Results Of Individual Models Research Articles

Ensemble physically based semi-distributed models for the rainfall-runoff process modeling in the data-scarce Katar catchment, Ethiopia

Abstract This study evaluates the performance of the soil and water assessment tool (SWAT), the hydrologiska byråns vattenbalansavdelning (HBV) and the hydrologic engineering center-hydrologic modeling system (HEC-HMS) for modeling rainfall-runoff in the data-scarce Katar catchment, Ethiopia. First, the rainfall-runoff process was simulated using the SWAT, HBV and HEC-HMS models individually. Second, simple average ensemble (SAE), weighted average ensemble (WAE) and neural network ensemble (NNE) techniques were developed by combining the results of individual models to improve overall accuracy. Statistical performance measures and flow duration curves (FDCs) were used to compare and evaluate the performance of the models. The results showed that the SWAT model outperformed the HBV and HEC-HMS models with the coefficient of determination (R2) and Nash–Sutcliffe efficiency (NSE) of 0.857 and 0.83 for calibration and 0.85 and 0.799 for validation, respectively. The ensemble result showed that NNE outperformed the SAE and WAE techniques, with NSE and R2 values of 0.924 and 0.925 for calibration and 0.896 and 0.904 for validation, respectively. The NNE technique improved the performance of SWAT, HBV and HEC-HMS by 12.14, 22.7 and 26.8%, respectively, in the validation phase. Overall, the results showed that ensemble modeling is a promising option for accurate modeling of the rainfall-runoff process.

Seismic Facies Segmentation Using Ensemble of Convolutional Neural Networks

The use of machine learning for seismic interpretation is a growing area of interest for researchers. Manual interpretation demands time and specialized effort. The use of machine learning model will expedite the process. The Convolutional Neural Networks (CNNs) are a class of deep learning algorithms used for images. In this paper, seismic facies segmentation using encoder-decoder architecture of CNNs is proposed. The proposed method filled the gap using a multimodel approach for seismic interpretation. The novelty of the model is that it is not limited to the current dataset and semantic segmentation models. The encoder-decoder architecture input and output size is the same, and it allows the labelling of each pixel of the image. Four models are trained on the open-sourced F3 block Netherlands dataset. Images of 128 × 128 were extracted from the data. Data augmentation is used in two of the models to increase the data size for better model learning. Results of individual models and their ensemble are compared. Ensemble is performed by taking the average of the probabilities of the classes obtained from the trained models. Ensemble gave the superior results. Seven classes are segmented with a global pixel accuracy (GPA) of 98.52%, mean class accuracy (MCA) of 96.88%, and mean intersection over union (MIoU) of 93.92%.

Improving Data and Prediction Quality of High-Throughput Perovskite Synthesis with Model Fusion.

Combinatorial fusion analysis (CFA) is an approach for combining multiple scoring systems using the rank-score characteristic function and cognitive diversity measure. One example is to combine diverse machine learning models to achieve better prediction quality. In this work, we apply CFA to the synthesis of metal halide perovskites containing organic ammonium cations via inverse temperature crystallization. Using a data set generated by high-throughput experimentation, four individual models (support vector machines, random forests, weighted logistic classifier, and gradient boosted trees) were developed. We characterize each of these scoring systems and explore 66 possible combinations of the models. When measured by the precision on predicting crystal formation, the majority of the combination models improves the individual model results. The best combination models outperform the best individual models by 3.9 percentage points in precision. In addition to improving prediction quality, we demonstrate how the fusion models can be used to identify mislabeled input data and address issues of data quality. In particular, we identify example cases where all single models and all fusion models do not give the correct prediction. Experimental replication of these syntheses reveals that these compositions are sensitive to modest temperature variations across the different locations of the heating element that can hinder or enhance the crystallization process. In summary, we demonstrate that model fusion using CFA can not only identify a previously unconsidered influence on reaction outcome but also be used as a form of quality control for high-throughput experimentation.

