Search USGS⌕ Search

SEARCH · Search USGS

Results for “Ecological Modeling”

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 109 records · Page 6Linked to original sources
USGS2019

Selecting a landscape model for natural resource management applications

Purpose of Review: Climate change and associated ecological impacts have challenged many conventional, observation-based approaches for predicting ecosystem and landscape responses to natural resource management. Complex spatial ecological models provide powerful, flexible tools which managers and others can use to make inferences about management impacts on future, no-analog landscape conditions. However, land managers who wish to use ecosystem and landscape models for natural resource applications are faced with the difficult task of deciding among many models that differ in important ways. Here, we summarize a process to aid managers in the selection of an appropriate model for natural resource management. Recent Findings: To guide management planning, scientifically credible information on how landscapes will respond to management actions under changing climate is required. Landscape models are increasingly used in a management context to evaluate of impacts of changing climate and interacting stressors on ecosystems and to test effects of alternative management options on desired conditions. However, the wide range of available models makes selection of appropriate and viable models a complex process. Summary: We present a series of preliminary steps to define critical scales of time, space, and ecological organization to guide an experimental design for a modeling project and then list a set of criteria for selecting a landscape or ecological model. Material presented includes the preliminary steps (crafting modeling objective, designing modeling project), organizational concerns (resources available, expertise on hand, timelines), and modeling details (complexity, design, documentation) of model selection.

Current Landscape Ecology Reports↗
USGS2016

Resprouting and seeding hypotheses: A test of the gap-dependent model using resprouting and obligate seeding subspecies of Arctostaphylos

Ecological factors favoring either postfire resprouting or postfire obligate seeding in plants have received considerable attention recently. Three ecological models have been proposed to explain patterns of these two life history types. In this study, we test these three models using data from California chaparral. We take an innovative approach to testing these models by not testing community or landscape patterns, but instead, investigating vegetation structure characteristic of four pairs of resprouting and (non-resprouting) obligate seeding subspecies of Arctostaphylos (Ericaceae), a dominant and diverse shrub genus in California chaparral. Data were analyzed for percentage bare ground, elevation, annual precipitation, number of fires, and time between fires and were compared independently for each subspecies pair. Results were consistently supportive of the gap-dependent model suggesting that obligate seeders are favored when post-disturbance gaps are large. Results were inconclusive or contrary to expectations for both of the other two models.

Plant Ecology↗
USGS2020

Testing prediction accuracy in short-term ecological studies

Applied ecology is based on an assumption that a management action will result in a predicted outcome. Testing the prediction accuracy of ecological models is the most powerful way of evaluating the knowledge implicit in this cause-effect relationship, however, the prevalence of predictive modeling and prediction testing are spreading slowly in ecology. The challenge of prediction testing is particularly acute for small-scale studies, because withholding data for prediction testing (e.g., via k -fold cross validation) can reduce model precision. However, by necessity small-scale studies are common. We use one such study that explored small mammal abundance along an elevational gradient to test prediction accuracy of models with varying degrees of information content. For each of three small mammal species, we conducted 5000 iterations of the following process: (1) randomly selected 75 % of the data to develop generalized linear models of species abundance that used detailed site measurements as covariates, (2) used an information theoretic approach to compare the top model with detailed covariates to habitat type-only and null models constructed with the same data, (3) tested those models’ ability to predict the 25 % of the randomly withheld data, and (4) evaluated prediction accuracy with a quadratic loss function. Detailed models fit the model-evaluation data best but had greater expected prediction error when predicting out-of-sample data relative to the habitat type models. Relationships between species and detailed site variables may be evident only within the framework of explicitly hierarchical analyses. We show that even with a small but relatively typical dataset ( n = 28 sampling locations across 125 km over two years), researchers can effectively compare models with different information content and measure models’ predictive power, thus evaluating their own ecological understanding and defining the limits of their inferences. Identifying the appropriate scope of inference through prediction testing is ecologically valuable and is attainable even with small datasets.

