Dynamic Linear Modeling Research Articles

Dynamic Linear Modeling for Characterizing and Predicting the Patterns of Summer Monsoon Rainfall in Northwest India

Dynamic Linear Modeling for Characterizing and Predicting the Patterns of Summer Monsoon Rainfall in Northwest India

The future ozone trends in changing climate simulated with SOCOLv4

Abstract. This study evaluates the future evolution of atmospheric ozone simulated with the Earth system model (ESM) SOCOLv4. Simulations have been performed based on two potential shared socioeconomic pathways (SSPs): the middle-of-the-road (SSP2-4.5) and fossil-fueled (SSP5-8.5) scenarios. The future changes in ozone, as well as in chemical drivers (NOx and CO) and temperature, were estimated between 2015 and 2099 and for several intermediate subperiods (i.e., 2015–2039, 2040–2069, and 2070–2099) via dynamic linear modeling. In both scenarios, the model projects a decline in tropospheric ozone in the future that starts in the 2030s in SSP2-4.5 and after the 2060s in SSP5-8.5 due to a decrease in concentrations of NOx and CO. The results also suggest a very likely ozone increase in the mesosphere and the upper and middle stratosphere, as well as in the lower stratosphere at high latitudes. Under SSP5-8.5, the ozone increase in the stratosphere is higher because of stronger cooling (>1 K per decade) induced by greenhouse gases (GHGs), which slows the catalytic ozone destruction cycles. In contrast, in the tropical lower stratosphere, ozone concentrations decrease in both experiments and increase over the middle and high latitudes of both hemispheres due to the speeding up of the Brewer–Dobson circulation, which is stronger in SSP5-8.5. No evidence was found of a decline in ozone levels in the lower stratosphere at mid-latitudes. In both future scenarios, the total column ozone is expected to be distinctly higher than present in middle to high latitudes and might be lower in the tropics, which causes a decrease in the mid-latitudes and an increase in the tropics in the surface level of ultraviolet radiation. The results of SOCOLv4 suggest that the stratospheric-ozone evolution throughout the 21st century is strongly governed not only by a decline in halogen concentration but also by future GHG forcing. In addition, the tropospheric-ozone column changes, which are mainly due to the changes in anthropogenic emissions of ozone precursors, also have a strong impact on the total column. Therefore, even though the anthropogenic halogen-loading problem has been brought under control to date, the sign of future ozone column changes, globally and regionally, is still unclear and largely depends on diverse future human activities. The results of this work are, thus, relevant for developing future strategies for socioeconomic pathways.

Climate and conspecific density inform phenotypic forecasting of juvenile Pacific salmon body size

Abstract Predicting effects of climate on fitness‐linked phenotypic traits, such as body size, is important for the management and conservation of species in the face of global change. During sensitive life stages, small changes in mean trait values can have large effects on survival and population productivity. The transition from freshwater to saltwater by migrating anadromous fishes such as Pacific salmon (genus Oncorhynchus) is a critical life history transition, where survival is mediated by the size of migrating individuals. For salmon that spend extended periods rearing in freshwater, the size at ocean entry (i.e., smolt length) may be sensitive to changes in freshwater conditions shaped by biotic and abiotic factors, yet long‐term phenotypic time series for exploring these responses are rare. We reconstructed a four‐decade time series of smolt length from archives of returning adult scales to quantify population‐specific responses to climate and conspecific density in a small watershed. Dynamic linear modeling found that the relationship between a proxy for cohort density and smolt length at ocean entry was consistently negative, suggestive of competition. In contrast, there was a positive, yet dynamic, relationship between a proxy for density of emerging fry during the second summer of growth and smolt length. The effect of temperature on smolt length was not consistent between two populations of sockeye salmon O. nerka that likely use distinct habitats within the watershed. A positive relationship between smolt length and temperature was only detected for the less abundant, early returning population. Predictions of smolt length showed variable responses under scenarios of increasing temperature and high and low densities of conspecifics. Collectively, these results reveal population‐specific responses to temperature and density, suggesting that local habitat conditions may filter larger‐scale climate drivers.

