USGS Science⌕ Search

SEARCH · USGS Science

Results for “Input”

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 433 records · Page 24Linked to original sources
USGS1999

Chlorinated hydrocarbon pesticides and polychlorinated biphenyls in sediment cores from San Francisco Bay

Sediment cores of known chronology from Richardson and San Pablo Bays in San Francisco Bay, CA, were analyzed for a suite of chlorinated hydrocarbon pesticides and polychlorinated biphenyls to reconstruct a historic record of inputs. Total DDTs (DDT = 2,4'- and 4,4'-dichlorodiphenyltrichloroethane and the metabolites, 2,4'- and 4,4'-DDE, -DDD) range in concentration from 4-21 ng/g and constitute a major fraction (> 84%) of the total pesticides in the top 70 cm of Richardson Bay sediment. A subsurface maximum corresponds to a peak deposition date of 1969-1974. The first measurable DDT levels are found in sediment deposited in the late 1930's. The higher DDT inventory in the San Pablo relative to the Richardson Bay core probably reflects the greater proximity of San Pablo Bay to agricultural activities in the watershed of the Sacramento and San Joaquin rivers. Total polychlorinated biphenyls (PCBs) occur at comparable levels in the two Bays (< 1-34 ng/g). PCBs are first detected in sediment deposited during the 1930's in Richardson Bay, about a decade earlier than the onset of detectable levels of DDTs. PCB inventories in San Pablo Bay are about a factor of four higher in the last four decades than in Richardson Bay, suggesting a distribution of inputs not as strongly weighed towards the upper reaches of the estuary as DDTs. The shallower subsurface maximum in PCBs compared to DDT in the San Pablo Bay core is consistent with the imposition of drastic source control measures four these constituents in 1970 and 1977 respectively. The observed decline in DDT and PCB levels towards the surface of both cores is consistent with a dramatic drop in the input of these pollutants once the effect of sediment resuspension and mixing is taken into account.

California↗
USGS2011

Relationships between silicic plutonism and volcanism: Geochemical evidence

Field associations (voluminous ash flow deposits, rhyolitic stocks and dykes, ring complexes), evidence of repeated influxes of mafic magma, and thermal constraints indicate that many high-level silicic plutons (magma chambers) acted as open systems for considerable parts of their history. The long thermal lifetime, as well as other evidence from the volcanic record, suggests that some such systems reached a quasi-steady state in which magma input was balanced by magma output for times longer than those required for crystallisation. Reconstruction of the evolution of large, long-lived caldera-forming systems, such as that of the Jemez Mountains, New Mexico, indicates that many chambers have lost a highly fractionated silicic cap, in some cases cyclically. Crystallised plutons may contain no obvious record of this evolutionary phase. Geochemical data from silicic ash flow deposits can be used to reconstruct the volcanic stage of pluton development. Many silicic systems, especially of alkaline affinity, apparently pass from a stage in which melt evolution is dominated by crystal-liquid processes to one in which other processes may also contribute to differentiation. Apparently, the transition is most readily achieved in volatile-rich, alkaline silicic systems emplaced in complex, ancient sialic crust of the cratons. Once established, the preservation of highly fractionated caps on magma chambers requires a balance between thermal input and cooling-induced crystallisation. If heat enters the system too quickly, the cap may get stirred into the dominant magma volume by convection. If heat input is too slow, the magma body will crystallise inward from the margins, and the plutonic-consolidation stage will begin. © 1988, Royal Society of Edinburgh. All rights reserved.

Transactions of the Royal Society of Edinburgh, Ea↗
USGS2007

Persistent chlordane concentrations in long island sound sediment: Implications from chlordane, 210 Pb, and 137 Cs profiles