Maine, New Hampshire↗
USGS2017

Multi-model comparison highlights consistency in predicted effect of warming on a semi-arid shrub

A number of modeling approaches have been developed to predict the impacts of climate change on species distributions, performance, and abundance. The stronger the agreement from models that represent different processes and are based on distinct and independent sources of information, the greater the confidence we can have in their predictions. Evaluating the level of confidence is particularly important when predictions are used to guide conservation or restoration decisions. We used a multi-model approach to predict climate change impacts on big sagebrush ( Artemisia tridentata ), the dominant plant species on roughly 43 million hectares in the western United States and a key resource for many endemic wildlife species. To evaluate the climate sensitivity of A. tridentata , we developed four predictive models, two based on empirically derived spatial and temporal relationships, and two that applied mechanistic approaches to simulate sagebrush recruitment and growth. This approach enabled us to produce an aggregate index of climate change vulnerability and uncertainty based on the level of agreement between models. Despite large differences in model structure, predictions of sagebrush response to climate change were largely consistent. Performance, as measured by change in cover, growth, or recruitment, was predicted to decrease at the warmest sites, but increase throughout the cooler portions of sagebrush's range. A sensitivity analysis indicated that sagebrush performance responds more strongly to changes in temperature than precipitation. Most of the uncertainty in model predictions reflected variation among the ecological models, raising questions about the reliability of forecasts based on a single modeling approach. Our results highlight the value of a multi-model approach in forecasting climate change impacts and uncertainties and should help land managers to maximize the value of conservation investments.

Global Change Biology↗
USGS2017

Model-based estimators of density and connectivity to inform conservation of spatially structured populations

Conservation and management of spatially structured populations is challenging because solutions must consider where individuals are located, but also differential individual space use as a result of landscape heterogeneity. A recent extension of spatial capture–recapture (SCR) models, the ecological distance model, uses spatial encounter histories of individuals (e.g., a record of where individuals are detected across space, often sequenced over multiple sampling occasions), to estimate the relationship between space use and characteristics of a landscape, allowing simultaneous estimation of both local densities of individuals across space and connectivity at the scale of individual movement. We developed two model-based estimators derived from the SCR ecological distance model to quantify connectivity over a continuous surface: (1) potential connectivity—a metric of the connectivity of areas based on resistance to individual movement; and (2) density-weighted connectivity (DWC)—potential connectivity weighted by estimated density. Estimates of potential connectivity and DWC can provide spatial representations of areas that are most important for the conservation of threatened species, or management of abundant populations (i.e., areas with high density and landscape connectivity), and thus generate predictions that have great potential to inform conservation and management actions. We used a simulation study with a stationary trap design across a range of landscape resistance scenarios to evaluate how well our model estimates resistance, potential connectivity, and DWC. Correlation between true and estimated potential connectivity was high, and there was positive correlation and high spatial accuracy between estimated DWC and true DWC. We applied our approach to data collected from a population of black bears in New York, and found that forested areas represented low levels of resistance for black bears. We demonstrate that formal inference about measures of landscape connectivity can be achieved from standard methods of studying animal populations which yield individual encounter history data such as camera trapping. Resulting biological parameters including resistance, potential connectivity, and DWC estimate the spatial distribution and connectivity of the population within a statistical framework, and we outline applications to many possible conservation and management problems.

