Monte Carlo Uncertainty Analysis Research Articles

Uncertainty in the modelling of spatial and temporal patterns of shallow groundwater flow paths: The role of geological and hydrological site information

Understanding the hydrological and hydrogeochemical responses of hillslopes and other small scale groundwater systems requires mapping the velocity and direction of groundwater flow relative to the controlling subsurface material features. Since point observations of subsurface materials and groundwater head are often the basis for modelling these complex, dynamic, three-dimensional systems, considerable uncertainties are inevitable, but are rarely assessed. This study explored whether piezometric head data measured at high spatial and temporal resolution over six years at a hillslope research site provided sufficient information to determine the flow paths that transfer nitrate leached from the soil zone through the shallow saturated zone into a nearby wetland and stream. Transient groundwater flow paths were modelled using MODFLOW and MODPATH, with spatial patterns of hydraulic conductivity in the three material layers at the site being estimated by regularised pilot point calibration using PEST, constrained by slug test estimates of saturated hydraulic conductivity at several locations. Subsequent Null Space Monte Carlo uncertainty analysis showed that this data was not sufficient to definitively determine the spatial pattern of hydraulic conductivity at the site, although modelled water table dynamics matched the measured heads with acceptable accuracy in space and time. Particle tracking analysis predicted that the saturated flow direction was similar throughout the year as the water table rose and fell, but was not aligned with either the ground surface or subsurface material contours; indeed the subsurface material layers, having relatively similar hydraulic properties, appeared to have little effect on saturated water flow at the site. Flow path uncertainty analysis showed that, while accurate flow path direction or velocity could not be determined on the basis of the available head and slug test data alone, the origin of well water samples relative to the material layers and site contour could still be broadly deduced. This study highlights both the challenge of collecting suitably informative field data with which to characterise subsurface hydrology, and the power of modern calibration and uncertainty modelling techniques to assess flow path uncertainty in hillslopes and other small scale systems.

Optimization for Early-Warning Monitoring Networks in Well Catchments Should Be Multi-objective, Risk-Prioritized and Robust Against Uncertainty

Abstract Groundwater abstraction wells are commonly protected by zones of restricted land use. Such well protection areas typically cannot cover the entire well catchment, and numerous risk sources remain. Each risk source could release contaminants at any time, affect the well earlier or later, and thus put the quality of supplied water at risk. In this context, it seems fortunate that most well catchments are equipped with monitoring networks. Such networks, however, often grew historically while following diverse purposes that changed with time. Thus, they are often inadequate (or at least suboptimal) as reliable risk control mechanism. We propose to optimize existing or new monitoring networks in a multi-objective setting. The different objectives are minimal costs, maximal reliability in detecting recent or future contaminant spills, and early detection. In a synthetic application scenario, we show that (1) these goals are in fact competing, and a multi-objective analysis is suitable, (2) the optimization should be made robust against predictive uncertainty through scenario-based or Monte Carlo uncertainty analysis, (3) classifying the risk sources (e.g., as severe, medium, almost tolerable) is useful to prioritize the monitoring needs and thus to obtain better compromise solutions under budgetary constraints, and (4) one can defend the well against risk sources at unknown locations through an adequate model for the residual risk. Overall, the concept brings insight into the costs of reliability, the costs of early warning, the costs of uncertainty, and into the trade-off between covering only severe risks versus the luxury situation of controlling almost tolerable risks as well.

Monte carlo uncertainty analysis for photothermal radiometry measurements using a curve fit process

Photothermal radiometry (PTR) has become a popular method to measure thermal properties of layered materials. Much research has been done to determine the capabilities of PTR, but with little uncertainty analysis. This study reports a Monte Carlo uncertainty analysis to quantify uncertainty of film diffusivity and effusivity measurements, presents a sensitivity study for each input parameter, compares linear and logarithmic spacing of data points on frequency scans, and investigates the validity of a one-dimensional heat transfer assumption. Logarithmic spacing of frequencies when taking data is found to be unequivocally superior to linear spacing, while the use of a higher-dimensional heat transfer model is only needed for certain measurement configurations. The sensitivity analysis supports the frequency spacing conclusion, as well as explains trends seen in the uncertainty data.