Analysis of CMIP6 atmospheric moisture fluxes and the implications for projections of future change in mean and heavy rainfall

AbstractMotivated by the wide range of projections of regional rainfall (or precipitation, pr) based on the CMIP5 ensemble of global climate models, we investigate the atmospheric moisture budget of idealized rising‐CO2 (1pctCO2) simulations from 10 models submitted to CMIP6. We use the new CMIP6 standard variables, the vertically integrated eastward and northward moisture flux, and provide a multi‐model assessment of these, their convergence (conv), along with water vapour path (prw). Seasonal climatologies of a 20‐year base climate for pr, prw, flux, and conv match well those from the ECMWF's Reanalysis 5. In addition, composites ‘H’ of quantities for months in the top decile of monthly rainfall, calculated at each model grid point, produce a field of heavy rainfall, on average double that of the mean. Spatial correlation coefficients (r) between the pr and conv fields are typically .7 in each season. This rises to .9 for the corresponding H fields, due to the important link between moisture convergence and heavy rainfall. Fields of standardized change (Δ), or change per degree of global warming, were calculated. The ensemble‐mean Δpr is only weakly correlated with the base climate pr, and likewise for conv. However, the prw change is highly correlated with the base prw, pointing to the importance of the “thermodynamic” change in that variable. Globally, the net Δconv is highly correlated with Δpr, around .9. This rises to .95 for the heavy‐rain composites. Relationships between changes in pr and other components hold at grid points also, when the individual model results are correlated. Clearly, there is a dependence in the change in pr with that in conv, particularly for heavy rainfall. Assessment of a larger ensemble from CMIP6, using the new flux variables, would improve the evaluation of these processes, and potentially allow more confident projections in future rainfall changes.

Predicting Arabic word reading: A cross-classified generalized random-effects analysis showing the critical role of morphology.

The distinctive features of the Arabic language and orthography offer opportunities to investigate multiple word characteristics at the item level. The aim of this paper was to model differences in word reading at the item level among 3rd grade native Arabic-speaking children (n = 303) using cross-classified generalized random-effects (CCGRE) analysis. The participants read 80 vowelized words that varied in multiple elements that may contribute to their decodability: number of letters, number of syllables, number of morphemes, ligaturing (connectivity), semantics (concrete vs. abstract), orthographic frequency, root type frequency, and part of speech. Morphological awareness (MA) was included as a person-level predictor. Results of individual models showed that MA, number of letters, number of syllables, number of morphemes, number of ligatures, orthographic frequency, and part of speech were significantly related to the probability of a correct response. However, when all predictors were entered simultaneously, only MA and number of morphemes remained significant. These results underscore the important role of morphology in the lexical structure of Arabic words and in Arabic word reading. Discussion focuses on the role of morphology in Arabic reading and the implications for intervention to improve word recognition in children learning to read Arabic.

Ensemble forecasts of air quality in eastern China – Part 2: Evaluation of the MarcoPolo–Panda prediction system, version 1

Abstract. An operational multimodel forecasting system for air quality has been developed to provide air quality services for urban areas of China. The initial forecasting system included seven state-of-the-art computational models developed and executed in Europe and China (CHIMERE, IFS, EMEP MSC-W, WRF-Chem-MPIM, WRF-Chem-SMS, LOTOS-EUROS, and SILAMtest). Several other models joined the prediction system recently, but are not considered in the present analysis. In addition to the individual models, a simple multimodel ensemble was constructed by deriving statistical quantities such as the median and the mean of the predicted concentrations. The prediction system provides daily forecasts and observational data of surface ozone, nitrogen dioxides, and particulate matter for the 37 largest urban agglomerations in China (population higher than 3 million in 2010). These individual forecasts as well as the multimodel ensemble predictions for the next 72 h are displayed as hourly outputs on a publicly accessible web site (http://www.marcopolo-panda.eu, last access: 27 March 2019). In this paper, the performance of the prediction system (individual models and the multimodel ensemble) for the first operational year (April 2016 until June 2017) has been analyzed through statistical indicators using the surface observational data reported at Chinese national monitoring stations. This evaluation aims to investigate (a) the seasonal behavior, (b) the geographical distribution, and (c) diurnal variations of the ensemble and model skills. Statistical indicators show that the ensemble product usually provides the best performance compared to the individual model forecasts. The ensemble product is robust even if occasionally some individual model results are missing. Overall, and in spite of some discrepancies, the air quality forecasting system is well suited for the prediction of air pollution events and has the ability to provide warning alerts (binary prediction) of air pollution events if bias corrections are applied to improve the ozone predictions.