Ecosphere↗
USGS2019

Digital mapping of ecological land units using a nationally scalable modeling framework

Ecological site descriptions (ESDs) and associated state-and-transition models (STMs) provide a nationally consistent classification and information system for defining ecological land units for management applications in the United States. Current spatial representations of ESDs, however, occur via soil mapping and are therefore confined to the spatial resolution used to map soils within a survey area. Land management decisions occur across a range of spatial scales and therefore require ecological information that spans similar scales. Digital mapping provides an approach for optimizing the spatial scale of modeling products to best serve decision makers and have the greatest impact in addressing land management concerns. Here, we present a spatial modeling framework for mapping ecological sites using machine learning algorithms, soil survey field observations, soil survey geographic databases, ecological site data, and a suite of remote sensing-based spatial covariates (e.g., hyper-temporal remote sensing, terrain attributes, climate data, land-cover, lithology). Based on the theoretical association between ecological sites and landscape biophysical properties, we hypothesized that the spatial distribution of ecological sites could be predicted using readily available geospatial data. This modeling approach was tested at two study areas within the western United States, representing 6.1 million ha on the Colorado Plateau and 7.5 million ha within the Chihuahuan Desert. Results show our approach was effective in mapping grouped ecological site classes (ESGs), with 10-fold cross-validation accuracies of 70% in the Colorado Plateau based on 1405 point observations across eight expertly-defined ESG classes and 79% in the Chihuahuan Desert based on 2589 point observations across nine expertly-defined ESG classes. Model accuracies were also evaluated using external-validation datasets; resulting in 56 and 44% correct classification for the Colorado Plateau and Chihuahuan Desert, respectively. National coverage of the training and covariate data used in this study provides opportunities for a consistent national-scale mapping effort of ecological sites.

Chihuahuan Desert, Colorado Plateau↗
USGS2011

Social network models predict movement and connectivity in ecological landscapes

Network analysis is on the rise across scientific disciplines because of its ability to reveal complex, and often emergent, patterns and dynamics. Nonetheless, a growing concern in network analysis is the use of limited data for constructing networks. This concern is strikingly relevant to ecology and conservation biology, where network analysis is used to infer connectivity across landscapes. In this context, movement among patches is the crucial parameter for interpreting connectivity but because of the difficulty of collecting reliable movement data, most network analysis proceeds with only indirect information on movement across landscapes rather than using observed movement to construct networks. Statistical models developed for social networks provide promising alternatives for landscape network construction because they can leverage limited movement information to predict linkages. Using two mark-recapture datasets on individual movement and connectivity across landscapes, we test whether commonly used network constructions for interpreting connectivity can predict actual linkages and network structure, and we contrast these approaches to social network models. We find that currently applied network constructions for assessing connectivity consistently, and substantially, overpredict actual connectivity, resulting in considerable overestimation of metapopulation lifetime. Furthermore, social network models provide accurate predictions of network structure, and can do so with remarkably limited data on movement. Social network models offer a flexible and powerful way for not only understanding the factors influencing connectivity but also for providing more reliable estimates of connectivity and metapopulation persistence in the face of limited data.

Proceedings of the National Academy of Sciences of↗
USGS2008

Coastal Louisiana ecosystem assessment and restoration program: The role of ecosystem forecasting in evaluating restoration planning in the Mississippi River Deltaic Plain

The development of ecosystem management plans to restore and rehabilitate natural resources requires an understanding of how specific ecological mechanisms regulate the structure and function of ecosystems. To achieve restoration goals, comprehensive plans and engineering designs must effectively change environmental drivers at the regional level to reduce stress conditions at the local environment that are responsible for ecosystem degradation. This document focuses on the Coastal Louisiana Ecosystem Assessment and Restoration (CLEAR) ecosystem forecasting framework and how it can be used to support the analysis of Louisiana’s coastal restoration plans. Specifically, the framework is designed to (1) develop and incorporate conceptual ecological models that can be used to integrate ecological needs and opportunities with engineering designs, (2) utilize wetland loss rates to describe the most likely “future without” scenario for a variety of ecosystem attributes, (3) estimate broad ecosystem responses to restoration alternatives based on processes associated with succession of geomorphic and ecological systems, and (4) calculate ecological benefits for incorporation into decision support tools associated with large-scale geomorphic and hydrologic processes. This paper provides a brief overview of the spatial framework and modular design of the CLEAR ecosystem forecasting framework and describes in greater detail the evolution of the landscape change module, concepts for its refinement, and how it was utilized in evaluating a coastal restoration alternative proposed in the Coastal Protection and Restoration Authority Preliminary Draft Master Plan. Such projections by the CLEAR forecasting framework can evaluate processes and conditions that result in sustainable coastal ecosystems with habitat functions that support higher trophic levels.