Subspace-Based Representation and Learning for Phonotactic Spoken Language Recognition

Phonotactic constraints can be employed to distinguish languages by representing a speech utterance as a multinomial distribution or phone events. In the present study, we propose a new learning mechanism based on subspace-based representation, which can extract concealed phonotactic structures from utterances, for language verification and dialect/accent identification. The framework mainly involves two successive parts. The first part involves subspace construction. Specifically, it decodes each utterance into a sequence of vectors filled with phone-posteriors and transforms the vector sequence into a linear orthogonal subspace based on low-rank matrix factorization or dynamic linear modeling. The second part involves subspace learning based on kernel machines, such as support vector machines and the newly developed subspace-based neural networks (SNNs). The input layer of SNNs is specifically designed for the sample represented by subspaces. The topology ensures that the same output can be derived from identical subspaces by modifying the conventional feed-forward pass to fit the mathematical definition of subspace similarity. Evaluated on the “General LR” test of NIST LRE 2007, the proposed method achieved up to 52%, 46%, 56%, and 27% relative reductions in equal error rates over the sequence-based PPR-LM, PPR-VSM, and PPR-IVEC methods and the lattice-based PPR-LM method, respectively. Furthermore, on the dialect/accent identification task of NIST LRE 2009, the SNN-based system performed better than the aforementioned four baseline methods.

Understanding the Relationships between Demand for Shared Ride Modes: Case Study using Open Data from New York City

The concept of shared travel, making trips with other users via a common vehicle, is far from novel. However, a changing technological climate has laid the tracks for new dynamically shared modes in the form of transportation network companies (TNCs), to substantially impact travel behavior. The current body of research on how these modal offerings impact the demand for existing shared modes (e.g., bikeshare, transit) is growing. However, a comprehensive investigation of the temporal evolution of the demand for TNCs and their relationship to other shared modes, is lacking. This research tackles this important limitation by analyzing ridership data for TNCs, taxi, subway, and Citi Bike in New York City using daily ridership data from January 2015 through June 2017. The primary objective was to understand the relationship between TNCs and other shared modal offerings while accounting for the influence of temporal trends and other exogenous factors. A dynamic linear modeling framework was formulated to accommodate time-dependent trends, periodicity, and time-varying exogenous factors on the demand for TNCs. As a preliminary work, the findings of this study reinforce the observed substitution relationship between taxis and TNCs. The results may also indicate a substitutional relationship between TNCs and Citi Bike, and a complementary relationship with subway, however these results still need to be explored further. With potentially impactful findings for planning and policymakers, the predictive model developed in the study can be used to carry out forecasting in support of short- and long-term operations and planning applications.

Dynamic linear modeling, identification and precise control of a walking piezo-actuated stage

Abstract In this paper, a dynamic model is firstly investigated for a flexible mechanical stage driven by a walking piezoelectric actuator (WPA). The developed model takes into account the WPA, the mechanical stage and moreover the connection part between them as an overall piezo-actuated stage. The model of the WPA is derived mainly from the electrical, piezoelectric and mechanical sides. Besides, the WPA and mechanical stage are treated as elements with lumped mass. Then the proposed model is identified based on the open-loop frequency response data. Finally, hybrid closed-loop controllers are designed for point-to-point (PTP) positioning and sinusoidal trajectory tracking control of the overall piezo-actuated stage. The hybrid control strategy includes displacement error feedback and velocity feed-forward control algorithms. Furthermore, the transient profile and tracking differentiator is proposed for quick settling of PTP positioning control, and the discrete-time repetitive controller (RC) is adopted to enhance the tracking precision for the periodic sinusoidal trajectory. Experimental results show that for the 300 μ m PTP positioning the settling time is 0.12 s to keep the steady error within 25 nm, and for the sinusoidal trajectory of 50 μ m amplitude the maximum tracking error is 1.99%. These results clearly demonstrate the high precision performance of the developed hybrid controllers for the walking piezo-actuated stage.

Multi‐decadal climate and fishing predictors of abundance for U.S. South Atlantic coastal fishes and invertebrates

AbstractAbundance of marine stocks fluctuates in response to both internal processes (e.g., density dependence) and exogenous drivers, including the physical environment, fishing, and trophodynamic interactions. In the United States, research investigating ecosystem drivers has been focused in data‐rich systems, primarily in the North Atlantic and North Pacific. To develop a more holistic understanding of important ecosystem drivers in the Southeast U.S. continental shelf Large Marine Ecosystem, we applied generalized linear and dynamic linear modeling to investigate the effects of climate and fishing covariates on the relative abundance trends of 71 demersal fish and invertebrate species sampled by a coastal trawl survey during 1990–2013. For the assemblage as a whole, fishing effects predominated over climate effects. In particular, changes in trawling effort within the penaeid shrimp fishery governed abundance trends of bony fishes, invertebrates, and elasmobranchs, a likely result of temporal changes in bycatch mortality. Changes in trawling intensity induced changes in overall community composition and appear to have altered trophic interactions among particular species. Among climate indices investigated, the Pacific Decadal Oscillation and the Western Bermuda High Index were most prevalent in well‐supported dynamic linear models. Observed annual abundance trends were synchronous among some taxonomically related species, highlighting similar responses to exogenous influences based on life history. This study strengthens the foundation for generating hypotheses and advancing ecosystem‐based fisheries research within the region.