Quantifying model uncertainty using Bayesian multi-model ensembles

Watershed models are essential tools to understand, quantify, and predict hydrologic processes and water quality responses from scales ranging from field to large river basins. However, the reliability of watershed models in a management context depends largely on inherent uncertainties in model predictions. The objective of this study is to quantify prediction uncertainty for flow, sediment, total nitrogen (TN), and total phosphorus (TP) resulting from model structure. We do this using using three process-based models: the Soil and Water Assessment Tool-Variable Source Area model (SWAT-VSA), the standard Soil and Water Assessment Tool (SWAT-ST), and the Chesapeake Bay Program’s Phase 6 Watershed Model (CBP-Model). We initialize each of the models using meteorological, soil, and land use data, and analyze outputs of flow, sediment, TN, and TP fluxes at the U.S. Geological Survey stream gauge at the downstream end of the Susquehanna River Basin in Conowingo, Maryland. Using these three models, we develop and compare two types of Bayesian models, a Bayesian Generalized (Non-) Linear Multilevel Model (BGMM), and a Bayesian Model Averaging (BMA) for flow, sediment, TN, and TP and 95% credible intervals. We compare the Bayesian models results against the individual model results, and straight model averaging (SMA) using a split time period analysis to assess their predictive strengths. Both Bayesian models provided substantially better predictions than the individual process-based models, and estimates of prediction uncertainty, which can enhance decision-making and improve watershed management by providing a risk based assessment of outcomes.

Atmospheric Modeling of 137Cs Plumes From the Fukushima Daiichi Nuclear Power Plant—Evaluation of the Model Intercomparison Data of the Science Council of Japan

AbstractSince the Fukushima Daiichi Nuclear Power Plant (FDNPP) accident in March 2011, atmospheric simulation models have improved our understanding of the atmospheric behavior of radionuclides. Model intercomparisons provide valuable and useful information for evaluating the validity and variability of individual model results. In this study, we compared results of seven atmospheric transport models used to simulate 137Cs released from the FDNPP to the atmosphere. All model results used in this analysis had been submitted for a model intercomparison project of the Science Council of Japan (2014, http://www.scj.go.jp/en/report/index.html). Here we assessed model performance by comparing model results with observed hourly atmospheric concentrations of 137Cs, with a particular focus on nine plumes over the Tohoku and Kanto regions. The intercomparison results showed that model performance in reproducing 137Cs concentrations was highly variable among different models and plumes. In general, models better reproduced plumes that passed over many observation stations. The performance among the models was consistent with the simulated wind fields and the source terms used. We also assessed model performance in relation to accumulated 137Cs deposition. Simulated areas of high 137Cs deposition were consistent with the simulated 137Cs plume pathways, though the models that best simulated atmospheric 137Cs concentrations were different from those that best simulated deposition. The ensemble mean of all models consistently reproduced atmospheric 137Cs concentrations and deposition well, suggesting that use of a multimodel ensemble results in more effective and consistent model performance.

Considerations for assessing model averaging of regression coefficients.

Model choice is usually an inevitable source of uncertainty in model-based statistical analyses. While the focus of model choice was traditionally on methods for choosing a single model, methods to formally account for multiple models within a single analysis are now accessible to many researchers. The specific technique of model averaging was developed to improve predictive ability by combining predictions from a set of models. However, it is now often used to average regression coefficients across multiple models with the ultimate goal of capturing a variable's overall effect. This use of model averaging implicitly assumes the same parameter exists across models so that averaging is sensible. While this assumption may initially seem tenable, regression coefficients associated with particular explanatory variables may not hold equivalent interpretations across all of the models in which they appear, making explanatory inference about covariates challenging. Accessibility to easily implementable software, concerns about being criticized for ignoring model uncertainty, and the chance to avoid having to justify choice of a final model have all led to the increasing popularity of model averaging in practice. We see a gap between the theoretical development of model averaging and its current use in practice, potentially leaving well-intentioned researchers with unclear inferences or difficulties justifying reasons for using (or not using) model averaging. We attempt to narrow this gap by revisiting some relevant foundations of regression modeling, suggesting more explicit notation and graphical tools, and discussing how individual model results arecombined to obtain a model averaged result. Our goal is to help researchers make informed decisions about model averaging and to encourage question-focused modeling over method-focused modeling.

