Search

The rapid expansion of Earth system model (ESM) data available from the Coupled Model Intercomparison Project Phase 6 (CMIP6) necessitates new methods to evaluate the performance and suitability of ESMs used for hydroclimate applications as these extremely large data volumes complicate stakeholder efforts to use new ESM outputs in updated climate vulnerability and impact assessments. We develop an analysis framework to inform ESM sub‐selection based on process‐oriented considerations and demonstrate its performance for a regional application in the US Pacific Northwest. First, a suite of global and regional metrics is calculated, using multiple historical observation datasets to assess ESM performance. These metrics are then used to rank CMIP6 models, and a culled ensemble of models is selected using a trend‐related diagnostics approach. This culling strategy does not dramatically change climate scenario trend projections in this region, despite retaining only 20% of the CMIP6 ESMs in the final model ensemble. The reliability of the culled trend projection envelope and model response similarity is also assessed using a perfect model framework. The absolute difference in temperature trend projections is reduced relative to the full ensemble compared to the model for each SSP scenario, while precipitation trend errors are largely unaffected. In addition, we find that the spread of the culled ensemble temperature and precipitation trends includes the trend of the "truth" model ∼83%‐92% of the time. This analysis demonstrates a reliable method to reduce ESM ensemble size that can ease use of ESMs for creating and understanding climate vulnerability and impact assessments. Plain Language Summary: This study provides an updated and rigorously tested method for evaluating the performance of climate models for applications relevant to water managers and other stakeholders. Using traditional metrics of climate model performance, both regional and global, as well as newly developed metrics based on processes important for the simulation of precipitation, we have created a generalizable, systematic, and succinct method for reducing the number of models to be considered for climate change impact applications. By reducing the number of relevant models to around 20% of the total models and not having a significant impact on future temperature and precipitation trend projections in doing so, we strongly reduce the computational effort needed to gain a realistic simulation of the future climate for a given regional impact. This method is tested with a variety of statistical tests and found to be reliable for our application over the Pacific Northwest United States. Key Points: A method for evaluation of climate models over the PNW using a combination of global and regional metrics has been developedThis allows for a reduced envelope of ESMs for impact applications without significantly affecting the future trend projections [ABSTRACT FROM AUTHOR]

Short-term probabilistic forecasts of the trajectory of the COVID-19 pandemic in the United States have served as a visible and important communication channel between the scientific modeling community and both the general public and decision-makers. Forecasting models provide specific, quantitative, and evaluable predictions that inform short-term decisions such as healthcare staffing needs, school closures, and allocation of medical supplies. Starting in April 2020, the US COVID-19 Forecast Hub (https://covid19forecasthub.org/) collected, disseminated, and synthesized tens of millions of specific predictions from more than 90 different academic, industry, and independent research groups. A multimodel ensemble forecast that combined predictions from dozens of groups every week provided the most consistently accurate probabilistic forecasts of incident deaths due to COVID-19 at the state and national level from April 2020 through October 2021. The performance of 27 individual models that submitted complete forecasts of COVID-19 deaths consistently throughout this year showed high variability in forecast skill across time, geospatial units, and forecast horizons. Two-thirds of the models evaluated showed better accuracy than a naïve baseline model. Forecast accuracy degraded as models made predictions further into the future, with probabilistic error at a 20-wk horizon three to five times larger than when predicting at a 1-wk horizon. This project underscores the role that collaboration and active coordination between governmental public-health agencies, academic modeling teams, and industry partners can play in developing modern modeling capabilities to support local, state, and federal response to outbreaks. [ABSTRACT FROM AUTHOR]

Aim: Citizen science data are increasingly used for modelling species distributions because they offer broad spatiotemporal coverage of local observations. However, such data are often collected without experimental design or set survey methods, raising the risk that bias and noise will compromise modelled predictions. We tested the ability of species distribution models (SDMs) built from these low‐structure citizen science data to match the quality of SDMs from systematically collected data and tested whether stringent data filtering improved predictions. Location: Northeastern USA. Methods: We evaluated models built from a rapidly growing dataset of avian occurrences reported by birders—eBird—against models built from four independent, systematically collected datasets. We developed SDMs for 96 species using both data sources and compared their predictive abilities. We also tested whether culling eBird data by applying stringent data filters on survey effort or observer expertise improved predictions. Results: We found that SDMs built from low‐structure citizen science data matched or exceeded performance of SDMs from systematically collected datasets for 12%–31% of species (x¯ = 22%), depending on the dataset. At least one culling option produced equivalent or better performance for 40%–70% of species (x¯ = 49%). Data culling by restricting survey effort improved predictions more than restricting by observer expertise. The optimal effort restriction differed by dataset, and for three of the datasets was further informed by species traits. Main conclusions: Species distribution models developed using low‐structure citizen science data sometimes performed as well as those from systematic data. Culling generally improved models, but results were heterogeneous, prohibiting clear recommendations for how to cull. Our results indicate that the growing availability of citizen science data holds potential for creating high‐quality spatial predictions, but that time should be invested in determining how best to cull datasets and that one‐size‐fits‐all solutions beyond basic outlier filtering may be hard to find. [ABSTRACT FROM AUTHOR]

• Model captured seasonal and regional variability of zooplankton observation dataset. • Modelled zooplankton biomass was highest in regions adjacent to areas of high mixing. • Zooplankton grazing was high in mixed regions despite low primary productivity. We used a three-dimensional coupled biophysical model to examine zooplankton dynamics in the Salish Sea, NE Pacific. First, we evaluated the two zooplankton classes of the SalishSeaCast model using a transboundary zooplankton dataset comprised of observation data from the Canadian and United States waters of the Salish Sea from 2015 to 2019. Model zooplankton classes correspond to micro- and meso-zooplankton whose biomass is tightly coupled to phytoplankton through modelled food web dynamics (Z1) and mesozooplankton with life cycle-based seasonal grazing impacts (Z2). Overall, the model effectively captured seasonal patterns in observed biomass, although with slightly higher biomass estimates for both Z1 and Z2 (Bias = 0.10 and 0.08 g C m−2, respectively). Model fit varied regionally, with a weaker model fit being observed in nearshore regions. In addition, an autumn peak in Z2 was observed in the model, but not in the observations, suggesting some seasonal variations in model fit. Following the model evaluation, we used the model to determine seasonal and regional patterns of zooplankton grazing. Seasonally, the main peak in modelled zooplankton biomass increased in April or May in most of the regions defined within the Salish Sea and was driven by grazing on diatoms. Regionally, depth-integrated zooplankton biomass was consistently highest in areas adjacent to regions of strong tidal mixing. In addition, model-based zooplankton grazing was highest in the tidally mixed regions where phytoplankton biomass was high due to advection into the region despite low primary productivity. Our model-based results provide an opportunity to examine bottom-up food web processes at spatio-temporal scales not achievable with in situ sampling and help to elucidate key drivers of lower trophic level dynamics within the Salish Sea. [ABSTRACT FROM AUTHOR]