Statistical performance of CO2 leakage detection using seismic travel time measurements

AbstractMonitoring for possible CO2 leakage is an important part of a safe and effective geological sequestration program. Seismic monitoring has been implemented in several pilot sequestration sites for site characterization and CO2 leakage detection. This study evaluates the detection power of seismic wave travel time measurements and statistical tests at different CO2 leakage rate levels. A simplified rock physics model is assumed for monitoring zones at sequestration sites and the effects of leakage‐induced changes in pressure and CO2 saturation on P‐wave travel times are modeled. The empirical distributions of detection power using the P‐wave travel time for four regions in the permeability‐porosity input space at four leakage levels are obtained from the Monte Carlo uncertainty analysis with a stochastic response surface method. The detection power using the P‐wave travel time measurements and test alone is generally not high enough, unless the porosity and the permeability of the monitoring zone are high, and/or a long period of time has elapsed since the leakage occurred. For monitoring layers with lower permeability and porosity, measurements from other monitoring techniques will likely be needed to increase the probability that leakage events are detected and addressed in a timely manner. © 2015 Society of Chemical Industry and John Wiley & Sons, Ltd

Markov Chain Monte Carlo (MCMC) uncertainty analysis for watershed water quality modeling and management

Watershed-scale water quality (WWQ) models are now widely used to support management decision-making. However, significant uncertainty in the model outputs remains a largely unaddressed issue. In recent years, Markov Chain Monte Carlo (MCMC), a category of formal Bayesian approaches for uncertainty analysis (UA), has become popular in the field of hydrological modeling, but its applications to WWQ modeling have been rare. This study systematically evaluated the applicability of MCMC in assessing the uncertainty of WWQ modeling, using Differential Evolution Adaptive Metropolis (DREAM(ZS)) and SWAT as the representative MCMC algorithm and WWQ model, respectively. The nitrate pollution in Newport Bay watershed was the case study for numerical experiments. It has been concluded that the efficiency and effectiveness of a MCMC algorithm would depend on some critical designs of the UA, including: (i) how many and which model parameters to be considered as random in the MCMC analysis; (ii) where to fix the non-random model parameters; and (iii) which criteria to stop the Markov Chain. The study results also indicate that the MCMC UA has to be management-oriented, that is, management objectives should be factored into the designs of the UA, rather than be considered after the UA.

Biochar and Activated Carbon for Enhanced Trace Organic Contaminant Retention in Stormwater Infiltration Systems

To assess the effectiveness of biochar and activated carbon (AC) for enhanced trace organic contaminant (TOrC) retention in stormwater infiltration systems, an approach combining forward-prediction modeling and laboratory verification experiments was employed. Batch and column tests were conducted using representative TOrCs and synthetic stormwater. Based on batch screening tests, two commercially available biochars (BN-biochar and MCG-biochar) and an AC were investigated. The AC exhibited the strongest sorption, followed by MCG-biochar and BN-biochar. Langmuir isotherms provided better fits to equilibrium data than Freundlich isotherms. Due to superior sorption kinetics, 0.2 wt % MCG-biochar in saturated sand columns retained TOrCs more effectively than 1.0 wt % BN-biochar. A forward-prediction intraparticle diffusion model based on the Langmuir isotherm adequately predicted column results when calibrated using only batch parameters, as indicated by a Monte Carlo uncertainty analysis. Case study simulations estimated that an infiltration basin amended with F300-AC or MCG-biochar could obtain sorption-retarded breakthrough times for atrazine of 54 or 5.8 years, respectively, at a 1 in./h infiltration rate. These results indicate that biochars or ACs with superior sorption capacity and kinetics can enhance TOrC retention in infiltration systems, and performance under various conditions can be predicted using results from batch tests.