Side by side: Modeling dyadic physiological linkage in strangers.

Physiological linkage (PL) refers to coordinated physiological responses among interacting partners (Feldman, 2012a), thought to offer mammals evolutionary advantages by promoting survival through social groups. Although PL has been observed in dyads who are familiar or have close relationships (e.g., parent-infant interactions, romantic couples), less is known with regard to PL in stranger dyads. The current study used dynamic linear time series modeling to assess cardiac interbeat interval linkage in 26 same-gender stranger dyads (17 female and 9 male dyads; 18-22 years old) while they spoke or wrote about emotional or neutral life events. The estimated coefficients in bivariate regression models indicated small but statistically significant PL effects for both male and female dyads. The PL effect was stronger for female dyads, extending to a lag of 4 seconds. For male dyads, the effect was statistically significant but weaker than for female dyads, extending only to a lag of 1 second. No statistically significant differences in PL were noted for type of task (i.e., baseline, writing, speaking, listening) or with differing task emotional content. Frequency domain analysis based on the estimated dynamic models yielded similar results. Our results suggest that PL can be detected among strangers in this setting and appears to be stronger and longer-lasting in women. Our findings are discussed in terms of the importance of biological synchrony in humans, gender differences, and possible implications for objective measurement of social reciprocity at a physiological level. (PsycINFO Database Record

Dynamic linear modeling of monthly electricity demand in Japan: Time variation of electricity conservation effect.

After the severe nuclear disaster in Fukushima, which was triggered by the Great East Japan earthquake in March 2011, nuclear power plants in Japan were temporarily shut down for mandatory inspections. To prevent large-scale blackouts, the Japanese government requested companies and households to reduce electricity consumption in summer and winter. It is reported that the domestic electricity demand had a structural decrease because of the electricity conservation effect (ECE). However, quantitative analysis of the ECE is not sufficient, and especially time variation of the ECE remains unclear. Understanding the ECE is important because Japan’s NDC (nationally determined contribution) assumes the reduction of CO2 emissions through aggressive energy conservation. In this study, we develop a time series model of monthly electricity demand in Japan and estimate time variation of the ECE. Moreover, we evaluate the impact of electricity conservation on CO2 emissions from power plants. The dynamic linear model is used to separate the ECE from the effects of other irrelevant factors (e.g. air temperature, economic production, and electricity price). Our result clearly shows that consumers’ electricity conservation behavior after the earthquake was not temporary but became established as a habit. Between March 2011 and March 2016, the ECE on industrial electricity demand ranged from 3.9% to 5.4%, and the ECE on residential electricity demand ranged from 1.6% to 7.6%. The ECE on the total electricity demand was estimated at 3.2%–6.0%. We found a seasonal pattern that the residential ECE in summer is higher than that in winter. The emissions increase from the shutdown of nuclear power plants was mitigated by electricity conservation. The emissions reduction effect was estimated at 0.82 MtCO2–2.26 MtCO2 (−4.5% on average compared to the zero-ECE case). The time-varying ECE is necessary for predicting Japan’s electricity demand and CO2 emissions after the earthquake.

Micro-level mechanisms of identity development: The role of emotional experiences in commitment development.