Regional climate models (RCMs) can simulate heavy precipitation more accurately than general circulation models (GCMs) through more realistic representation of topography and mesoscale processes. Analogue methods of downscaling, which identify the large-scale atmospheric conditions associated with heavy precipitation, can also produce more accurate and precise heavy precipitation frequency in GCMs than the simulated precipitation. In this study, we examine the performances of the analogue method versus direct simulation, when applied to RCM and GCM simulations, in detecting present-day and future changes in summer (JJA) heavy precipitation over the Midwestern United States. We find analogue methods are comparable to MERRA-2 and its bias-corrected precipitation in characterizing the occurrence and interannual variations of observed heavy precipitation events, all significantly improving upon MERRA precipitation. For the late twentieth-century heavy precipitation frequency, RCM precipitation improves upon the corresponding driving GCM with greater accuracy yet comparable inter-model discrepancies, while both RCM- and GCM-based analogue results outperform their model-simulated precipitation counterparts in terms of accuracy and model consensus. For the projected trends in heavy precipitation frequency through the mid twenty-first century, analogue method also manifests its superiority to direct simulation with reduced intermodel disparities, while the RCM-based analogue and simulated precipitation do not demonstrate a salient improvement (in model consensus) over the GCM-based assessment. However, a number of caveats preclude any overall judgement, and further work—over any region of interest—should include a larger sample of GCMs and RCMs as well as ensemble simulations to comprehensively account for internal variability. [ABSTRACT FROM AUTHOR]

Dynamic evaluation of the fully coupled Weather Research and Forecasting (WRF)– Community Multi-scale Air Quality (CMAQ) model ozone simulations over the contiguous United States (CONUS) using two decades of simulations covering the period from 1990 to 2010 is conducted to assess how well the changes in observed ozone air quality are simulated by the model. The changes induced by variations in meteorology and/or emissions are also evaluated during the same timeframe using spectral decomposition of observed and modeled ozone time series with the aim of identifying the underlying forcing mechanisms that control ozone exceedances and making informed recommendations for the optimal use of regional-scale air quality models. The evaluation is focused on the warm season's (i.e., May–September) daily maximum 8-hr (DM8HR) ozone concentrations, the 4th highest (4th) and average of top 10 DM8HR ozone values (top10), as well as the spectrally-decomposed components of the DM8HR ozone time series using the Kolmogorov-Zurbenko (KZ) filter. Results of the dynamic evaluation are presented for six regions in the U.S., consistent with the National Oceanic and Atmospheric Administration (NOAA) climatic regions. During the earlier 11-yr period (1990–2000), the simulated and observed regional average trends are not statistically significant. During the more recent 2000–2010 period, all observed trends are statistically significant and WRF-CMAQ captures the observed downward trend in the Southwest and Midwest but under-predicts the downward trends in observations for the other regions. Observational analysis reveals that it is the magnitude of the long-term forcing that dictates the maximum ozone exceedance potential; there is a strong linear relationship between the long-term forcing and the 4th highest or the average of the top10 ozone concentrations in both observations and model output. This finding indicates that improving the model's ability to reproduce the long-term component will also enable better simulation of ozone extreme values that are of interest to regulatory agencies. [ABSTRACT FROM AUTHOR]

A new, high-resolution (4 km), gridded land surface dataset produced with the Land Information System (LIS) is introduced, and the first set of synthesis of key hydroclimatic variables is reported. The dataset is produced over a 33-yr time period (1980-2012) for the U.S. Midwest with the intent to aid the agricultural community in understanding hydroclimatic impacts on crop production and decision-making in operational practices. While approximately 20 hydroclimatic variables are available through the LIS dataset, the focus here is on soil water content, soil temperature, and evapotranspiration. To assess the performance of the model, the LIS dataset is compared with in situ hydrometeorological observations across the study domain and with coarse-resolution reanalysis products [NARR, MERRA, and NLDAS-2 (phase 2 of the North American Land Data Assimilation System)]. In agricultural regions such as the U.S. Midwest, finescale hydroclimatic mapping that links the regional scale to the field scale is necessary. The new dataset provides this link as an intermediate-scale product that links point observations and coarse gridded datasets. In general, the LIS dataset compares well with in situ observations and coarser gridded products in terms of both temporal and spatial patterns, but cases of strong disagreement exist particularly in areas with sandy soils. The dataset is made available to the broader research community as an effort to fill the gap in spatial hydroclimatic data availability. [ABSTRACT FROM AUTHOR]

Previously developed models for predicting absolute risk of invasive epithelial ovarian cancer have included a limited number of risk factors and have had low discriminatory power (area under the receiver operating characteristic curve (AUC) <0.60). Because of this, we developed and internally validated a relative risk prediction model that incorporates 17 established epidemiologic risk factors and 17 genome-wide significant single nucleotide polymorphisms (SNPs) using data from 11 case-control studies in the United States (5,793 cases; 9,512 controls) from the Ovarian Cancer Association Consortium (data accrued from 1992 to 2010). We developed a hierarchical logistic regression model for predicting case-control status that included imputation of missing data. We randomly divided the data into an 80% training sample and used the remaining 20% for model evaluation. The AUC for the full model was 0.664. A reduced model without SNPs performed similarly (AUC = 0.649). Both models performed better than a baseline model that included age and study site only (AUC = 0.563). The best predictive power was obtained in the full model among women younger than 50 years of age (AUC = 0.714); however, the addition of SNPs increased the AUC the most for women older than 50 years of age (AUC = 0.638 vs. 0.616). Adapting this improved model to estimate absolute risk and evaluating it in prospective data sets is warranted. [ABSTRACT FROM AUTHOR]

Land transformation from grassland to cropland in the Northern Great Plains (NGP) has become a growing concern among many stakeholders. A growing body of work has sought to determine the amount and rate of land use change with less emphasis on the systemic structures or feedback processes of land use decisions. This paper presents the development of a system dynamics simulation model to integrate ecological, economic, and social components influencing land use decisions, including cattle ranching, cropland production, rural communities, land quality, and public policies. Evaluation indicated that the model satisfactorily predicted historical land, agricultural commodity, and rural community data from the model structure. Reference modes for key variables, including the farmland area, were characterized by a bias correction of 0.999, root mean squared error of prediction of 0.053, R2 of 0.921, and concordance correlation coefficient of 0.0959. The model was robust under extreme and varying sensitivity tests, as well as adequately predicting land use under changing system context. The model's major contributions were the inclusion of decision-making feedbacks from economic and social signals with connectivity to land quality and elasticity values that drive land transformation. Limitations include lack of spatial input and output capabilities useful for visual interfacing. [ABSTRACT FROM AUTHOR]