Concentrations of chlordane, a banned termiticide and pesticide, were examined in recently collected surficial sediment (10 sites) and sediment cores (4 sites) in Long Island Sound (LIS).The highest chlordane concentrations were observed in western LIS, near highly urbanized areas. Chlordane concentrations did not decrease significantly in the past decade when compared to the data collected in 1996, consistent with the observation of near-constant chlordane levels in blue mussel tissues collected during the same time period. Chlordane concentrations in many of the sites exceeded levels above which harmful effects on sediment-dwelling organisms are expected to frequently occur. Chlordane concentrations in two of the four sediment cores showed a peak below the sediment surface, suggesting reduced chlordane inputs in recent years. The lack of a chlordane concentration maximum below the sediment surface in the other two cores, coupled with the lack of a well-defined 137 Cs peak, indicated significant sediment mixing. Simulations of 137 Cs and 210 Pb profiles in sediment cores with a simple sediment-mixing model were used to constrain both the deposition rate and the bioturbation rate of the sediment. Simulations of the chlordane profiles indicated continued chlordane input to LIS long after chlordane was phased out in the U.S. Continued chlordane input and significant sediment mixing may have contributed to the persistent chlordane concentrations in surficial sediment, which poses long-term threats to benthic organisms in LIS.

New York↗
USGS2000

Dissolved organic nitrogen budgets for upland, forested ecosystems in New England

Relatively high deposition of nitrogen (N) in the northeastern United States has caused concern because sites could become N saturated. In the past, mass-balance studies have been used to monitor the N status of sites and to investigate the impact of increased N deposition. Typically, these efforts have focused on dissolved inorganic forms of N (DIN = NH4-N + NO3-N) and have largely ignored dissolved organic nitrogen (DON) due to difficulties in its analysis. Recent advances in the measurement of total dissolved nitrogen (TDN) have facilitated measurement of DON as the residual of TDN - DIN. We calculated DON and DIN budgets using data on precipitation and streamwater chemistry collected from 9 forested watersheds at 4 sites in New England. TDN in precipitation was composed primarily of DIN. Net retention of TDN ranged from 62 to 89% (4.7 to 10 kg ha-1 yr-1) of annual inputs. DON made up the majority of TDN in stream exports, suggesting that inclusion of DON is critical to assessing N dynamics even in areas with large anthropogenic inputs of DIN. Despite the dominance of DON in streamwater, precipitation inputs of DON were approximately equal to outputs. DON concentrations in streamwater did not appear significantly influenced by seasonal biological controls, but did increase with discharge on some watersheds. Streamwater NO3-N was the only fraction of N that exhibited a seasonal pattern, with concentrations increasing during the winter months and peaking during snowmelt runoff. Concentrations of NO3-N varied considerably among watersheds and are related to DOC:DON ratios in streamwater. Annual DIN exports were negatively correlated with streamwater DOC:DON ratios, indicating that these ratios might be a useful index of N status of upland forests.

Biogeochemistry↗
USGS2002

Cycling of beryllium and carbon through hillslope soils in Iowa

Isotopes of Be and C were used to reconstruct loess accumulation, hillslope evolution, and agricultural modification in soils of western Iowa. While both elements are derived from additions by the atmosphere (via plants in the case of carbon), the differences in element cycling allow erosional and depositional processes to be separated from biochemical processing. Based on 10Be, loess accumulation likely occurred simultaneously with hillslope degradation. Rates of loess accumulation declined five-fold between early stages (late Pleistocene and early Holocene) and later stages (late Holocene) of accumulation, but the absolute timing of accumulation requires independent dating methods. Based on 14C measurements, plant inputs and decomposition are significant near the surface, but below 1-1.5 m carbon inputs are minimal and decomposition is nearly arrested. The amount of carbon below 1.5 m is constant (0.1%) and is composed of soil organic matter that was buried by loess. Agricultural modification results in a dramatic redistribution of 10Be through soil erosion and deposition. By contrast, the redistribution of soil organic matter is masked by the rapid cycling of C through the topsoil as it continually decomposes and is replaced by plant inputs.

Biogeochemistry↗
USGS2020

A spatially explicit, empirical estimate of tree-based biological nitrogen fixation in forests of the United States