Based on Marcia's theory, many researchers consider exploration and commitment as the main processes in identity development. Although some identity theorists have hypothesized that emotional experience may also be an important part of the mechanisms of identity development, empirical research to investigate this claim has been lagging behind. In this study, we shed light on the role of emotional experiences in micro-level commitment dynamics, and compare this to the role of exploration. We take a within-individual approach, and particularly focus on educational commitment. We collected weekly measurements among 103 first year university students over several months, resulting in 22 to 30 measurements for each individual. Every week, the students reported an important experience and accompanying positive and negative emotions, their level of educational exploration and commitment. We generated linear growth models for each individual separately, using Dynamic Linear Modeling. These individual models generate regression weights that indicate how strong the impact is of exploration, positive and negative emotional experiences on changes in micro-level commitment for each individual. Our main finding is that both positive and negative emotional experiences are indeed related to changes in educational commitment. Positive experiences, but surprisingly, also negative experiences, are related to increases in educational commitment for the majority of individuals. Moreover, for the large majority of individuals, the impact of emotional experiences is larger than the impact of exploration. Therefore, we conclude that it is highly likely that emotional experiences are an essential part of the micro-level mechanisms of identity development. (PsycINFO Database Record

Dynamic linear modeling of a refrigeration process with electronic expansion valve actuator

Usually, commercial control solutions for superheat control still use PID controllers as a standard. Although there are several applications of advanced control in refrigeration processes in the literature, there isn't a consensus about the optimal control solution for each system. The implementation of advanced control algorithms ultimately depends on accurate process knowledge in the form of dynamic mathematical models. This study aims to take a first step toward the designing an adaptive stochastic MPC controller for superheat control in an R404 refrigeration cycle with electronic expansion valve by developing stochastic dynamic models of the process. Both time-varying and time-invariant versions of the models are identified. Statistical validation results show whitening of the residuals of the time-invariant models, creating a basis for comparison. The recursive estimation of the time-varying parameters was realized with the Kalman Filter and the Forgetting Factor algorithms. Results of validation tests by simulation show good results, with average output errors between 0.05 and 1.39°C, indicating that the ARMAX with time-varying parameters may be a good presentation for this system.

Energy prediction of CUDA application instances using dynamic regression models

GPGPUs no longer seem to be an inconsequential component of supercomputing architectures, and a section of HPC application developers no longer refrain from utilizing GPGPUs. CUDA, in general, has remained a successful computing platform for those architectures. Thousands of scientific applications from various domains, such, as bio-informatics, HEP, and so forth, have been accelerated using CUDA in the past few years. In fact, the energy consumption issue still remains a serious challenge for the HPC and GPGPU communities. This paper proposes energy prediction approaches using dynamic regression models, such as parallel dynamic random forest modeling (P-DynRFM), dynamic random forest modeling (DynRFM), dynamic support vector machines (DynSVM), and dynamic linear regression modeling (DynLRM). These models identify energy efficient CUDA application instances while considering the block size, grid size, and the other tunable parameters, such as problem size. The predictions of CUDA application instances have been attained by executing a few CUDA application instances and predicting the other CUDA application instances based on the performance metrics of applications, such as number of instructions, memory issues, and so forth. The proposed energy prediction mechanisms were evaluated with CUDA applications such as Nbody and Particle Simulations on two GPGPU machines. The proposed dynamic prediction mechanisms achieved a 50.26 to 61.23 percentage of energy/performance prediction improvements when compared to the classical prediction models; and, the parallel implementation of the dynamic RFM (P-DynRFM) recorded over 83 percentage points of prediction time improvements.

Evaluating the relationships between watershed physiography, land use patterns, and phosphorus loading in the bay of Quinte basin, Ontario, Canada

Abstract The Bay of Quinte, a Z-shaped embayment at the northeastern end of Lake Ontario, represents a characteristic case of an area of concern (AOC) where scientific uncertainty underlies management efforts to address eutrophication problems. The present study attempts to examine the relationships between total phosphorus (TP) (i.e., concentrations from point and non-point sources, net export, and yield) and watershed attributes, including land use cover and physiographic characteristics. Dynamic linear modeling is applied to assess temporal trends in the flow rates in five major tributaries of the area. Urbanized catchments near the Bay of Quinte show more dynamic flow patterns with significant within- and among-year variability and stronger response to local precipitation patterns. The “flashy” behavior of urban land is characterized by relatively high net TP export and TP yield. Self-organizing map analysis, classified the watershed into six distinct spatial clusters, representing two agricultural, one forested, one urban, one pasture-dominated, and one transitional area. Our study shows that agricultural and urban sites profoundly shape riverine TP dynamics. We also provide evidence that tributaries draining agricultural areas exhibit considerable variability, depending on management practices and soil properties. Excessive fertilizer or manure application in croplands, management practices (e.g., crop types and tillage methods) and soil hydrological characteristics (e.g., hydraulic conductivity and permeability) generally determine soil phosphorus level and consequent losses through erosion and runoff from agricultural sites. In a similar manner, pasture-based grazing systems appear to play an important role in modulating the fate and transport of phosphorus in the Bay of Quinte watershed.