Case management is a $6 billion industry that employs over 34,000 people in the United States alone. Traditionally, case management has been utilized to help patients navigate the health care system and to coordinate care in the hope of lowering costs and achieving better health outcomes. However, since enrollment into these programs is typically either universal or limited to very sick patients, studies on the cost-effectiveness of case management programs find that their performance is mixed, at best. In this article we posit an opportunity to improve outcomes and lower costs by targeting certain patients for case management and early intervention. Utilizing modern data mining methods, we develop a methodology to identify these patients, who we describe as “jumpers” because their costs are currently low but will increase significantly in the near future. Given the performance of the prediction models, we also show that unless case management can prevent over 7.5% of health care cost increases, it may benefit enrolled members but will not reduce overall costs. The article then introduces a performance bounding methodology that characterizes the best obtainable prediction accuracy on a given data set. The derived upper bound demonstrates that searching for jumpers presents a far more challenging prediction problem than identifying future high-cost members, which is the traditional approach to selecting case management candidates. [ABSTRACT FROM AUTHOR]

Abstract: Changes in climate and emissions will affect future air quality. In this work, simulations of regional air quality during current (2001–2005) and future (2026–2030) winter and summer are conducted with the newly released CMAQ version 5.0 to examine the impacts of simulated future climate and anthropogenic emission projections on air quality over the U.S. Current meteorological and chemical predictions are evaluated against observations to assess the model's capability in reproducing the seasonal differences. WRF and CMAQ capture the overall observational spatial patterns and seasonal differences. Biases in model predictions are attributed to uncertainties in emissions, boundary conditions, and limitations in model physical and chemical treatments as well as the use of a coarse grid resolution. Increased temperatures (up to 3.18 °C) and decreased ventilation (up to 157 m in planetary boundary layer height) are found in both future winter and summer, with more prominent changes in winter. Increases in future temperatures result in increased isoprene and terpene emissions in winter and summer, driving the increase in maximum 8-h average O3 (up to 5.0 ppb) over the eastern U.S. in winter while decreases in NO x emissions drive the decrease in O3 over most of the U.S. in summer. Future PM2.5 concentrations in winter and summer and many of its components decrease due to decreases in primary anthropogenic emissions and the concentrations of secondary anthropogenic pollutants as well as increased precipitation in winter. Future winter and summer dry and wet deposition fluxes are spatially variable and increase with decreasing surface resistance and precipitation, respectively. They decrease with a decrease in ambient particulate concentrations. Anthropogenic emissions play a more important role in summer than in winter for future O3 and PM2.5 levels, with a dominance of the effects of significant emission reductions over those of climate change on future PM2.5 levels. [Copyright &y& Elsevier]

Spatial modeling over broad scales can potentially direct conservation efforts to areas with high species-specific abundances. We examined the performance of regional models for predicting bird abundance at spatial scales typically addressed in conservation planning. Specifically, we used point count data on wood thrush ( Hylocichla mustelina) and blue-winged warbler ( Vermivora cyanoptera) from 2 time periods (1995-1998 and 2006-2007) to evaluate the ability of regional models derived via Bayesian hierarchical techniques to predict bird abundance. We developed models for each species within Bird Conservation Region (BCR) 23 in the upper midwestern United States at 800-ha, 8,000-ha, and approximately 80,000-ha scales. We obtained count data from the Breeding Bird Survey and land cover data from the National Land Cover Dataset (1992). We evaluated predictions from the best models, as defined by an information-theoretic criterion, using point count data collected within an ecological subregion of BCR 23 at 131 count stations in the 1990s and again in 2006-2007. Competing (Deviance Information Criteria <5) blue-winged warbler models accounted for 67% of the variability and suggested positive associations with forest edge and proportion of forest at the 8,000-ha scale, and negative associations with forest patch area (800 ha) and wetness (800 ha and 80,000 ha). The regional model performed best for blue-winged warbler predicted abundances from point counts conducted in Iowa during 1995-1996 ( rs = 0.57; P = 0.14), the survey period that most closely aligned with the time period of data used for regional model construction. Wood thrush models exhibited positive correlations with point count data for all survey areas and years combined ( rs = 0.58, P ≤ 0.001). In comparison, blue-winged warbler models performed worse as time increased between the point count surveys and vintage of the model building data ( rs = 0.03, P = 0.92 for Iowa and rs = 0.13, P = 0.51 for all areas, 2006-2007), likely related to the ephemeral nature of their preferred early successional habitat. Species abundance and sensitivity to changing habitat conditions seems to be an important factor in determining the predictive ability of regional models. Hierarchical models can be a useful tool for concentrating efforts at the scale of management units and should be one of many tools used by land managers, but we caution that the utility of such models may decrease over time for species preferring relatively ephemeral habitats if model inputs are not updated accordingly. © 2012 The Wildlife Society. [ABSTRACT FROM AUTHOR]

Abstract: We use long-term, coincident O3 and temperature measurements at the regionally representative US Environmental Protection Agency Clean Air Status and Trends Network (CASTNet) over the eastern US from 1988 through 2009 to characterize the surface O3 response to year-to-year fluctuations in weather, for the purpose of evaluating global chemistry-climate models. We first produce a monthly climatology for each site over all available years, defined as the slope of the best-fit line (m O3-T) between monthly average values of maximum daily 8-hour average (MDA8) O3 and monthly average values of daily maximum surface temperature (T max). Applying two distinct statistical approaches to aggregate the site-specific measurements to the regional scale, we find that summer time m O3-T is 3–6ppbK−1 (r =0.5–0.8) over the Northeast, 3–4ppbK−1 (r =0.5–0.9) over the Great Lakes, and 3–6ppbK−1 (r =0.2–0.8) over the Mid-Atlantic. The Geophysical Fluid Dynamics Laboratory (GFDL) Atmospheric Model version 3 (AM3) global chemistry-climate model generally captures the seasonal variations in correlation coefficients and m O3-T despite biases in both monthly mean summertime MDA8 O3 (up to +10 to +30ppb) and daily T max (up to +5K) over the eastern US. During summer, GFDL AM3 reproduces m O3-T over the Northeast (m O3-T =2–6ppbK−1; r =0.6–0.9), but underestimates m O3-T by 4ppbK−1 over the Mid-Atlantic, in part due to excessively warm temperatures above which O3 production saturates in the model. Combining T max biases in GFDL AM3 with an observation-based m O3-T estimate of 3ppbK−1implies that temperature biases could explain up to 5–15ppb of the MDA8 O3 bias in August and September though correcting for excessively cool temperatures would worsen the O3 bias in June. We underscore the need for long-term, coincident measurements of air pollution and meteorological variables to develop process-level constraints for evaluating chemistry-climate models used to project air quality responses to climate change. [Copyright &y& Elsevier]