Quantifying human impacts on the nitrogen (N) cycle and investigating natural ecosystem N cycling depend on the magnitude of inputs from natural biological nitrogen fixation (BNF). Here, we present two bottom‐up approaches to quantify tree‐based symbiotic BNF based on forest inventory data across the coterminous United States and SE Alaska. For all major N‐fixing tree genera, we quantify BNF inputs using (1) ecosystem N accretion rates (kg N ha −1 yr −1 ) scaled with spatial data on tree abundance and (2) percent of N derived from fixation (%N dfa ) scaled with tree N demand (from tree growth rates and stoichiometry). We estimate that trees fix 0.30–0.88 Tg N yr −1 across the study area (1.4–3.4 kg N ha −1 yr −1 ). Tree‐based N fixation displays distinct spatial variation that is dominated by two genera, Robinia (64% of tree‐associated BNF) and Alnus (24%). The third most important genus, Prosopis , accounted for 5%. Compared to published estimates of other N fluxes, tree‐associated BNF accounted for 0.59 Tg N yr −1 , similar to asymbiotic (0.37 Tg N yr −1 ) and understory symbiotic BNF (0.48 Tg N yr −1 ), while N deposition contributed 1.68 Tg N yr −1 and rock weathering 0.37 Tg N yr −1 . Overall, our results reveal previously uncharacterized spatial patterns in tree BNF that can inform large‐scale N assessments and serve as a model for improving tree‐based BNF estimates worldwide. This updated, lower BNF estimate indicates a greater ratio of anthropogenic to natural N inputs, suggesting an even greater human impact on the N cycle.

Global Biogeochemical Cycles↗
USGS2022

Integrating Bayesian networks to forecast sea-level rise impacts on barrier island characteristics and habitat availability

Evaluation of sea-level rise (SLR) impacts on coastal landforms and habitats is a persistent need for informing coastal planning and management, including policy decisions, particularly those that balance human interests and habitat protection throughout the coastal zone. Bayesian networks (BNs) are used to model barrier island change under different SLR scenarios that are relevant to management and policy decisions. BNs utilized here include a shoreline change model and two models of barrier island biogeomorphological evolution at different scales (50 and 5 m). These BNs were then linked to another BN to predict habitat availability for piping plovers ( Charadrius melodus ), a threatened shorebird reliant on beach habitats. We evaluated the performance of the two linked geomorphology BNs and further examined error rates by generating hindcasts of barrier island geomorphology and habitat availability for 2014 conditions. Geomorphology hindcasts revealed that model error declined with a greater number of known inputs, with error rates reaching 55% when multiple outputs were hindcast simultaneously. We also found that, although error in predictions of piping plover nest presence/absence increased when outputs from the geomorphology BNs were used as inputs in the piping plover habitat BN, the maximum error rate for piping plover habitat suitability in the fully-linked BNs was only 30%. Our findings suggest this approach may be useful for guiding scenario-based evaluations where known inputs can be used to constrain variables that produce higher uncertainty for morphological predictions. Overall, the approach demonstrates a way to assimilate data and model structures with uncertainty to produce forecasts to inform coastal planning and management.

New York↗
USGS2022

Evaluating the sensitivity of multi-dimensional model predictions of salmon habitat to the source of remotely sensed river bathymetry

Multi-dimensional numerical models are fundamental tools for investigating biophysical processes in aquatic ecosystems. Remote sensing techniques increase the feasibility of applying such models at riverscape scales, but tests of model performance on large rivers have been limited. We evaluated the potential to develop two-dimensional (2D) and three-dimensional (3D) hydrodynamic models for a 1.6-km reach of a large gravel-bed river using three sources of remotely sensed river bathymetry. We estimated depth from hyperspectral image data acquired from conventional and uncrewed aircraft and multispectral satellite imagery. Our results indicated that modeled water depth errors were similar between 2D and 3D models, with depth residuals that were comparable to the uncertainty associated with the bathymetry used as input. We found good agreement between measured and modeled depth-averaged velocities generated by 2D and 3D models, while 3D models provided superior predictions of near-bed velocities. We found that optimal model performance occurred for lower flow resistance values than previously reported in the literature, possibly as a consequence of the high-resolution bathymetry used as model input. Model predictions of winter-run Chinook salmon ( Oncorhynchus tshawytscha ) spawning and rearing habitat were not sensitive to the source of bathymetric information, but bioenergetic predictions related to adult holding costs were influenced by the input bathymetry. Our results suggest that hyperspectral imagery acquired from piloted and/or uncrewed aircraft can be used to map the bathymetry of clear-flowing, relatively shallow large rivers with sufficient accuracy to support multi-dimensional flow model development; models developed from multispectral satellite imagery had more limited predictive capability.