Implications of freshwater flux data from the CMIP5 multimodel output across a set of Northern Hemisphere drainage basins

AbstractThe multimodel ensemble of the Coupled Model Intercomparison Project, Phase 5 (CMIP5) synthesizes the latest research in global climate modeling. The freshwater system on land, particularly runoff, has so far been of relatively low priority in global climate models, despite the societal and ecosystem importance of freshwater changes, and the science and policy needs for such model output on drainage basin scales. Here we investigate the implications of CMIP5 multimodel ensemble output data for the freshwater system across a set of drainage basins in the Northern Hemisphere. Results of individual models vary widely, with even ensemble mean results differing greatly from observations and implying unrealistic long‐term systematic changes in water storage and level within entire basins. The CMIP5 projections of basin‐scale freshwater fluxes differ considerably more from observations and among models for the warm temperate study basins than for the Arctic and cold temperate study basins. In general, the results call for concerted research efforts and model developments for improving the understanding and modeling of the freshwater system and its change drivers. Specifically, more attention to basin‐scale water flux analyses should be a priority for climate model development, and an important focus for relevant model‐based advice for adaptation to climate change.

GECC: Gene Expression Based Ensemble Classification of Colon Samples.

Gene expression deviates from its normal composition in case a patient has cancer. This variation can be used as an effective tool to find cancer. In this study, we propose a novel gene expressions based colon classification scheme (GECC) that exploits the variations in gene expressions for classifying colon gene samples into normal and malignant classes. Novelty of GECC is in two complementary ways. First, to cater overwhelmingly larger size of gene based data sets, various feature extraction strategies, like, chi-square, F-Score, principal component analysis (PCA) and minimum redundancy and maximum relevancy (mRMR) have been employed, which select discriminative genes amongst a set of genes. Second, a majority voting based ensemble of support vector machine (SVM) has been proposed to classify the given gene based samples. Previously, individual SVM models have been used for colon classification, however, their performance is limited. In this research study, we propose an SVM-ensemble based new approach for gene based classification of colon, wherein the individual SVM models are constructed through the learning of different SVM kernels, like, linear, polynomial, radial basis function (RBF), and sigmoid. The predicted results of individual models are combined through majority voting. In this way, the combined decision space becomes more discriminative. The proposed technique has been tested on four colon, and several other binary-class gene expression data sets, and improved performance has been achieved compared to previously reported gene based colon cancer detection techniques. The computational time required for the training and testing of 208 × 5,851 data set has been 591.01 and 0.019 s, respectively.

Evolution of the hydro-climate system in the Lake Baikal basin

Summary Climatic changes can profoundly alter hydrological conditions in river basins. Lake Baikal is the deepest and largest freshwater reservoir on Earth, and has a unique ecosystem with numerous endemic animal and plant species. We here identify long-term historical (1938–2009) and projected future hydro-climatic trends in the Selenga River Basin, which is the largest sub-basin (>60% inflow) of Lake Baikal. Our analysis is based on long-term river monitoring and historical hydro-climatic observation data, as well as ensemble mean and 22 individual model results of the Coupled Model Intercomparison Project, Phase 5 (CMIP5). Study of the latter considers a historical period (from 1961) and projections for 2010–2039 and 2070–2099. Observations show almost twice as fast warming as the global average during the period 1938–2009. Decreased intra-annual variability of river discharge over this period indicates basin-scale permafrost degradation. CMIP5 ensemble projections show further future warming, implying continued permafrost thaw. Modelling of runoff change, however, is highly uncertain, with many models (64%) and their ensemble mean failing to reproduce historical behaviour, and with indicated future increase being small relative to the large differences among individual model results.