Abstract: We evaluate the simulations of SO2 and sulfate using the Community Multiscale Air Quality model (CMAQ) version 4.6 with the observations over the United States in 2002. MM5 was used for meteorological simulations. While the general seasonal cycles of SO2 and sulfate are simulated well by the model, we find significant systematic biases in the summer. The model low bias in sulfate is considerably more severe than the model bias in SO2. Both ACM and RADM schemes are used in the model to test the sensitivities of simulated sulfate to cloud processing. We carry out detailed modeling analysis and diagnostics for July 2002. Compared to satellite observations of cloud liquid water path, CMAQ cloud modules greatly overestimates convective (sub-grid) precipitating clouds, leading to large overestimation of sulfate wet scavenging. Limiting convective precipitating cloud fraction in the cloud modules to <10% and hence significantly reducing wet scavenging lead to much improved agreement between simulated and observed sulfate. The average lifetime of sulfate in the model increases from 1–2 days to 3–4 days for July. We show that a potential model problem of excessive wet scavenging of sulfate does not necessarily lead to apparent problems in model simulations of sulfate wet deposition rate compared to surface observations. In general, there is still a lack of direct observational constraints from air quality monitoring measurements on model simulated cloud processing of SO2 and sulfate. [Copyright &y& Elsevier]

The energy sector comprises approximately two-thirds of global total greenhouse gas emissions. For this and other reasons, renewable energy resources including wind power are being increasingly harnessed to provide electricity generation potential with negligible emissions of carbon dioxide. The wind energy resource is naturally a function of the climate system because the "fuel" is the incident wind speed and thus is determined by the atmospheric circulation. Some recent articles have reported historical declines in measured near-surface wind speeds, leading some to question the continued viability of the wind energy industry. Here we briefly articulate the challenges inherent in accurately quantifying and attributing historical tendencies and making robust projections of likely future wind resources. We then analyze simulations from the current generation of regional climate models and show, at least for the next 50 years, the wind resource in the regions of greatest wind energy penetration will not move beyond the historical envelope of variability. Thus this work suggests that the wind energy industry can, and will, continue to make a contribution to electricity provision in these regions for at least the next several decades. [ABSTRACT FROM AUTHOR]

Abstract: The CALIOPE project, funded by the Spanish Ministry of the Environment, aims at establishing an air quality forecasting system for Spain. With this goal, CALIOPE modeling system was developed and applied with high resolution (4km×4km, 1h) using the HERMES emission model (including emissions of resuspended particles from paved roads) specifically built up for Spain. The present study provides an evaluation and the assessment of the modeling system, coupling WRF-ARW/HERMES/CMAQ/BSC-DREAM8b for a full-year simulation in 2004 over Spain. The evaluation focuses on the capability of the model to reproduce the temporal and spatial distribution of gas phase species (NO2, O3, and SO2) and particulate matter (PM10) against ground-based measurements from the Spanish air quality monitoring network. The evaluation of the modeling results on an hourly basis shows a strong dependency of the performance of the model on the type of environment (urban, suburban and rural) and the dominant emission sources (traffic, industrial, and background). The O3 chemistry is best represented in summer, when mean hourly variability and high peaks are generally well reproduced. The mean normalized error and bias meet the recommendations proposed by the United States Environmental Protection Agency (US-EPA) and the European regulations. Modeled O3 shows higher performance for urban than for rural stations, especially at traffic stations in large cities, since stations influenced by traffic emissions (i.e., high-NO x environments) are better characterized with a more pronounced daily variability. NO x /O3 chemistry is better represented under non-limited-NO2 regimes. SO2 is mainly produced from isolated point sources (power generation and transformation industries) which generate large plumes of high SO2 concentration affecting the air quality on a local to national scale where the meteorological pattern is crucial. The contribution of mineral dust from the Sahara desert through the BSC-DREAM8b model helps to satisfactorily reproduce episodic high PM10 concentration peaks at background stations. The model assessment indicates that one of the main air quality-related problems in Spain is the high level of O3. A quarter of the Iberian Peninsula shows more than 30days exceeding the value 120μgm−3 for the maximum 8-h O3 concentration as a consequence of the transport of O3 precursors downwind to/from the Madrid and Barcelona metropolitan areas, and industrial areas and cities in the Mediterranean coast. [Copyright &y& Elsevier]

Abstract: The ratios of observed organic carbon (OC) to elemental carbon (EC) from the rural sites of the IMPROVE network are analyzed for the 5-year period from 2000 to 2004. Among these years, nationwide OC/EC peaks are observed most consistently in the summer of 2002. Several potential factors are analyzed, including biomass burning, secondary organic aerosols (SOA) formation from biogenic sources and in-cloud processing, long-range transport from East Asia, and meteorological conditions over the U.S. We find that biomass burning and SOA formation make the most significant contributions using the global GEOS-Chem model simulations. The effect of model estimated in-cloud SOA formation is significant compared to the estimate of (non-cloud) biogenic SOA. The impacts of Canadian and western U.S. fires are larger than fires in Russia or Mexico in summer. The dry meteorological condition of the summer of 2002 tends to promote higher OC/EC ratios by inducing larger fire emissions, SOA formation, and a longer OC lifetime. [ABSTRACT FROM AUTHOR]

Abstract: We look at how well several alternative Taylor-rule specifications describe Federal Reserve policy decisions in real time, using the newly developed test for non-nested model selection in the presence of (possible) parameter instability. Further, we isolate those Taylor-rule features that are most important for achieving relatively strong real-time performance. Key features of our preferred rule, which is robust to changing economic conditions, are the partial adjustment of the federal funds rate toward an equilibrium rate that depends on the unemployment rate and forward-looking inflation measures. We conclude by presenting an empirical application to show the policy relevance of our preferred rule in the context of the 2008–2009 recession. [ABSTRACT FROM AUTHOR]