A Bayesian assessment of the mercury and PCB temporal trends in lake trout (Salvelinus namaycush) and walleye (Sander vitreus) from lake Ontario, Ontario, Canada

Polychlorinated biphenyls (PCBs) and total mercury (THg) are two of the most prevalent contaminants, resulting in restrictive advisories on consuming fish from the Laurentian Great Lakes. The goal of this study is to examine the temporal trends of the two contaminants in walleye (Sander vitreus) and lake trout (Salvelinus namaycush) for Lake Ontario. We employed Bayesian inference techniques to parameterize three different strategies of time series analysis: dynamic linear, exponential decay, and mixed-order modeling. Our analysis sheds light on the role of different covariates (length, lipid content) that can potentially hamper the detection of the actual temporal patterns of fish contaminants. Both PCBs and mercury demonstrate decreasing temporal trends in lake trout males and females. Decreasing PCB trends are evident in walleye, but the mean annual mercury levels are characterized by a “wax and wane” pattern, suggesting that specific fish species may not act as bio-indicators for all contaminants. This finding may be attributed to the shifts in energy trophodynamics along with the food web alterations induced from the introduction of non-native species, the intricate nature of the prey–predator interactions, the periodicities of climate factors, and the year-to-year variability of the potentially significant fluxes from atmosphere or sediments. Finally, a meaningful risk assessment exercise will be to elucidate the role of within-lake fish contaminant variability and evaluate the potential bias introduced when drawing inference from pooled datasets.

Building a comprehensive mill-level database for the Industrial Sectors Integrated Solutions (ISIS) model of the U.S. pulp and paper sector.

Air emissions from the U.S. pulp and paper sector have been federally regulated since 1978; however, regulations are periodically reviewed and revised to improve efficiency and effectiveness of existing emission standards. The Industrial Sectors Integrated Solutions (ISIS) model for the pulp and paper sector is currently under development at the U.S. Environmental Protection Agency (EPA), and can be utilized to facilitate multi-pollutant, sector-based analyses that are performed in conjunction with regulatory development. The model utilizes a multi-sector, multi-product dynamic linear modeling framework that evaluates the economic impact of emission reduction strategies for multiple air pollutants. The ISIS model considers facility-level economic, environmental, and technical parameters, as well as sector-level market data, to estimate the impacts of environmental regulations on the pulp and paper industry. Specifically, the model can be used to estimate U.S. and global market impacts of new or more stringent air regulations, such as impacts on product price, exports and imports, market demands, capital investment, and mill closures. One major challenge to developing a representative model is the need for an extensive amount of data. This article discusses the collection and processing of data for use in the model, as well as the methods used for building the ISIS pulp and paper database that facilitates the required analyses to support the air quality management of the pulp and paper sector.

Fish contamination in Lake Erie: An examination of temporal trends of organochlorine contaminants and a Bayesian approach to consumption advisories

When examining environmental levels of organic contaminants, much of our focus has been placed on fish due to their greater potential to bioaccumulate and their direct linkage with human as a staple of their diet. Contaminant levels in Great Lakes fish communities have been closely monitored over the last few decades, and the resulting information has been indispensable in guiding consumption advisories. In this study, we first conducted an analysis of temporal trends of three organochlorines (hexachlorobenzene, octachlorostyrene, and α-hexachlorocyclohexane) in five Lake Erie fish species using dynamic linear modeling, while explicitly considering fish length and lipid content as covariates. Our results indicate that the levels of the three compounds have been steadily decreasing during the late 1970s to 2007, although there were instances in which the fish organochlorine contents exhibited fluctuations through time. The second part of our analysis focused on the development of a Bayesian framework to update fish consumption advisories. We present a methodology that incorporates the uncertainty in contaminant predictions and the natural variability in fish length and lipid content, while remaining flexible for different human weights and diet patterns. We then illustrate our Bayesian framework for two important contaminants in the Great Lakes region, mercury and PCBs. We established thresholds for each contaminant based on the tolerable daily intake (TDI) values and made predictive statements about the probability of exceedance of these critical levels. Our study also discusses how failure to account for model uncertainty/error can have profound implications for the credibility of the predictive risk assessment statements derived. The proposed Bayesian approach to fish consumption advisories can serve as a valuable framework for year-specific, highly customizable risk assessment statements that also account for the inherent variability in natural systems.