The Value of Automobile Surveys in Reducing the Buyer’s Risk

Consumers use quality surveys to reduce risk in purchasing a car. The J. D. Power's initial quality survey of automobile buyers reports the number of problems found in the first 90 days of ownership. It is important that this information be clear and concise for the public to understand. In 2006, the surveys were changed to combine two measures of quality, number of defects and design problems, into a single number. In prior years, only the number of defects was used. The survey design change combined with issues related to the methodology in aggregating values leads to confusion and inconsistencies over time as the difference between cars becomes insignificant. The conclusion recommends changes to correct these problems. INTRODUCTION Quality survey companies follow the basic information system's process of collecting the data from the assessment documents, creating a database of the information, analyzing the data, and disseminating the probability of a defect or the number of problems per product (O'Brien & Marakas, 2009). This paper examines the information process of a survey company and identifies specific problems that may lead to incorrect conclusions. These issues result from the design of survey, computational errors, and lack of continuity in the values over time. The example used in this analysis is taken from the automobile industry. Each year, numerous automobile quality surveys are reported by several information services. Driven by the results of these studies, car manufactures have substantially improved vehicle quality. The results from the J. D. Power's Initial Quality Surveys (Tews & Perryman, 2008) provide consumers with information on the awards for best car in a segment. These awards are used extensively in advertising and influence purchasing decisions. Before purchasing a car, consumers review the available literature on the quality of a product to minimize their risk of purchasing a defective car. Collecting all this information reduces the risk of making a decision with partial or imperfect information (Heiser & Render, 2008). The analysis of the survey process focuses on correctly calculating values and the consistency in the final results. First, the J.D. Power's methodology that is used to aggregate the results of individual models into nameplates, manufacturers and country is examined. Next, with the number of problems becoming statistically insignificant, a qualitative measure is added to the number of defects, a quantitative measure, resulting in a confusing and inconsistent set of values for the new surveys. Throughout the paper, suggestions for improvement are made. The conclusion recommends changes in the survey, a new methodology for aggregating data, and separate surveys for qualitative and quantitative measures to improve the information content

Comparison of eight published static finite element models of the intact lumbar spine: Predictive power of models improves when combined together

Finite element (FE) model studies have made important contributions to our understanding of functional biomechanics of the lumbar spine. However, if a model is used to answer clinical and biomechanical questions over a certain population, their inherently large inter-subject variability has to be considered. Current FE model studies, however, generally account only for a single distinct spinal geometry with one set of material properties. This raises questions concerning their predictive power, their range of results and on their agreement with in vitro and in vivo values.Eight well-established FE models of the lumbar spine (L1-5) of different research centers around the globe were subjected to pure and combined loading modes and compared to in vitro and in vivo measurements for intervertebral rotations, disc pressures and facet joint forces.Under pure moment loading, the predicted L1-5 rotations of almost all models fell within the reported in vitro ranges, and their median values differed on average by only 2° for flexion-extension, 1° for lateral bending and 5° for axial rotation. Predicted median facet joint forces and disc pressures were also in good agreement with published median in vitro values. However, the ranges of predictions were larger and exceeded those reported in vitro, especially for the facet joint forces. For all combined loading modes, except for flexion, predicted median segmental intervertebral rotations and disc pressures were in good agreement with measured in vivo values.In light of high inter-subject variability, the generalization of results of a single model to a population remains a concern. This study demonstrated that the pooled median of individual model results, similar to a probabilistic approach, can be used as an improved predictive tool in order to estimate the response of the lumbar spine.

Southeastern U.S. Rainfall Prediction in the North American Multi-Model Ensemble

Abstract The present study investigates the predictive skill of the North American Multi-Model Ensemble (NMME) system for intraseasonal-to-interannual (ISI) prediction with focus on southeastern U.S. precipitation. The southeastern United States is of particular interest because of the typically short-lived nature of above- and below-normal extended rainfall events allowing for focus on seasonal prediction, as well as the tendency for more predictability in the winter months. Included in this study is analysis of the forecast quality of the NMME system when predicting above- and below-normal rainfall and individual rainfall events, with particular emphasis on results from the 2007 dry period. Both deterministic and probabilistic measures of skill are utilized in order to gain a more complete understanding of how accurately the system predicts precipitation at both short and long lead times and to investigate the multimodel aspect of the system as compared to using an individual predictive model. The NMME system consistently shows low systematic error and relatively high skill in predicting precipitation, particularly in winter months as compared to individual model results.