Louisiana↗
USGS2012

General methods for sensitivity analysis of equilibrium dynamics in patch occupancy models

Sensitivity analysis is a useful tool for the study of ecological models that has many potential applications for patch occupancy modeling. Drawing from the rich foundation of existing methods for Markov chain models, I demonstrate new methods for sensitivity analysis of the equilibrium state dynamics of occupancy models. Estimates from three previous studies are used to illustrate the utility of the sensitivity calculations: a joint occupancy model for a prey species, its predators, and habitat used by both; occurrence dynamics from a well-known metapopulation study of three butterfly species; and Golden Eagle occupancy and reproductive dynamics. I show how to deal efficiently with multistate models and how to calculate sensitivities involving derived state variables and lower-level parameters. In addition, I extend methods to incorporate environmental variation by allowing for spatial and temporal variability in transition probabilities. The approach used here is concise and general and can fully account for environmental variability in transition parameters. The methods can be used to improve inferences in occupancy studies by quantifying the effects of underlying parameters, aiding prediction of future system states, and identifying priorities for sampling effort.

Ecology↗
USGS2015

Impacts of climate change on land-use and wetland productivity in the Prairie Pothole Region of North America

Wetland productivity in the Prairie Pothole Region (PPR) of North America is closely linked to climate. A warmer and drier climate, as predicted, will negatively affect the productivity of PPR wetlands and the services they provide. The effect of climate change on wetland productivity, however, will not only depend on natural processes (e.g., evapotranspiration), but also on human responses. Agricultural land use, the predominant use in the PPR, is unlikely to remain static as climate change affects crop yields and prices. Land use in uplands surrounding wetlands will further affect wetland water budgets and hence wetland productivity. The net impact of climate change on wetland productivity will therefore depend on both the direct effects of climate change on wetlands and the indirect effects on upland land use. We examine the effect of climate change and land-use response on semipermanent wetland productivity by combining an economic model of agricultural land-use change with an ecological model of wetland dynamics. Our results suggest that the climate change scenarios evaluated are likely to have profound effects on land use in the North and South Dakota PPR, with wheat displacing other crops and pasture. The combined pressure of land-use and climate change significantly reduces wetland productivity. In a climate scenario with a +4 °C increase in temperature, our model predicts that almost the entire region may lack the wetland productivity necessary to support wetland-dependent species.

North Dakota, South Dakota↗
USGS2017

What determines water temperature dynamics in the San Francisco Bay-Delta system?

Water temperature is an important factor determining estuarine species habitat conditions. Water temperature is mainly governed by advection (e.g., from rivers) and atmospheric exchange processes varying strongly over time (day-night, seasonally) and the spatial domain. On a long time scale, climate change will impact water temperature in estuarine systems due to changes in river flow regimes, air temperature, and sea level rise. To determine which factors govern estuarine water temperature and its sensitivity to changes in its forcing, we developed a process-based numerical model (Delft3D Flexible Mesh) and applied it to a well-monitored estuarine system (the San Francisco Estuary) for validation. The process-based approach allows for detailed process description and a physics-based analysis of governing processes. The model was calibrated for water year 2011 and incorporated 3-D hydrodynamics, salinity intrusion, water temperature dynamics, and atmospheric coupling. Results show significant skill in reproducing temperature observations on daily, seasonal, and yearly time scales. In North San Francisco Bay, thermal stratification is present, enhanced by salinity stratification. The temperature of the upstream, fresh water Delta area is captured well in 2-D mode, although locally—on a small scale—vertical processes (e.g., stratification) may be important. The impact of upstream river temperature and discharge and atmospheric forcing on water temperatures differs throughout the Delta, possibly depending on dispersion and residence times. Our modeling effort provides a sound basis for future modeling studies including climate change impact on water temperature and associated ecological modeling, e.g., clam and fish habitat and phytoplankton dynamics.