Examination of temporal DDT trends in Lake Erie fish communities using dynamic linear modeling

The industrial pesticide dichlorodiphenyltrichloroethane (DDT) was initially heralded for its effectiveness against malaria and agricultural pests, but was eventually banned in North America during the 1970s due to growing concerns about its detrimental impacts on wildlife. Despite the successful termination of its commercial application, the persistent and bioaccumulative nature of DDT has resulted in lingering concentrations in aquatic food webs, particularly in upper trophic-level fish species. In this study, we used dynamic linear modeling to examine temporal trends of four DDT compounds (p,p′-DDT, o,p′-DDT, p,p′-DDE, and p,p′-DDD) in nine fish species in Lake Erie from 1976 to 2007, while considering both fish length and lipid content as covariates. Our results indicate that the levels of both p,p′-DDT and o,p′-DDT have been decreasing, often to the detection limit, with slowing decline rates during the second half of the study period. The p,p′-DDE levels were much more variable, exhibiting large fluctuations through time (though usually with a net downward trajectory), with the annual rates of change of the corresponding concentrations remaining negative or (more recently) near zero. Similarly, p,p′-DDD levels fluctuated (though to a lesser degree) over time, with gradually slowing decline rates in many fish species, such as smallmouth bass and freshwater drum. The results are in agreement with our understanding that DDE and DDD are degradation products of p,p′-DDT, and thus continue to be produced, as DDT is broken down. Declining trends observed for nearly all congeners and fish species indicate reduction of DDT risks in the Lake Erie fish communities.

Bayesian dynamic linear modeling for exploring the impact of recent financial crisis on Japan Credit Default Swap market

This paper investigates the effects of recent subprime financial crisis on Japan Credit Default Swap (CDS) market. We first analyze the relationship between the log return series of the reference rates of a CDS contract and the hazard rate. This provides a theoretical foundation for the use of correlation of the log CDS returns as a representation of credit risk correlation. In the dynamic Bayesian linear modeling framework, we consider an algorithm that allow us to obtain dynamic Bayesian updates for the correlation among the reference rates of an underlying CDS contract. Data from the Japan CDS market is analyzed using the proposed methodology. An empirical analyses on the data segmented by different economic environments are carried out. Results indicate that the estimated implied default correlation captures market structure very well and provides useful information for credit risk management.

Long-term changes in fish mercury levels in the historically impacted English-Wabigoon River system (Canada)

The English-Wabigoon River system in Northwestern Ontario, Canada, was one of the most heavily mercury-contaminated waterways in the world due to historical discharges in the 1960s from a chlor-alkali plant. This study examines long-term (1970-2010) monitoring data to assess temporal trends in mercury contamination in Walleye, Northern Pike and Lake Whitefish, three species important for sport and subsistence fishing in this region, using dynamic linear modeling and piecewise regression. For all lakes and species, there is a significant decline (36-94%) in mercury concentrations through time; however, there is evidence that this decline is either slowing down or levelling off. Concentrations in the English-Wabigoon fish are elevated, and may still present a potential health risk to humans consuming fish from this system. Various biotic and abiotic factors are examined as possible explanations to slowing rates of decline in mercury concentrations observed in the mid-1980s.

Temporal PCB and mercury trends in Lake Erie fish communities: A dynamic linear modeling analysis

We performed dynamic linear modeling analysis on fish contaminant data collected from the Ontario Ministry of the Environment and Environment Canada to examine long-term trends of total mercury (THg) and polychlorinated biphenyls (PCBs) in Lake Erie. Several sport fish species (walleye, smallmouth bass, rainbow trout) with differences in their diet habits, food competition strategies and foraging patterns are characterized by weakly increasing trends of their THg levels in Lake Erie after the mid- or late 1990s. Similarly, our analysis shows that the decline rates of the PCB body burdens in white bass, smallmouth bass, freshwater drum and whitefish have slowed down or have switched to weakly increasing rates over the last decade. Our analysis also provides evidence that the rainbow trout and coho salmon PCB concentrations have been decreasing steadily but the associated rates were fairly weak. The systematic shifts in energy trophodynamics along with the food web alterations induced from the introduction of non-native species, the new role of the sediments as a net contaminant source, and the potentially significant fluxes from the atmosphere stand out as some of the hypotheses proposed to explain the limited Lake Erie response in recent years to the various contamination mitigation strategies.