California↗
USGS2024

Longitudinal and seasonal changes of organic matter sources through a semi-arid river-reservoir system

The quality and quantity of organic matter (OM) in a river system directly affects ecosystem health; thus, managers benefit from an in-depth understanding of the drivers and sources of OM. In the Snake River, a highly altered river-reservoir system in the semi-arid western United States, OM production and loading are key drivers of reservoir anoxia, which leads to several deleterious processes such as mercury methylation. However, sources and quantities of OM to the Snake River, and the effects of impoundment on OM moving through the river-reservoir system, are not well understood. Particulate organic carbon (POC), dissolved organic carbon (DOC), particulate nitrogen (PN), chlorophyll a (chl- a ), and δ 15 N–PN and δ 13 C–POC isotopic ratios were measured bi-weekly for over 2 years at four locations through the Snake River Hells Canyon Reservoir Complex to determine spatial and temporal patterns of OM quantities and sources. POC concentrations increased through the riverine zone upstream of the reservoirs, likely due to in situ primary production and/or inputs from tributaries and agricultural drains; then decreased through the most upstream reservoir likely due to particle settling. Isotopic ratios and other OM source indicators (δ 15 N–PN, δ 13 C–POC, POC:PN, chl- a :POC) show that the dominant source of particulate OM was phytoplankton with seasonal terrestrial/macrophytic inputs. Results highlight the effects of major tributary and agricultural drain inputs, primary production, and impoundment on OM composition and concentration through a large river-reservoir system and may inform water quality management efforts in this and similar systems.

Idaho, Oregon↗
USGS1996

Thermal and mass implications of magmatic evolution in the Lassen volcanic region, California, and minimum constraints on basalt influx to the lower crust

We have analyzed the heat and mass demands of a petrologic model of basalt-driven magmatic evolution in which variously fractionated mafic magmas mix with silicic partial melts of the lower crust. We have formulated steady state heat budgets for two volcanically distinct areas in the Lassen region: the large, late Quaternary, intermediate to silicic Lassen volcanic center and the nearby, coeval, less evolved Caribou volcanic field. At Caribou volcanic field, heat provided by cooling and fractional crystallization of 52 km 3 of basalt is more than sufficient to produce 10 km of rhyolitic melt by partial melting of lower crust. Net heat added by basalt intrusion at Caribou volcanic field is equivalent to an increase in lower crustal heat flow of ∼7 mW m −2 , indicating that the field is not a major crustal thermal anomaly. Addition of cumulates from fractionation is offset by removal of erupted partial melts. A minimum basalt influx of 0.3 km 3 (km 2 Ma) −1 is needed to supply Caribou volcanic field. Our methodology does not fully account for an influx of basalt that remains in the crust as derivative intrusives. On the basis of comparison to deep heat flow, the input of basalt could be ∼3 to 7 times the amount we calculate. At Lassen volcanic center, at least 203 km 3 of mantle-derived basalt is needed to produce 141 km 3 of partial melt and drive the volcanic system. Partial melting mobilizes lower crustal material, augmenting the magmatic volume available for eruption at Lassen volcanic center; thus the erupted volume of 215 km 3 exceeds the calculated basalt input of 203 km 3 . The minimum basalt input of 1.6 km 3 (km 2 Ma) −1 is >5 times the minimum influx to the Caribou volcanic field. Basalt influx high enough to sustain considerable partial melting, coupled with locally high extension rate, is a crucial factor in development of Lassen volcanic center; in contrast, Caribou volcanic field has failed to develop into a large silicic center primarily because basalt supply there has been insufficient.

Journal of Geophysical Research B: Solid Earth↗
USGS2019

Prominence of the tropics in the recent rise of global nitrogen pollution

Nitrogen (N) pollution is shaped by multiple processes, the combined effects of which remain uncertain, particularly in the tropics. We use a global land biosphere model to analyze historical terrestrial-freshwater N budgets, considering the effects of anthropogenic N inputs, atmospheric CO2, land use, and climate. We estimate that globally, land currently sequesters 11 (10–13)% of annual N inputs. Some river basins, however, sequester >50% of their N inputs, buffering coastal waters against eutrophication and society against greenhouse gas-induced warming. Other basins, releasing >25% more than they receive, are mostly located in the tropics, where recent deforestation, agricultural intensification, and/or exports of land N storage can create large N pollution sources. The tropics produce 56 ± 6% of global land N pollution despite covering only 34% of global land area and receiving far lower amounts of fertilizers than the extratropics. Tropical land use should thus be thoroughly considered in managing global N pollution.