California↗
USGS2023

Grizzly bear movement models predict habitat use for nearby populations

Conservation planning and decision-making can be enhanced by ecological models that reliably transfer to times and places beyond those where models were developed. Transferrable models can be especially helpful for species of conservation concern, such as grizzly bears ( Ursus arctos ). Currently, only four grizzly bear populations remain in the contiguous United States. We evaluated transferability of previously derived individual-based, integrated step selection functions (iSSFs) developed from GPS-collared grizzly bears in the Northern Continental Divide Ecosystem by applying them within the nearby Selkirk (SE), Cabinet-Yaak (CYE), and Greater Yellowstone Ecosystems (GYE). We simulated 100 replicates of 5000 steps for each iSSF in each ecosystem, summarized relative use into 10 equal-area classes for each sex, and overlaid GPS locations from bears in the SE, CYE, and GYE on resulting maps. Spearman rank correlations between numbers of locations and class rank were ≥ 0.96 within each study area, indicating models were highly predictive of grizzly bear space use in these nearby populations. Assessment of models using smaller subsets of data in space and time demonstrated generally high predictive accuracy for females. Although generally high across space and time, predictive accuracy for males was low within some watersheds and in summer within the SE and CYE, potentially due to seasonal effects, vegetation, and food assemblage differences. Altogether, these results demonstrated high transferability of our models to landscapes in the Northern Rocky Mountains, suggesting they may be used to evaluate habitat suitability and connectivity throughout the region to benefit conservation planning.

Idaho, Montana, Washington, Wyoming↗
USGS2012

Numerical models of salt marsh evolution: ecological, geomorphic, and climatic factors

Salt marshes are delicate landforms at the boundary between the sea and land. These ecosystems support a diverse biota that modifies the erosive characteristics of the substrate and mediates sediment transport processes. Here we present a broad overview of recent numerical models that quantify the formation and evolution of salt marshes under different physical and ecological drivers. In particular, we focus on the coupling between geomorphological and ecological processes and on how these feedbacks are included in predictive models of landform evolution. We describe in detail models that simulate fluxes of water, organic matter, and sediments in salt marshes. The interplay between biological and morphological processes often produces a distinct scarp between salt marshes and tidal flats. Numerical models can capture the dynamics of this boundary and the progradation or regression of the marsh in time. Tidal channels are also key features of the marsh landscape, flooding and draining the marsh platform and providing a source of sediments and nutrients to the marsh ecosystem. In recent years, several numerical models have been developed to describe the morphogenesis and long-term dynamics of salt marsh channels. Finally, salt marshes are highly sensitive to the effects of long-term climatic change. We therefore discuss in detail how numerical models have been used to determine salt marsh survival under different scenarios of sea level rise.

Reviews of Geophysics↗
USGS2012

Simulation of groundwater flow and hydrologic effects of groundwater withdrawals from the Kirkwood-Cohansey aquifer system in the Pinelands of southern New Jersey