An Idiographic Examination of Day-to-Day Patterns of Substance Use Craving, Negative Affect, and Tobacco Use Among Young Adults in Recovery

Psychological constructs, such as negative affect and substance use cravings that closely predict relapse, show substantial intraindividual day-to-day variability. This intraindividual variability of relevant psychological states combined with the “one day at a time” nature of sustained abstinence warrant a day-to-day investigation of substance use recovery. This study examines day-to-day associations among substance use cravings, negative affect, and tobacco use among 30 college students in 12-step recovery from drug and alcohol addictions. To account for individual variability in day-to-day process, it applies an idiographic approach. The sample of 20 males and 10 females (mean age = 21) was drawn from members of a collegiate recovery community at a large university. Data were collected with end-of-day data collections taking place over an average of 26.7 days. First-order vector autoregression models were fit to each individual predicting daily levels of substance use cravings, negative affect, and tobacco use from the same 3 variables 1 day prior. Individual model results demonstrated substantial interindividual differences in intraindividual recovery process. Based on estimates from individual models, cluster analyses were used to group individuals into 2 homogeneous subgroups. Group comparisons demonstrate distinct patterns in the day-to-day associations among substance use cravings, negative affect, and tobacco use, suggesting the importance of idiographic approaches to recovery management and that the potential value of focusing on negative affect or tobacco use as prevention targets depends on idiosyncratic processes.

The analysis of the constitutive model effect on convergence of results with the experiment for FEM modeling of 3D issues with high elastic-plastic deformations

This paper presents the results of experimental researches and FEM modeling for tension process of an original plane specimen with notch. The specimen was made of the low carbon steel sheet with thickness of 4 mm. The modeling was performed with three different constitutive relationships, which are available in commercial software for modeling the non-linear issues with high strains. The three-dimensional numerical model was based on the experimental model. The results of individual models were compared with the experimental results. The special attention was paid to convergence of force characteristics and modeled specimen shapes in areas of the highest elastic-plastic deformations. Moreover, for the model of the highest convergence with the experiment the ductile fracture phase was modeled until total separation of specimen material.

Simulation of the time-variable gravity field by means of coupled geophysical models

Abstract. Time variable gravity fields, reflecting variations of mass distribution in the system Earth is one of the key parameters to understand the changing Earth. Mass variations are caused either by redistribution of mass in, on or above the Earth's surface or by geophysical processes in the Earth's interior. The first set of observations of monthly variations of the Earth gravity field was provided by the US/German GRACE satellite mission beginning in 2002. This mission is still providing valuable information to the science community. However, as GRACE has outlived its expected lifetime, the geoscience community is currently seeking successor missions in order to maintain the long time series of climate change that was begun by GRACE. Several studies on science requirements and technical feasibility have been conducted in the recent years. These studies required a realistic model of the time variable gravity field in order to perform simulation studies on sensitivity of satellites and their instrumentation. This was the primary reason for the European Space Agency (ESA) to initiate a study on ''Monitoring and Modelling individual Sources of Mass Distribution and Transport in the Earth System by Means of Satellites''. The goal of this interdisciplinary study was to create as realistic as possible simulated time variable gravity fields based on coupled geophysical models, which could be used in the simulation processes in a controlled environment. For this purpose global atmosphere, ocean, continental hydrology and ice models were used. The coupling was performed by using consistent forcing throughout the models and by including water flow between the different domains of the Earth system. In addition gravity field changes due to solid Earth processes like continuous glacial isostatic adjustment (GIA) and a sudden earthquake with co-seismic and post-seismic signals were modelled. All individual model results were combined and converted to gravity field spherical harmonic series, which is the quantity commonly used to describe the Earth's global gravity field. The result of this study is a twelve-year time-series of 6-hourly time variable gravity field spherical harmonics up to degree and order 180 corresponding to a global spatial resolution of 1 degree in latitude and longitude. In this paper, we outline the input data sets and the process of combining these data sets into a coherent model of temporal gravity field changes. The resulting time series was used in some follow-on studies and is available to anybody interested.