Nature Communications↗
USGS2024

Continental-scale nutrient and contaminant delivery by Pacific salmon

The movement of large amounts of nutrients by migrating animals has ecological benefits for recipient food webs1,2 that may be offset by co-transported contaminants3,4. Salmon spawning migrations are archetypal of this process, carrying marine-derived materials to inland ecosystems where they stimulate local productivity but also enhance contaminant exposure5,6,7. Pacific salmon abundance and biomass are higher now than in the last century, reflecting substantial shifts in community structure8 that probably altered nutrient versus contaminant delivery. Here we combined nutrient and contaminant concentrations with 40 years of annual Pacific salmon returns to quantify how changes in community structure influenced marine to freshwater inputs to western North America. Salmon transported tonnes of nutrients and kilograms of contaminants to freshwaters annually. Higher salmon returns (1976–2015) increased salmon-derived nutrient and contaminant inputs by 30% and 20%, respectively. These increases were dominated by pink salmon, which are short-lived, feed lower in marine food webs than other salmon species, and had the highest nutrient-to-contaminant ratios. As a result, the delivery of nutrients increased at a greater rate than the delivery of contaminants, and salmon inputs became more ecologically beneficial over time. Even still, contaminant loadings may represent exposure concerns for some salmon predators. The Pacific salmon example demonstrates how long-term environmental changes interact with nutrient and contaminant movement across large spatial scales and provides a model for exploring similar patterns with other migratory species.

Nature↗
USGS2011

Use of pharmaceuticals and pesticides to constrain nutrient sources in coastal groundwater of northwestern Long Island, New York, USA

In developed, non-agricultural, unsewered areas, septic systems and fertilizer application to lawns and gardens represent two major sources of nitrogen to coastal groundwater, in addition to atmospheric input. This study was designed to distinguish between these two possible nitrogen sources by analyzing groundwater samples for pharmaceutical residuals, because fertilizers do not contain any of these pharmaceuticals, but domestic wastewater commonly does. In addition, several herbicides and insecticides used in lawn treatment were analyzed as indicators of nitrogen delivery to groundwater from fertilizers. Groundwater samples were taken through piezometres at shoreline sites in unsewered areas surrounding Northport Harbor and in sewered areas adjacent to Manhasset Bay (hereafter referred to as "Northport" and "Manhasset", respectively), both in northwestern Long Island, USA. Excessive nitrogen loading has led to reduced dissolved oxygen concentrations in Long Island Sound, and the groundwater contribution to the nitrogen budget is poorly constrained. The frequent detection of the anticonvulsant compound carbamazepine in groundwater samples of the Northport Harbor area (unsewered), together with the fact that few pesticides associated with lawn applications were detected, suggests that wastewater input and atmospheric input are the likely sources of nitrogen in the Northport groundwater. High concentrations of nitrogen were also detected in the Manhasset (sewered) groundwater. The low detection frequency and concentration of carbamazepine, however, suggest that the sewer system effectively intercepts nitrogen from wastewater there. The likely sources of nitrogen in the Manhasset groundwater are atmospheric deposition and lawn fertilizers, as this area is densely populated.

New York↗
USGS2001

User interface for ground-water modeling: Arcview extension

Numerical simulation for ground-water modeling often involves handling large input and output data sets. A geographic information system (GIS) provides an integrated platform to manage, analyze, and display disparate data and can greatly facilitate modeling efforts in data compilation, model calibration, and display of model parameters and results. Furthermore, GIS can be used to generate information for decision making through spatial overlay and processing of model results. Arc View is the most widely used Windows-based GIS software that provides a robust user-friendly interface to facilitate data handling and display. An extension is an add-on program to Arc View that provides additional specialized functions. An Arc View interface for the ground-water flow and transport models MODFLOW and MT3D was built as an extension for facilitating modeling. The extension includes preprocessing of spatially distributed (point, line, and polygon) data for model input and postprocessing of model output. An object database is used for linking user dialogs and model input files. The Arc View interface utilizes the capabilities of the 3D Analyst extension. Models can be automatically calibrated through the Arc View interface by external linking to such programs as PEST. The efficient pre- and postprocessing capabilities and calibration link were demonstrated for ground-water modeling in southwest Kansas.