The Kirkwood-Cohansey aquifer system is an important source of present and future water supply in southern New Jersey. Because this unconfined aquifer system also supports sensitive wetland and aquatic habitats within the New Jersey Pinelands (Pinelands), water managers and policy makers need up-to-date information, data, and projections that show the effects of potential increases in groundwater withdrawals on these habitats. Finite-difference groundwater flow models (MODFLOW) were constructed for three drainage basins (McDonalds Branch Basin, 14.3 square kilometers (km 2 ); Morses Mill Stream Basin, 21.63 km 2 ; and Albertson Brook Basin, 52.27 km 2 ) to estimate the effects of potential increases in groundwater withdrawals on water levels and the base-flow portion of streamflow, in wetland and aquatic habitats. Three models were constructed for each drainage basin: a transient model consisting of twenty-four 1-month stress periods (October 2004 through September 2006); a transient model to simulate the 5- to 10-day aquifer tests that were performed as part of the study; and a high-resolution, steady-state model used to assess long-term effects of increased groundwater withdrawals on water levels in wetlands and on base flow. All models were constructed with the same eight-layer structure. The smallest horizontal cell dimensions among the three model areas were 150 meters (m) for the 24-month transient models, 10 m for the steady-state models, and 3 m for the transient aquifer-test models. Boundary flows of particular interest to this study and represented separately are those for wetlands, streams, and evapotranspiration. The final variables calibrated from both transient models were then used in steady-state models to assess the long-term effects of increased groundwater withdrawals on water levels in wetlands and on base flow. Results of aquifer tests conducted in the three study areas illustrate the effects of withdrawals on water levels in wetlands and on base flow. Pumping stresses at aquifer-test sites resulted in measurable drawdown in each observation well installed for the tests. The magnitude of drawdown in shallow wetland observation wells at the end of pumping ranged from 5.5 to 16.7 centimeters (cm). The stresses induced by the respective tests reduced the flow of the smallest stream (McDonalds Branch) by 75 percent and slightly reduced flow in a side channel of Morses Mill Stream, but did not measurably affect the flow of Morses Mill Stream or Albertson Brook. Results of aquifer-test simulations were used to refine the estimates of hydraulic properties used in the models and to confirm the ability of the model to replicate observed hydrologic responses to pumping. Steady-state sensitivity simulation results for a variety of single well locations and depths were used to define overall “best-case” (smallest effect on wetland water levels and base flow) and “worst-case” (greatest effect on wetland water levels and base flow) groundwater withdrawal configurations. “Best-case” configurations are those for which the extent of the wetland areas within a 1-kilometer (km) radius of the withdrawal well is minimized, the well is located at least 100 m and as far from wetland boundaries as possible, and the withdrawal is from a deep well (50–90 m deep). “Worst-case” configurations are those for which the extent of wetlands within a 1-km radius of the withdrawal well is maximized, the well is located 100 m or less from a wetland boundary, and the withdrawal is from a relatively shallow well (30–67 m deep). “Best-” and “worst-case” simulations were applied by locating hypothetical wells across the study areas and assigning groundwater withdrawals so that the sum of the withdrawals for the basin is equal to 5, 10, 15, and 30 percent of overall recharge. The results were compared to the results of simulations of no groundwater withdrawals. Results for withdrawals of 5 percent of recharge show that the area of wetland water-level decline that exceeded 15 cm was as much as 1.5 percent of the total wetland area for the “best-case” simulations and as much as 9.7 percent of the total wetland area for the “worst-case” simulations. For the same withdrawals, base-flow reduction was as much as 5.1 percent for the “best-case” simulations and as much as 8.6 percent for the “worst-case” simulations. Results for withdrawals of 30 percent of recharge show that the area of wetland water-level decline that exceeded 15 cm was as much as 70 percent of the total wetland area for the “best-case” simulations and as much as 84 percent of the total wetland area for the “worst-case” simulations. For the same withdrawals, base-flow reduction was as much as 30 percent for the “best-case” simulations and as much as 51 percent for the “worst-case” simulations. Results for withdrawals of 10 and 15 percent of recharge show decreased water levels and base flow that are intermediate between those simulated for 5 and 30 percent of recharge. Several approaches for applying the results of this study to other parts of the Pinelands were explored. An analytical-modeling technique based on the Thiem equation and image-well theory was developed to estimate local drawdown distributions resulting from withdrawals in other areas within the Pinelands. Results of example applications of this technique were compared with those of the numerical simulations used in this study and were shown to be useful. Differences among the three basins in the simulated percentage of basin wetlands affected by drawdown were found to be related to the proximity of wetlands to streams, the proximity of wetlands to pumped wells, and the vertical conductance of the aquifer system. These factors formed the basis for an index of wetland vulnerability to drawdown. An empirically-derived model based on the Gompertz function and the wetland vulnerability index was developed, tested, and shown to be an effective means to evaluate potential drawdown in wetlands at a basin scale throughout the Pinelands. Base-flow reduction can be estimated from generalized results of the numerical models, estimates of evapotranspiration reduction, or available regional groundwater flow models. These approaches could be used to evaluate alternative water-supply strategies and, in conjunction with ecological-modeling results, to determine maximum basin withdrawal rates within the limits of acceptable ecological change.