Abstract: This paper compares spatial and seasonal variations and temporal trends in modeled and measured concentrations of sulfur and nitrogen compounds in wet and dry deposition over an 18-year period (1988–2005) over a portion of the northeastern United States. Substantial emissions reduction programs occurred over this time period, including Title IV of the Clean Air Act Amendments of 1990 which primarily resulted in large decreases in sulfur dioxide (SO2) emissions by 1995, and nitrogen oxide (NOx) trading programs which resulted in large decreases in warm season NOx emissions by 2004. Additionally, NOx emissions from mobile sources declined more gradually over this period. The results presented here illustrate the use of both operational and dynamic model evaluation and suggest that the modeling system largely captures the seasonal and long-term changes in sulfur compounds. The modeling system generally captures the long-term trends in nitrogen compounds, but does not reproduce the average seasonal variation or spatial patterns in nitrate. [ABSTRACT FROM AUTHOR]

The main purpose of this study was to evaluate the ability of a human exposure-response model, which describes ozone-induced changes in forced expiratory volume in 1 second (FEV1) across a wide range of dynamic exposure conditions, to predict responses in independent data. We first conducted an n-fold cross-validation of the model using samples of the original EPA data from which the model was developed. We then identified seven more recently published studies with controlled exposures to a wide range of ozone exposure patterns relevant to the current ambient ozone health standard and used the model to calculate the mean predicted responses for the exposure conditions of the individual studies that we compared to the mean observed responses reported in these studies. The n-fold cross-validation indicated good internal agreement between mean predicted and mean observed responses in the original data used to develop the model. The model accurately captured the patterns of response in each of the seven independent studies with a tendency to overpredict the observed responses by about 1 percentage point of FEV1 decrement on average. We conclude that the model is currently capable of predicting human FEV1 responses across a wide range of dynamic exposure conditions and anticipate further improvements in predictions with the addition of low-concentration exposure data. [ABSTRACT FROM AUTHOR]

This study introduces a top-down strategy for model evaluation and selection under uncertainty in which watershed model structures with increasing complexity are applied to twelve watersheds across a hydro-climatic gradient within the United States (US). The models'' complexities and their related assumptions provide an indication of the dominant controls on the watershed response at the inter-annual, intra-annual, monthly, and daily time scales as captured in the water balance signatures (or metrics) used in this study. The ability of the models to capture the water balance signatures is evaluated in an ensemble framework with respect to their reliability (Is the model ensemble capturing the observed signature?) and with their shape (Is the model structure capable of representing an observed signature's variability?). Model selection is automated by combining the reliability and shape performance measures in a fuzzy rule system. Our results suggest that the framework can be tuned to function as a screening tool that formalizes our model selection process. This fuzzy model selection framework enhances our ability to automatically select parsimonious model structures for large databases of watersheds and therefore provides an important step towards understanding how controls on the watershed response vary with landscape and climatic characteristics. This understanding further advances our ability for model-based watershed classification. [Copyright &y& Elsevier]

An operational model evaluation procedure is described to quantitatively assess the relative skill among several regional-scale air quality models simulating various percentiles of the cumulative frequency distribution of observed daily maximum 8-h ozone concentrations. Bootstrap sampling is used to characterize the variability in the observed percentile values, thereby providing a means for assessing whether the differences seen between model predictions are significant. The procedure was designed to facilitate model inter-comparisons, since all that is needed to implement the procedure is for each modeler to provide a listing of the daily maximum 8-h ozone concentration predictions for a summer season for grid cells containing ozone monitors. Available ozone modeling results for the summer of 2002 from four regional-scale air quality simulations are used here to illustrate the results that can be obtained. These simulations were conducted using the Community Multi-Scale Air Quality (CMAQ) model with somewhat different setups. The modeling domains were different, but there is a region in the central Eastern United States where ozone estimates from all four simulations are available. Our objective is to describe the inter-comparison procedure, to illustrate the results obtained, and to stimulate discussions on how similar procedures might be developed and improved in the future. [Copyright &y& Elsevier]

Many problems and challenges of ecosystem management currently are driven by the rapid pace and spatial extent of landscape change. Parks and reserves within areas of high human population density are especially challenged to meet the recreational needs of local populations and to preserve valued environmental resources. The complex problem of managing multiple objectives and multiple resources requires an enormous quantity of information, and conceptual models have been proposed as tools for organizing and interpreting this information. Academics generally prefer a bottom-up approach to model construction that emphasizes ecologic theory and process, whereas managers often use a top-down approach that takes advantage of existing information to address more pragmatic objectives. The authors propose a formal process for developing, applying, and testing conceptual models to be used in landscape monitoring that reconciles these seemingly opposing perspectives. The four-step process embraces the role of hypothesis testing in the development of models and evaluation of their utility. An example application of the process to a network of national parks in and around Washington, DC illustrates the ability of the approach to systematically identify monitoring data that would both advance ecologic theory and inform management decisions. [ABSTRACT FROM AUTHOR]

SAFE_AIR II is a model which simulates transport and diffusion of airborne pollutants emitted in the low atmosphere above complex orography at both local and regional scale, under non-stationary and inhomogeneous emission and meteorological conditions. The aim of this paper is the assessment of the performance of the dispersive module (P6) of SAFE_AIR II in simulating pollutant concentrations due to releases from area sources. This analysis has been carried out comparing the results of model simulations with the data collected in a U.S. Environmental Protection Agency Meteorological Wind Tunnel. Three different scenarios have been investigated in order to understand the effects of the shape and orientation of the area source. To compare measured concentrations against the corresponding predicted ones, both graphical analyses and appropriate statistical indices have been used. The capability of the code to simulate area sources has been outlined. [Copyright &y& Elsevier]

Abstract: Operational, diagnostic, and comparative evaluations of two one-atmosphere regional models were performed for the full calendar year 2002 in support of regional haze regulatory applications in the eastern US. Using consistent emissions, meteorological and air quality data sets, the community multi-scale air quality and comprehensive air quality model with extensions models were exercised on a nested 36/12km grid system and evaluated across a broad range of time and space scales for numerous gas-phase and fine particulate species derived from routine and research-grade ambient measurements at six monitoring networks. Performance by both models for speciated fine particulate matter (PM) across the eastern US ranged from quite good (e.g., SO4 2−) to poor (e.g., soil). For most species, model bias was higher in the winter and lower (usually negative) in the summer suggesting potential issues related to vertical mixing (e.g., too little in winter), temporal allocation of emissions, and/or other model science processes or inputs. These results may be used to (a) guide one-atmosphere model refinements, (b) improve data input preparation procedures, (c) evaluate methods for rigorous, stressful performance testing, and (d) clarify the uncertainty in model estimates for regional haze and PM2.5 control strategy programs. [Copyright &y& Elsevier]