The influence of foreign vs. North American emissions on surface ozone in the US

Abstract. As part of the Hemispheric Transport of Air Pollution (HTAP; http:// www.htap.org) project, we analyze results from 15 global and 1 hemispheric chemical transport models and compare these to Clean Air Status and Trends Network (CASTNet) observations in the United States (US) for 2001. Using the policy-relevant maximum daily 8-h average ozone (MDA8 O3) statistic, the multi-model ensemble represents the observations well (mean r2=0.57, ensemble bias = +4.1 ppbv for all US regions and all seasons) despite a wide range in the individual model results. Correlations are strongest in the northeastern US during spring and fall (r2=0.68); and weakest in the midwestern US in summer (r2=0.46). However, large positive mean biases exist during summer for all eastern US regions, ranging from 10–20 ppbv, and a smaller negative bias is present in the western US during spring (~3 ppbv). In nearly all other regions and seasons, the biases of the model ensemble simulations are ≤5 ppbv. Sensitivity simulations in which anthropogenic O3-precursor emissions (NOx + NMVOC + CO + aerosols) were decreased by 20% in four source regions: East Asia (EA), South Asia (SA), Europe (EU) and North America (NA) show that the greatest response of MDA8 O3 to the summed foreign emissions reductions occurs during spring in the West (0.9 ppbv reduction due to 20% emissions reductions from EA + SA + EU). East Asia is the largest contributor to MDA8 O3 at all ranges of the O3 distribution for most regions (typically ~0.45 ppbv) followed closely by Europe. The exception is in the northeastern US where emissions reductions in EU had a slightly greater influence than EA emissions, particularly in the middle of the MDA8 O3 distribution (response of ~0.35 ppbv between 35–55 ppbv). EA and EU influences are both far greater (about 4x) than that from SA in all regions and seasons. In all regions and seasons O3-precursor emissions reductions of 20% in the NA source region decrease MDA8 O3 the most – by a factor of 2 to nearly 10 relative to foreign emissions reductions. The O3 response to anthropogenic NA emissions is greatest in the eastern US during summer at the high end of the O3 distribution (5–6 ppbv for 20% reductions). While the impact of foreign emissions on surface O3 in the US is not negligible – and is of increasing concern given the recent growth in Asian emissions – domestic emissions reductions remain a far more effective means of decreasing MDA8 O3 values, particularly those above 75 ppb (the current US standard).

Wind onset and withdrawal of Asian summer monsoon and their simulated performance in AMIP models

This study defines the concepts of wind onset and wind withdrawal to describe the abrupt seasonal variations of wind direction and circulation of the Asian monsoon. The patterns of wind onset and withdrawal show that the earliest wind onset in the tropical monsoon regions is found over equator around 70°–100°E and the southernmost South China Sea (SCS) and western Kalimantan, and the wind withdrawal shows a southward progression in tropics compared to the wind onset. A notable temporal boundary is found around 25°N in the subtropical western North Pacific (WNP), which may be related to the northward advance and southward retreat of the western Pacific subtropical high. The angle amplitudes of wind vectors in wind onset and withdrawal have distinct regional differences in Asian monsoon regions. Since the process of monsoon onset (withdrawal) may include several onsets of different variables without simultaneity, the relationships of the wind onset and withdrawal with the abrupt change of other variables (e.g. reversal of zonal wind, reversal of meridional wind, outgoing longwave radiation (OLR), precipitation) are investigated. The results indicate that the temporal discrepancies in different monsoon regions confirmed the asynchronous onsets. It also implies that the wind onset might be a good omen for monsoon precipitation in most regions since it is slightly earlier than rainy season onset. Seven Atmospheric Model Intercomparison Project (AMIP) models from Intergovernmental Panel on Climate Change (IPCC) Fourth Assessment Report (AR4) are validated against observations mentioned above. Generally, the simulations of the multi-model ensemble mean are better than any individual model results. And the simulations of wind withdrawal are better than those of wind onset. For wind onset, IAP-FGOALS-1.0g, MIROC3.2 (medres) and MPI-ECHAM5 simulate reasonably well. For wind retreat, most models can capture the behaviors in tropics. However, there are still some discrepancies in a few models to simulate the dates of sudden change of monsoon wind direction. Moreover, most of models cannot reproduce the onset and withdrawal of both rainfall and OLR. The relationship between these discrepancies and the shortcomings of precipitation simulation is crucial for further investigating in the future.