New Jersey↗
USGS2021

A machine learning approach to modeling streamflow with sparse data in ungaged watersheds on the Wyoming Range, Wyoming, 2012–17

Scant availability of streamflow data can impede the utility of streamflow as a variable in ecological models of aquatic and terrestrial species, especially when studying small streams in watersheds that lack streamgages. Streamflow data at fine resolution and broad extent were needed by collaborators for ecological research on small streams in several ungaged watersheds of southwestern Wyoming, where streamflow data are sparse. To improve the utility of sparse streamflow data to ecological research in ungaged watersheds, we developed a machine learning approach in R for modeling spatially and temporally continuous monthly streamflow from 2012 through 2017 in three semiarid montane-steppe watersheds (with drainage areas of 26–55 square miles and mean elevations of 8,031–8,455 feet) on the Wyoming Range in the upper Green River Basin. A machine learning streamflow (MLFLOW) model was calibrated and validated with 971 discrete streamflow observations and 24 static and dynamic predictor variables derived from geospatial and time series data on climatic, physiographic, and anthropogenic characteristics affecting streamflow. The predictor variables were temporally and spatially conditioned to amplify the relation of predictor variables to monthly streamflow. The MLFLOW model had satisfactory agreement between observed and predicted streamflow (coefficient of determination [ R 2 ]=0.80, Nash-Sutcliffe efficiency [NSE]=0.79, NSE with log-transformed data [logNSE]=0.82, and percent bias [PBIAS]=0.7 percent). NSE and logNSE indicated the MLFLOW model performed equally well for high and low flows, and PBIAS indicated the MLFLOW model did not overpredict or underpredict monthly streamflow. Streamflow predictions seemed to well represent the annual hydrograph within the study area during the study period. The most important variables (statistically important in the MLFLOW model) for explaining monthly streamflow were temporally and spatially conditioned dynamic climatic variables, mostly precipitation and snow water equivalent. Importance of the static and dynamic variables did not differ substantially among the three watersheds but differed considerably among the 6 years. Monthly streamflow increased with increasing precipitation, snow water equivalent, and drainage area but decreased with increasing forest cover, elevation, evapotranspiration, and temperature. The MLFLOW model was most sensitive to selection of dynamic climatic variables. Unconditioned dynamic climatic variables alone explained 54 percent of the variance ( R 2 =0.54) in monthly streamflow, whereas adding static physiographic and anthropogenic variables only explained 12 percent more of the variance ( R 2 =0.66). Also, spatial conditioning of all variables together with temporal conditioning of dynamic variables increased the variance explained in the MLFLOW model by another 14 percent ( R 2 =0.80). The MLFLOW model also had greater sensitivity to temporal than to spatial differences in the data. For the MLFLOW model trained with observations from all watersheds and years or for models trained with observations from all except one watershed or 1 year left out sequentially, performance was better in testing on observations from each watershed than from each year separately. Also, performance was better for models fitted to fewer sites than to fewer months of observations. The greatest utility of the modeling approach is the ease of use and the speed of processing input data, running the model, and interpreting the model output, whereas the greatest limitation is the need for spatially and temporally representative streamflow observations to drive the model. Although familiarity with R is necessary, only a working knowledge of hydrology (for selecting appropriate predictor variables and evaluating the quality of streamflow observations) and a rudimentary understanding of machine learning models are needed. Therefore, this modeling approach is practicable for other scientists who work with water but who are not hydrologists.