Abstract: The National Oceanic and Atmospheric Administration (NOAA), in partnership with the United States Environmental Protection Agency (EPA), are developing an operational, nationwide Air Quality Forecasting (AQF) system. An experimental phase of this program, which couples NOAA''s Eta meteorological model with EPA''s Community Multiscale Air Quality (CMAQ) model, began operation in June of 2004 and has been providing forecasts of ozone (O3) concentrations over the northeastern United States. An important component of this AQF system has been the development and implementation of an evaluation protocol. Accordingly, a suite of statistical metrics that facilitates evaluation of both discrete- and categorical-type forecasts was developed and applied to the system in order to characterize its performance. The results reveal that the AQF system performed reasonably well in this inaugural season (mean domain wide correlation coefficient=0.59), despite anomalously cool and wet conditions that were not conducive to the formation of O3. Due in part to these conditions, the AQF system overpredicted concentrations, resulting in a mean bias of +10.2ppb (normalized mean bias=+22.8%). In terms of error, the domain-wide root mean square error averaged 15.7ppb (normalized mean error=28.1%) for the period. Examination of the discrete and categorical metrics on a daily basis revealed that the AQF system''s level of performance was closely related to the synoptic-scale meteorology impacting the domain. The model performed very well during periods when anticyclones, characterized by clear skies, dominated. Conversely, periods characterized by extensive cloud associated with fronts and/or cyclones, resulted in poor model performance. Subsequent analysis revealed that factors associated with CMAQ''s cloud cover scheme contributed to this overprediction. Accordingly, changes to the cloud schemes are currently underway that are expected to significantly improve the AQF system''s performance in anticipation of its second year of operation. [Copyright &y& Elsevier]

Abstract: The ability of the US Environmental Protection Agency''s Community Multi-scale Air Quality (CMAQ) model to simulate the wet deposition of ammonium during 8-week winter and summer periods in 2001 is evaluated using observations from the National Acid Deposition Program (NADP) monitoring sites. The objective of this study is to ascertain the effects of precipitation simulations and emissions on CMAQ simulations of deposition. In both seasons, CMAQ tends to underpredict the deposition amounts. Based on the co-located measurements of ammonium wet deposition and precipitation at the NADP sites and on estimated precipitation amounts for each grid cell, Bayesian statistical methods are used to estimate ammonium wet deposition over all grid cells in the study region. To assess the effect of precipitation on the CMAQ simulations, our statistical method is run twice for each time period, using the simulated precipitation information provided to CMAQ and precipitation estimates based on data collected by the cooperative observer network. During the winter period when stratiform-type precipitation dominates, precipitation amounts do not seem to be a major factor in CMAQ''s ability to simulate the wet deposition of ammonium. However, during the summer period when precipitation is mainly generated by convective processes, small portions of the region are identified in which problems with precipitation simulations may be adversely affecting CMAQ''s estimates. [Copyright &y& Elsevier]

Two types of state-switching models for U.S. real output have been proposed: models that switch randomly between states and models that switch states deterministically, as in the threshold autoregressive model of Potter. These models have been justified primarily on how well they fit the sample data, yielding statistically significant estimates of the model coefficients. Here we propose a new approach to the evaluation of an estimated nonlinear time series model that provides a complement to existing methods based on in-sample fit or on out-of-sample forecasting. In this new approach, a battery of distinct nonlinearity tests is applied to the sample data, resulting in a set of p-values for rejecting the null hypothesis of a linear generating mechanism. This set of p-values is taken to be a "stylized fact" characterizing the nonlinear serial dependence in the generating mechanism of the time series. The effectiveness of an estimated nonlinear model for this time series is then evaluated in terms of the congruence between this stylized fact and a set of nonlinearity test results obtained from data simulated using the estimated model. In particular, we derive a portmanteau statistic based on this set of nonlinearity test p-values that allows us to test the proposition that a given model adequately captures the nonlinear serial dependence in the sample data. We apply the method to several estimated state-switching models of U.S. real output. [ABSTRACT FROM AUTHOR]

Abstract: This study evaluates the ability of the Community Multiscale Air Quality (CMAQ) model to simulate the spatial variability of summertime ozone (O3) at the surface and in the free troposphere over the continental United States. Simulated surface O3 concentrations are compared with 987 Air Quality System (AQS) sites and 123 Clean Air Status and Trends Network (CASTNet) sites. CMAQ''s ability to reproduce surface observations varies with O3 concentration. The model best simulates observed O3 for intermediate concentrations (40–60ppbv), while over-(under-) predicting at lower (higher) levels. CMAQ reproduces surface O3 for a wide range of conditions (30–80ppbv) with a normalized mean error (NME) less than 35% and normalized mean bias (NMB) lying between ±15% for the whole domain. Although systematically over-predicting O3 in the east and under-predicting it in the western United States, CMAQ is able to reproduce 1- and 8-h daily maxima with a cross-domain mean bias (MB) of 1 and 8ppbv, or NMB of 8% and 25%, respectively. The model underestimates observed O3 at rural sites (MB=−5ppbv, NMB=−5% and NME=23% with a 40ppbv cut-off value) and over-predicts it at urban and suburban sites by a similar magnitude (MB=6ppbv, NMB=7% and NME=25%). Apparent errors and biases decrease when data is averaged over longer periods, suggesting that most evaluation statistics are dependent on the time scale of data aggregation. Therefore, performance criteria should specify an averaging period (e.g., 1- or 8-h) and not be independent of averaging period as some current model evaluation studies imply. Comparisons of vertical profiles of simulated O3 with ozonesonde data show both overestimation and underestimation by 10–20ppbv in the lower troposphere and a consistent under-prediction in the upper troposphere. Vertical O3 distributions are better simulated when lateral boundary conditions obtained from the global Model of Ozone and Related Tracers version 2 (MOZART-2) are used, but under-prediction remains. The assumption of zero-flux at the top boundary and the resulting exclusion of the contribution of stratosphere–troposphere exchange to tropospheric O3 concentrations limit the ability of CMAQ to reproduce O3 concentrations in the upper troposphere. [Copyright &y& Elsevier]

