estimates of probable fire behavior. Methodologies for the processing and analysis of Landsat MSS and digital terrain data for the mapping of U.S. Forest Service fuel types were developed and demonstrated. Key elements in the mapping process were the development of a fuels terrain distribution model which provided a statistical description of the topographic distribution patterns of fuels within spectral classes, and secondly, the application of a layered classifier which incorporated the spectral and terrain data in a two-stage maximum likelihood classification framework for the mapping of fuels.

Montana↗
USGS2019

Predicting species distributions: unifying model selection and scale optimization for multi-scale occupancy models

Geographic distributions are a basic component of a species’ ecology, and predicting distributions is a fundamental task of conservation and resource management. Reliable prediction depends on identification of appropriate scales of effect for environmental data, and scale-optimization techniques are thus desirable to identify optimal scales for predictor variables. Recent statistical developments have also advanced methods of model selection based explicitly on predictive ability, which differ from commonly-used methods that regulate model structures via anticipated predictive performance. Such methods are beginning to permeate into species distribution models (SDMs), yet there remains no consensus methodology for developing optimally-predictive multi-scale SDMs when covariate data are collected over a range of scales. Thus, we compared the performance of common approaches for scale optimization and model selection in terms of their ability to produce optimally predictive multi-scale Bayesian occupancy models for predicting a species distribution, using models of the breeding distribution for King Rails (Rallus elegans) as a case study. Our results demonstrate sizable gains in predictive performance for hierarchical occupancy models selected explicitly via their ability to predict out-of-sample data using the logarithmic scoring rule, as compared to models selected using information criteria (DIC and WAIC). Information criteria commonly selected individual covariates, as well as scales of effect for those covariates, with suboptimal predictive performance. Performance of models selected using the logarithmic scoring rule was robust across method of scale optimization, which was not true for models selected using DIC and WAIC. Thus, we empirically demonstrate benefits of study designs that enable covariate and scale selection based explicitly on predictive ability. Our results also imply that more careful consideration of what constitutes an optimal scale is warranted in many ecological studies, as the meaning of optimal is not independent of the technique used for scale selection.

Ecosphere↗
USGS1985

Landsat-faciliated vegetation classification of the Kenai National Wildlife Refuge and adjacent areas, Alaska

A Landsat-based vegetation map was prepared for Kenai National Wildlife Refuge and adjacent lands, 2 million and 2.5 million acres respectively. The refuge lies within the middle boreal sub zone of south central Alaska. Seven major classes and sixteen subclasses were recognized: forest (closed needleleaf, needleleaf woodland, mixed); deciduous scrub (lowland and montane, subalpine); dwarf scrub (dwarf shrub tundra, lichen tundra, dwarf shrub and lichen tundra, dwarf shrub peatland, string bog/wetlands); herbaceous (graminoid meadows and marshes); scarcely vegetated areas ; water (clear, moderately turbid, highly turbid); and glaciers. The methodology employed a cluster-block technique. Sample areas were described based on a combination of helicopter-ground survey, aerial photo interpretation, and digital Landsat data. Major steps in the Landsat analysis involved: preprocessing (geometric connection), spectral class labeling of sample areas, derivation of statistical parameters for spectral classes, preliminary classification of the entree study area using a maximum-likelihood algorithm, and final classification through ancillary information such as digital elevation data. The vegetation map (scale 1:250,000) was a pioneering effort since there were no intermediate-sclae maps of the area. Representative of distinctive regional patterns, the map was suitable for use in comprehensive conservation planning and wildlife management.

Alaska↗
USGS2015

Standardization of reflectance measurements in dispersed organic matter: results of an exercise to improve interlaboratory agreement