Journal of Hydrologic Engineering↗
USGS2008

Effects of spatial resolution ratio in image fusion

In image fusion, the spatial resolution ratio can be defined as the ratio between the spatial resolution of the high-resolution panchromatic image and that of the low-resolution multispectral image. This paper attempts to assess the effects of the spatial resolution ratio of the input images on the quality of the fused image. Experimental results indicate that a spatial resolution ratio of 1:10 or higher is desired for optimal multisensor image fusion provided the input panchromatic image is not downsampled to a coarser resolution. Due to the synthetic pixels generated from resampling, the quality of the fused image decreases as the spatial resolution ratio decreases (e.g. from 1:10 to 1:30). However, even with a spatial resolution ratio as small as 1:30, the quality of the fused image is still better than the original multispectral image alone for feature interpretation. In cases where the spatial resolution ratio is too small (e.g. 1:30), to obtain better spectral integrity of the fused image, one may downsample the input high-resolution panchromatic image to a slightly lower resolution before fusing it with the multispectral image.

International Journal of Remote Sensing↗
USGS2009

Dynamics in phosphorus retention in wetlands upstream of Delavan Lake, Wisconsin

A phosphorus budget was constructed for Delavan Lake Inlet, a perennial riverine wetland with submersed and floating aquatic vegetation in southeastern Wisconsin, to better understand the phosphorus dynamics in natural wetlands and the role of wetlands in lake-rehabilitation efforts. During the growing season, the inlet served as a net source of phosphorus , primarily due to the release of phosphorus from the sediments. More phosphorus was released from the sediments of the inlet (600 kg) than was input from the upstream watershed (460 kg). This release was caused by high pH associated with high photosynthetic activity. During the remainder of the year, the inlet served as a net sink for phosphorus , retaining 6% of die phosphorus input from the watershed. Over the entire year, this wetland was a net source of over 500 kg of phosphorus to downstream Delavan Lake. A constructed riverine wetland upstream of Delavan Lake Inlet demonstrated a similar periodic release of phosphorus . However, in this case, the summer release of phosphorus was less than that trapped during the remainder of the year. The constructed wetland served as a net sink for approximately 20% of the input phosphorus on an annual time scale. The role of existing and constructed wetlands as phosphorus traps is complex. Wetlands can act as a source or a sink for phosphorus depending on the ambient conditions in die wetland . Howa wetland fits into a rehabilitation plan depends upon its net retention efficiency and the importance of the periodic releases of phosphorus to downstream waters.

Wisconsin↗
USGS2020

Importance of accurately quantifying internal loading in developing phosphorus reduction strategies for a chain of shallow lakes

The Winnebago Pool is a chain of 4 shallow lakes in Wisconsin. Because of high external phosphorus (P) inputs to the lakes, the lakes became highly eutrophic, with much P contained in their sediments. In developing a total maximum daily load (TMDL) for these lakes, it is important to determine how their phosphorus concentrations should respond to changes in external P loading. In many TMDLs, internal P loading is assumed to be negligible or it is estimated based on sediment release rates and dissolved oxygen conditions in the lake, and each lake is considered independently. To evaluate these assumptions, internal P loading and external P loading were quantified by developing detailed P budgets for the Winnebago Pool chain of lakes. This information was then inputted into 2 eutrophication models (BATHTUB and Jensen models), which were used to simulate the steady-state and transient effects of various P reduction strategies. The importance of internal P loading varied among lakes, from being a minor source to representing almost 60% of the summer P input. Model results indicate that each lake responds to external P reductions, but internal loading can delay the lake responses, especially in the most downstream lake, Lake Winnebago, where internal P loading was most important to its summer P budget. Accurately quantifying net internal P loading and using this information in lake models are important in evaluating how large shallow lakes should respond to P reduction strategies, setting realistic expectations from watershed P reductions, and guiding TMDL efforts.

Wisconsin↗
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↗