While fluctuations in meteorological and air quality variables occur on a continuum of spatial scales, the horizontal grid spacing of coupled meteorological and photochemical models sets a lower limit on the spatial scales that they can resolve. However, both computational costs and data requirements increase significantly with increasing grid resolution. Therefore, it is important to examine, for any given application, whether the expected benefit of increased grid resolution justifies the extra costs. In this study, we examine temperature and ozone observations and model predictions for three high ozone episodes that occurred over the northeastern United States during the summer of 1995. In the first set of simulations, the meteorological model RAMS4a was run with three two-way nested grids of 108/36/12 km grid spacing covering the United States and the photochemical model UAM-V was run with two grids of 36/12 km grid spacing covering the eastern United States. In the second set of simulations, RAMS4a was run with four two-way nested grids of 108/36/12/4 km grid spacing and UAM-V was run with three grids of 36/12/4 km grid spacing with the finest resolution covering the northeastern United States. Our analysis focuses on the comparison of model predictions for the finest grid domain of the simulations, namely, the region overlapping the 12 km and 4 km domains. A comparison of 12 kmversus4 km fields shows that the increased grid resolution leads to finer texture in the model predictions; however, comparisons of model predictions with observations do not reveal the expected improvement in the predictions. While high-resolution modeling has scientific merit and potential uses, the currently available monitoring networks, in conjunction with the scarceness of highly resolved spatial input data and the limitations of model formulation, do not allow confirmation of the expected superiority of the high-resolution model predictions. [ABSTRACT FROM AUTHOR]

To study the potential impacts of climate change on air quality and public health over the eastern United States, a coupled global/regional-scale modeling system consisting of the NASA-Goddard Institute for Space Studies Atmosphere–Ocean model, the MM5 mesoscale meteorological model and the Community Multiscale Air Quality (CMAQ) model for air quality has been developed. Evaluation results of the modeling system used to simulate climate and ozone air quality over the eastern United States during the five summers of 1993–1997 are presented in this paper. The results indicate that MM5 and CMAQ capture interannual and synoptic-scale variability present in surface temperature and ozone observations in the current climate, while the magnitude of fluctuations on shorter time scales is underestimated. A comparison of observed and predicted spatial patterns of daily maximum ozone concentrations shows best performance in predicting patterns for average and above-average ozone concentrations. The frequency distributions of the duration of extreme heat and ozone events show similar features for both model predictions and observations. Finally, application of a synoptic map-typing procedure reveals that the MM5/CMAQ system succeeded in simulating the average ozone concentrations associated with several frequent pressure patterns, indicating that the effects of synoptic-scale meteorology on ozone concentrations are captured by the modeling system. It is concluded that the GCM/MM5/CMAQ system is a suitable tool for the simulation of summertime surface temperature and ozone air quality conditions over the eastern United States in the present climate. [Copyright &y& Elsevier]

This study develops a regional econometric model to extend and complement input-output economic impact analysis. Following the tradition of "linking" econometric and input-output models, this work extends the previous linear econometric specification to a more accurate nonlinear form. It also extends previous specification of these models to address the issue of job creation and job capture by residents. The resulting empirical model is based on a cross-sectional econometric analysis of all Washington counties. The use of the estimated econometric model is illustrated with a real job growth scenario using the growth in jobs between 1990 and 1997 for each Washington county. The simultaneous equation model evaluated by examining how closely the estimated population variable tracks its actual values in 1997. The root mean square (RMS) percent error for population was 5.46 percent. Most of the new jobs are captured by immigrants instead of county residents. However the positive long-run effect of job growth on labor force participation remains, though its effect is small. [ABSTRACT FROM AUTHOR]

This study compares several transportation-related air quality models. Two line-source models were used: Caline4 (California Department of Transportation) and Hiway2 (US Environmental Protection Agency). Two mobile-source models, Mobile5b (US Environmental Protection Agency) and COPERT3 (European Environment Agency), along with real-world emission factors were used and evaluated as well. Model predictions of NOx concentrations were compared to measured values at two sites in Israel, differing by fleet composition and physical layout (‘at-grade’ and a ‘cut/depressed’ road sections). The process indicated that emission factors generated by COPERT3 are the most appropriate for free flowing traffic situations in Israel. Predictions by both line-source models were similar when applied to ‘at-grade’ road sections. When applied to ‘cut/depressed’ sections, Hiway2 better predicted concentrations during unstable conditions, while Caline4 better predicted concentrations during stable conditions and peak concentrations. [Copyright &y& Elsevier]

Net energy and protein systems (hereafter called feed evaluation systems) offer the possibility to formulate rations by matching feed values (e.g., net energy and metabolizable protein) with animal requirements. The accuracy and precision of this approach relies heavily on the quantification of various animal digestive and metabolic responses to dietary changes. Therefore, the aims of the current study were, first, to evaluate the predicted responses to dietary changes of total-tract digestibility (including organic matter, crude protein, and neutral detergent fiber) and nitrogen (N) flows at the duodenum (including microbial N and undigested feed N together with endogenous N) against measurements from published studies by 2 different feed evaluation systems. These feed evaluation systems were the recently updated Institut National de la Recherche Agronomique (INRA, 2018) and the older, yet widely used, National Research Council (NRC, 2001) system. The second objective was to estimate the accuracy and precision of predicting milk yield responses based on values of net energy (NE L ) and metabolizable protein (MP) supply predicted by the 2 feed evaluation systems. For this, published studies, with experimentally induced changes in either NE L or MP content, were used to calibrate the relationship of NE L and MP supply, with milk component yields. Based on the slope, root mean square prediction error, and concordance correlation coefficient (CCC), the results obtained show that total nonammonia nitrogen flow at the duodenum was predicted with similar accuracy and precision, but considerably better prediction was achieved when the INRA model was used to predict organic matter and neutral detergent fiber digestibility responses. The average NE L and MP content predicted by both models was similar, but NE L and MP content of individual diets differed substantially between both models as indicated by determination coefficients of 0.45 (NE L content) and 0.50 (MP content). Despite these differences, this work shows that when response equations are calibrated with NE L and MP values either from the INRA model or from the NRC model, the accuracy and precision (slope, root mean square prediction error, and CCC) of the predicted milk component yields responses is similar between the models. The lowest accuracy and precision were observed for milk fat yield response, with CCC values in the range of 0.37 to 0.40, compared with milk lactose and protein yields responses for which CCC values were in the range of 0.75 to 0.81., (© 2020, The Authors. Published by Elsevier Inc. and Fass Inc. on behalf of the American Dairy Science Association®. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).)