Low energy scattering cross section ratios ofN14(p,p)N14

Background: The slowest reaction in the first CNO cycle is $^{14}\mathrm{N}(p,\ensuremath{\gamma})^{15}\mathrm{O}$, therefore its rate determines the overall energy production efficiency of the entire cycle. The cross section presents several strong resonance contributions, especially for the ground-state transition. Some of the properties of the corresponding levels in the $^{15}\mathrm{O}$ compound nucleus remain uncertain, which affects the uncertainty in extrapolating the capture cross section to the low energy range of astrophysical interest.Purpose: The $^{14}\mathrm{N}(p,\ensuremath{\gamma})^{15}\mathrm{O}$ cross section can be described by using the phenomenological $R$ matrix. Over the energy range of interest, only the proton and $\ensuremath{\gamma}$-ray channels are open. Since resonance capture makes significant contributions to the $^{14}\mathrm{N}(p,\ensuremath{\gamma})^{15}\mathrm{O}$ cross section, resonant proton scattering data can be used to provide additional constraints on the $R$-matrix fit of the capture data.Methods: A 4 MV KN Van de Graaff accelerator was used to bombard protons onto a windowless gas target containing enriched $^{14}\mathrm{N}$ gas over the proton energy range from ${E}_{p}=1.0$ to 3.0 MeV. Scattered protons were detected at ${\ensuremath{\theta}}_{\text{lab}}=90, {120}^{\ensuremath{\circ}}, {135}^{\ensuremath{\circ}}, {150}^{\ensuremath{\circ}}$, and ${160}^{\ensuremath{\circ}}$ using ruggedized silicon detectors. In addition, a 10 MV FN Tandem Van de Graaff accelerator was used to accelerate protons onto a solid Adenine (${\mathrm{C}}_{5}{\mathrm{H}}_{5}{\mathrm{N}}_{5}$) target, of natural isotopic abundance, evaporated onto a thin self-supporting carbon backing, over the energy range from ${E}_{p}=1.8$ to 4.0 MeV. Scattered protons were detected at 28 angles between ${\ensuremath{\theta}}_{\text{lab}}=30.{4}^{\ensuremath{\circ}}$ and $167.{7}^{\ensuremath{\circ}}$ by using silicon photodiode detectors.Results: Relative cross sections were extracted from both measurements. While the relative cross sections do not provide as much constraint as absolute measurements, they greatly reduce the dependence of the data on otherwise significant systematic uncertainties, which are more difficult to quantify. The data are fit simultaneously using an $R$-matrix analysis and level energies and proton widths are extracted. Even with relative measurements, the statistics and large angular coverage of the measurements result in more confident values for the energies and proton widths of several levels; in particular, the broad resonance at ${E}_{\text{c.m.}}=2.21$ MeV, which corresponds to the $3/{2}^{+}$ level at ${E}_{x}=9.51$ MeV in $^{15}\mathrm{O}$. In particular, the $s$- and $d$-wave angular-momentum channels are separated.Conclusion: The relative cross sections provide a consistent set of data that can be used to better constrain a full multichannel $R$-matrix extrapolation of the capture data. It has been demonstrated how the scattering data reduce the uncertainty through a preliminary Monte Carlo uncertainty analysis, but several other issues remain that make large contributions to the uncertainty, which must be addressed by further capture and lifetime measurements.

Investigation of Drag-Modulated Supersonic Inflatable Aerodynamic Decelerators for Sounding Rocket Payloads

The goal of this investigation is to understand the sizing and performance of supersonic inflatable aerodynamic decelerators for Earth-based sounding rocket applications. The recovery system under examination is composed of a supersonic inflatable aerodynamic decelerator and a guided parafoil system to achieve sub-100 m miss distances. Three supersonic inflatable aerodynamic decelerator configurations (tension cone, attached isotensoid, and trailing isotensoid) are examined using the metrics of decelerator mass, aerodynamic performance, and vehicle integration. In terms of aerodynamic performance, the tension cone is the preferred choice for the sizes investigated. The attached isotensoid was shown to be the most mass efficient decelerator, whereas the trailing isotensoid was found to be the more ideal decelerator for vehicle integration. A three-degree-of-freedom trajectory simulation is used in conjunction with Monte Carlo uncertainty analysis to assess the landed accuracy capability of the proposed architectures. In 95% of the cases examined, the drag-modulated inflatable aerodynamic decelerator provides arrivals within the 10 km parafoil capability region, meeting the sub-100 m landed recovery goals. In 76% of the cases examined, the drag-modulated inflatable aerodynamic decelerator arrives within 5 km of this target zone.

金属矿冶区稻米锰的健康风险分析

通过对湖南某金属矿冶区稻米锰含量和当地居民膳食结构进行系统调查，考察调查区域的农产品安全现状。依据美国EPA推荐的非致癌污染物健康风险模型进行Monte-Carlo不确定性分析，采用Crystal ball软件对研究区域居民膳食结构进行模拟分析，评估不同人群长期食用当地大米的锰致癌风险。结果表明：该矿冶区居民食用当地大米，锰的健康风险较低，但应警惕儿童易感人群的锰健康风险。 Manganese content in the local rice and the rice dietary of residents in a metallurgy area were in-vestigated on-the-spot. Based on the Monte-Carlo uncertainty analysis of the risk model for health hazard recommended by the USEPA, Mn carcinogenic risk of different populations was evaluated with Crystal ball software. All the results showed that manganese might not be one of the risk fac-tors for an adult exposed through rice consumption.