Vitrinite reflectance generally is considered the most robust thermal maturity parameter available for application to hydrocarbon exploration and petroleum system evaluation. However, until 2011 there was no standardized methodology available to provide guidelines for vitrinite reflectance measurements in shale. Efforts to correct this deficiency resulted in publication of ASTM D7708: Standard test method for microscopical determination of the reflectance of vitrinite dispersed in sedimentary rocks . In 2012-2013, an interlaboratory exercise was conducted to establish precision limits for the D7708 measurement technique. Six samples, representing a wide variety of shale, were tested in duplicate by 28 analysts in 22 laboratories from 14 countries. Samples ranged from immature to overmature (0.31-1.53% R o ), from organic-lean to organic-rich (1-22 wt.% total organic carbon), and contained Type I (lacustrine), Type II (marine), and Type III (terrestrial) kerogens. Repeatability limits (maximum difference between valid repetitive results from same operator, same conditions) ranged from 0.03-0.11% absolute reflectance, whereas reproducibility limits (maximum difference between valid results obtained on same test material by different operators, different laboratories) ranged from 0.12-0.54% absolute reflectance. Repeatability and reproducibility limits degraded consistently with increasing maturity and decreasing organic content. However, samples with terrestrial kerogens (Type III) fell off this trend, showing improved levels of reproducibility due to higher vitrinite content and improved ease of identification. Operators did not consistently meet the reporting requirements of the test method, indicating that a common reporting template is required to improve data quality. The most difficult problem encountered was the petrographic distinction of solid bitumens and low-reflecting inert macerals from vitrinite when vitrinite occurred with reflectance ranges overlapping the other components. Discussion among participants suggested this problem could not be easily corrected via kerogen concentration or solvent extraction and is related to operator training and background. No statistical difference in mean reflectance was identified between participants reporting bitumen reflectance vs. vitrinite reflectance vs. a mixture of bitumen and vitrinite reflectance values, suggesting empirical conversion schemes should be treated with caution. Analysis of reproducibility limits obtained during this exercise in comparison to reproducibility limits from historical interlaboratory exercises suggests use of a common methodology (D7708) improves interlaboratory precision. Future work will investigate opportunities to improve reproducibility in high maturity, organic-lean shale varieties.

Marine and Petroleum Geology↗
USGS2011

Effects of sample size, number of markers, and allelic richness on the detection of spatial genetic pattern

The influence of study design on the ability to detect the effects of landscape pattern on gene flow is one of the most pressing methodological gaps in landscape genetic research. To investigate the effect of study design on landscape genetics inference, we used a spatially-explicit, individual-based program to simulate gene flow in a spatially continuous population inhabiting a landscape with gradual spatial changes in resistance to movement. We simulated a wide range of combinations of number of loci, number of alleles per locus and number of individuals sampled from the population. We assessed how these three aspects of study design influenced the statistical power to successfully identify the generating process among competing hypotheses of isolation-by-distance, isolation-by-barrier, and isolation-by-landscape resistance using a causal modelling approach with partial Mantel tests. We modelled the statistical power to identify the generating process as a response surface for equilibrium and non-equilibrium conditions after introduction of isolation-by-landscape resistance. All three variables (loci, alleles and sampled individuals) affect the power of causal modelling, but to different degrees. Stronger partial Mantel r correlations between landscape distances and genetic distances were found when more loci were used and when loci were more variable, which makes comparisons of effect size between studies difficult. Number of individuals did not affect the accuracy through mean equilibrium partial Mantel r , but larger samples decreased the uncertainty (increasing the precision) of equilibrium partial Mantel r estimates. We conclude that amplifying more (and more variable) loci is likely to increase the power of landscape genetic inferences more than increasing number of individuals.

Molecular Ecology Resources↗
USGS2026

ENSO and PDO drive shoreline position anomalies in the U.S. Pacific Northwest

Sandy beaches act as buffers against various coastal hazards but are vulnerable to episodic (seasonal) and chronic (interannual) erosion. Understanding the variation of shoreline position, a key metric in coastal morphology, over a spectrum of time scales is therefore crucial in assessing hazard vulnerability. Long-standing research has investigated the role of El Niño-Southern Oscillation (ENSO), the dominant mode of climate variability in the Pacific Basin, in seasonal shoreline variability. Yet, ENSO’s chronic influence—and that of another Pacific climate mode, the Pacific Decadal Oscillation (PDO)—on shoreline anomalies remains poorly understood. Here, we examine the variability of sandy beaches in the US Pacific Northwest, a ∼750 km long coastal region on the US West Coast. We leverage 40 years (1984–2024) of shoreline data from publicly available Earth-observing (Landsat) satellite imagery at a high spatial resolution (>10,000 shore-normal transects at 50-m alongshore spacing) and employ Convergent Cross Mapping (CCM), a methodology for inferring causality in dynamical systems. We discover that strong El Niño years are signified by erosion (75.1% of transects), and strong La Niña years exhibit accretional behavior (73.4% of transects). Furthermore, we establish, for the first time, that both ENSO and PDO exert a statistically significant control on interannual shoreline variability, particularly on the alongshore component (in 95 and 100% of littoral cells, respectively), with water level fluctuations playing a critical role. This effort advances our understanding of the seasonal-to-interannual interactions between Pacific Basin climate variability and the PNW’s coastal morphodynamics, with implications for sediment management and coastal adaptation.

California, Oregon, Washington↗