Previous studies have proposed that model performance statistics from earlier photochemical grid model (PGM) applications can be used to benchmark performance in new PGM applications. A challenge in implementing this approach is that limited information is available on consistently calculated model performance statistics that vary spatially and temporally over the U.S. Here, a consistent set of model performance statistics are calculated by year, season, region, and monitoring network for PM 2.5 and its major components using simulations from versions 4.7.1-5.2.1 of the Community Multiscale Air Quality (CMAQ) model for years 2007–2015. The multi-year set of statistics is then used to provide quantitative context for model performance results from the 2015 simulation. Model performance for PM 2.5 organic carbon in the 2015 simulation ranked high (i.e., favorable performance) in the multi-year dataset, due to factors including recent improvements in biogenic secondary organic aerosol and atmospheric mixing parameterizations in CMAQ. Model performance statistics for the Northwest region in 2015 ranked low (i.e., unfavorable performance) for many species in comparison to the 2007–2015 dataset. This finding motivated additional investigation that suggests a need for improved speciation of wildfire PM 2.5 emissions and modeling of boundary layer dynamics near water bodies. Several limitations were identified in the approach of benchmarking new model performance results with previous results. Since performance statistics vary widely by region and season, a simple set of national performance benchmarks (e.g., one or two targets per species and statistic) as proposed previously are inadequate to assess model performance throughout the U.S. Also, trends in model performance statistics for sulfate over the 2007 to 2015 period suggest that model performance for earlier years may not be a useful reference for assessing model performance for recent years in some cases. Comparisons of results from the 2015 base case with results from five sensitivity simulations demonstrated the importance of parameterizations of NH 3 surface exchange, organic aerosol volatility and production, and emissions of crustal cations for predicting PM 2.5 species concentrations. • A new method is demonstrated for benchmarking photochemical grid model performance. • Model performance statistics for 2015 are compared with statistics for 2007–2015. • Sensitivity simulations provide insights about processes driving PM 2.5 concentrations. [ABSTRACT FROM AUTHOR]

The goal of customer retention campaigns, by design, is to add value and enhance the operational efficiency of businesses. For organizations that strive to retain their customers in saturated, and sometimes fast moving, markets such as the telecommunication and banking industries, implementing customer churn prediction models that perform well and in accordance with the business goals is vital. The expected maximum profit (EMP) measure is tailored toward this problem by taking into account the costs and benefits of a retention campaign and estimating its worth for the organization. Unfortunately, the measure assumes fixed and equal customer lifetime value (CLV) for all customers, which has been shown to not correspond well with reality. In this article, we extend the EMP measure to take into account the variability in the lifetime values of customers, thereby basing it on individual characteristics. We demonstrate how to incorporate the heterogeneity of CLVs when CLVs are known, when their prior distribution is known, and when neither is known. By taking into account individual CLVs, our proposed approach of measuring model performance gives novel insights when deciding on a customer retention campaign. The method is dependent on the characteristics of the customer base as is compliant with modern business analytics and accommodates the data-driven culture that has manifested itself within organizations.

This work evaluated the National Research Council (NRC) dairy model (2001) predictions of rumen undegradable (RUP) and degradable (RDP) protein compared with measured postruminal non-ammonia, nonmicrobial (NANMN) and microbial N flows. Models were evaluated using the root mean squared prediction error (RMSPE) as a percent of the observed mean, mean and slope biases as percentages of mean squared prediction error (MSPE), and concordance correlation coefficient (CCC). The NRC (2001) over-estimated NANMN by 18% and under-estimated microbial N by 14%. Both responses had large mean biases (19% and 20% of MSPE, respectively), and NANMN had a slope bias (22% of MSPE). The NRC NANMN estimate had high RMSPE (46% of observed mean) and low CCC (0.37); updating feed library A, B, and C protein fractions and degradation rate (K d ) estimates with newer literature only marginally improved fit. The re-fit NRC models for NANMN and microbial N had CCC of 0.89 and 0.94, respectively. When compared with a prediction of NANMN as a static mean fraction of N intake, the re-derived NRC approach did not have improved fit. A protein system of intermediate complexity was derived in an attempt to estimate NANMN with improved fit compared with the static mean NANMN model. In this system, postruminal appearance of A, B, and C protein fractions were predicted in a feed-type specific manner rather than from estimated passage and degradation rates. In a comparison to independent data achieved through cross-validation, the new protein system improved RMSPE (34 vs. 36% of observed mean) and CCC (0.42 vs. 0.30) compared with the static mean NANMN model. When the NRC microbial N equation was re-derived, the RDP term dropped from the model. Consequently, 2 new microbial protein equations were formulated, both used a saturating (increasing at a decreasing rate) form: one saturated with respect to TDN and the other saturated over increasing intakes of rumen degraded starch and NDF. Both equations expressed maximal microbial N production as a linear function of RDP intake. The function relating microbial N to intake of rumen degradable carbohydrate improved RMSPE (24 vs. 28% of the observed mean) and CCC (0.63 vs 0.30) compared with the re-derived NRC model. The newly derived equations showed modest improvements in model fit and improved capacity to account for known biological effects; however, substantial variability in NANMN and microbial N estimates remained unexplained., (The Authors. Published by the Federation of Animal Science Societies and Elsevier Inc. on behalf of the American Dairy Science Association®. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).)

Evaluation of ration balancing systems such as the National Research Council (NRC) Nutrient Requirements series is important for improving predictions of animal nutrient requirements and advancing feeding strategies. This work used a literature data set (n = 550) to evaluate predictions of total-tract digested neutral detergent fiber (NDF), fatty acid (FA), crude protein (CP), and nonfiber carbohydrate (NFC) estimated by the NRC (2001) dairy model. Mean biases suggested that the NRC (2001) lactating cow model overestimated true FA and CP digestibility by 26 and 7%, respectively, and under-predicted NDF digestibility by 16%. All NRC (2001) estimates had notable mean and slope biases and large root mean squared prediction error (RMSPE), and concordance (CCC) ranged from poor to good. Predicting NDF digestibility with independent equations for legumes, corn silage, other forages, and nonforage feeds improved CCC (0.85 vs. 0.76) compared with the re-derived NRC (2001) equation form (NRC equation with parameter estimates re-derived against this data set). Separate FA digestion coefficients were derived for different fat supplements (animal fats, oils, and other fat types) and for the basal diet. This equation returned improved (from 0.76 to 0.94) CCC compared with the re-derived NRC (2001) equation form. Unique CP digestibility equations were derived for forages, animal protein feeds, plant protein feeds, and other feeds, which improved CCC compared with the re-derived NRC (2001) equation form (0.74 to 0.85). New NFC digestibility coefficients were derived for grain-specific starch digestibilities, with residual organic matter assumed to be 98% digestible. A Monte Carlo cross-validation was performed to evaluate repeatability of model fit. In this procedure, data were randomly subsetted 500 times into derivation (60%) and evaluation (40%) data sets, and equations were derived using the derivation data and then evaluated against the independent evaluation data. Models derived with random study effects demonstrated poor repeatability of fit in independent evaluation. Similar equations derived without random study effects showed improved fit against independent data and little evidence of biased parameter estimates associated with failure to include study effects. The equations derived in this analysis provide interesting insight into how NDF, starch, FA, and CP digestibilities are affected by intake, feed type, and diet composition., (The Authors. Published by the Federation of Animal Science Societies and Elsevier Inc. on behalf of the American Dairy Science Association®. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).)