Wyoming↗
USGS2024

Assessing the added value of antecedent streamflow alteration information in modeling stream biological condition

In stream systems, disentangling relationships between biology and flow and subsequent prediction of these relationships to unsampled streams is a common objective of large-scale ecological modeling. Often, streamflow metrics are derived from aggregating continuous streamflow records available at a subset of stream gages into long-term flow regime descriptors. Despite demonstrated value, shortcomings of these long-term approaches include spatial restriction to locations with long-term continuous flow records (commonly, biased toward larger systems) and omission of potentially ecologically important short-term (i.e., ≤1 year) antecedent streamflow information. We used long-term flow regime and short-term antecedent streamflow alteration information to evaluate relative performance in modeling stream fish biological condition. We compared results to understand whether short-term antecedent streamflow information improved models of fish biological condition. Results indicated that models incorporating short-term antecedent data performed better than those relying solely on long-term flow regime data (kappa statistic = 0.29 and 0.23, respectively) and improved prediction accuracy among stream sizes and in six of nine ecoregions. Additionally, models relying solely on short-term streamflow information performed similarly to those with only long-term streamflow information (kappa = 0.23). Incorporating short-term antecedent streamflow metrics may provide added ecological information not fully captured by long-term flow regime summaries in macroscale modeling efforts or perform similarly to long-term streamflow data when long-term data are not available.

Science of the Total Environment↗
USGS2008

Use of an integrated flow model to estimate ecologically relevant hydrologic characteristics at stream biomonitoring sites

We developed an integrated hydroecological model to provide a comprehensive set of hydrologic variables representing five major components of the flow regime at 856 aquatic-invertebrate monitoring sites in New Jersey. The hydroecological model simulates streamflow by routing water that moves overland and through the subsurface from atmospheric delivery to the watershed outlet. Snow accumulation and melt, evapotranspiration, precipitation, withdrawals, discharges, pervious- and impervious-area runoff, and lake storage were accounted for in the water balance. We generated more than 78 flow variables, which describe the frequency, magnitude, duration, rate of change, and timing of flow events. Highly correlated variables were filtered by principal component analysis to obtain a non-redundant subset of variables that explain the majority of the variation in the complete set. This subset of variables was used to evaluate the effect of changes in the flow regime on aquatic-invertebrate assemblage structure at 856 biomonitoring sites. We used non-metric multidimensional scaling (NMS) to evaluate variation in aquatic-invertebrate assemblage structure across a disturbance gradient. We employed multiple linear regression (MLR) analysis to build a series of MLR models that identify the most important environmental and hydrologic variables driving the differences in the aquatic-invertebrate assemblages across the disturbance gradient. The first axis of NMS ordination was significantly related to many hydrologic, habitat, and land-use/land-cover variables, including the average number of annual storms producing runoff, ratio of 25-75% exceedance flow (flashiness), diversity of natural stream substrate, and the percentage of forested land near the stream channel (forest buffer). Modifications in the hydrologic regime as the result of changes in watershed land use appear to promote the retention of highly tolerant aquatic species; in contrast, species that are sensitive to hydrologic instability and other anthropogenic disturbance become much less prevalent. We also found strong relations between an index of invertebrate-assemblage impairment, its component metrics, and the primary disturbance gradient. The process-oriented watershed modeling approach used in this study provides a means to evaluate how natural landscape features interact with anthropogenic factors and assess their effects on flow characteristics and stream ecology. By combining watershed modeling and indirect ordination techniques, we were able to identify components of the hydrologic regime that have a considerable effect on aquatic-assemblage structure and help in developing short- and long-term management measures that mitigate the effects of anthropogenic disturbance in stream systems.

New Jersey↗