Efficiency enhancement of optimized Latin hypercube sampling strategies: Application to Monte Carlo uncertainty analysis and meta-modeling

The majority of literature regarding optimized Latin hypercube sampling (OLHS) is devoted to increasing the efficiency of these sampling strategies through the development of new algorithms based on the combination of innovative space-filling criteria and specialized optimization schemes. However, little attention has been given to the impact of the initial design that is fed into the optimization algorithm, on the efficiency of OLHS strategies. Previous studies, as well as codes developed for OLHS, have relied on one of the following two approaches for the selection of the initial design in OLHS: (1) the use of random points in the hypercube intervals (random LHS), and (2) the use of midpoints in the hypercube intervals (midpoint LHS). Both approaches have been extensively used, but no attempt has been previously made to compare the efficiency and robustness of their resulting sample designs. In this study we compare the two approaches and show that the space-filling characteristics of OLHS designs are sensitive to the initial design that is fed into the optimization algorithm. It is also illustrated that the space-filling characteristics of OLHS designs based on midpoint LHS are significantly better those based on random LHS. The two approaches are compared by incorporating their resulting sample designs in Monte Carlo simulation (MCS) for uncertainty propagation analysis, and then, by employing the sample designs in the selection of the training set for constructing non-intrusive polynomial chaos expansion (NIPCE) meta-models which subsequently replace the original full model in MCSs. The analysis is based on two case studies involving numerical simulation of density dependent flow and solute transport in porous media within the context of seawater intrusion in coastal aquifers. We show that the use of midpoint LHS as the initial design increases the efficiency and robustness of the resulting MCSs and NIPCE meta-models. The study also illustrates that this relative improvement decreases with increasing number of sample points and input parameter dimensions. Since the computational time and efforts for generating the sample designs in the two approaches are identical, the use of midpoint LHS as the initial design in OLHS is thus recommended.

Suitability and uncertainty of two models for the simulation of ammonia dispersion from a pig farm located in an area with frequent calm conditions

We used two atmospheric dispersion models (ADMS and AERMOD) to simulate the short-range dispersion of ammonia emitted by two pig farms to assess their suitability in situations with frequent calm meteorological conditions. Simulations were carried out both using constant and temporally-varying emission rates to evaluate the effect on the model predictions. Monthly and annual mean concentrations predicted by the models at locations within one kilometre of the farms were compared with measured values. AERMOD predicted higher concentrations than ADMS (by a factor of 6–7, on average) and predicted the atmospheric concentrations more accurately for both the monthly and annual simulations. The differences between the concentrations predicted by the two models were mainly the result of different calm wind speed thresholds used by the models. The use of temporally-varying emission rates improved the performance of both models for the monthly and annual simulations with respect to the constant emission simulations. A Monte Carlo uncertainty analysis based on the inputs judged to be the most uncertain for the selected case study estimated a prediction uncertainty of ± a factor of two for both models with most of this due to uncertainty in emission rates.

Comparative Life Cycle Assessment of ceramic versus concrete roof tiles in the Brazilian context

Abstract The Brazilian ceramic industry is responsible for providing more than 90% of the roof coverings and wall bricks in the country, producing more than 15 billion pieces per year. In order to compare the life cycle impacts of ceramic versus concrete roofing tiles and identify potential improvements in ceramic products, we carried out a life cycle impact assessment of both products. This study aimed to compare the life cycle impacts of ceramic and concrete roof coverage over 1 m 2 , with an assumed life time of 20 years in Brazil. Nine different sensitivity analyses were carried out followed by a Monte Carlo uncertainty analysis to verify the robustness of the study. The results show that ceramic tiles appear to have less impact than concrete tiles on Climate Change, Resource Depletion and Water Withdrawal, while for the remaining damage categories, Human Health and Ecosystem Quality, the difference between the two alternatives was too low to be considered significant. The use of wood chips led to significant impacts, mainly related to respiratory inorganics. Assessment of the data quality identified that the data is of generally high or acceptable quality. The sensitivity analysis and uncertainty assessment show that the conclusions are robust.

A decomposition‐based approach to uncertainty analysis of feed‐forward multicomponent systems

SUMMARYTo support effective decision making, engineers should comprehend and manage various uncertainties throughout the design process. Unfortunately, in today's modern systems, uncertainty analysis can become cumbersome and computationally intractable for one individual or group to manage. This is particularly true for systems comprised of a large number of components. In many cases, these components may be developed by different groups and even run on different computational platforms. This paper proposes an approach for decomposing the uncertainty analysis task among the various components comprising a feed‐forward system and synthesizing the local uncertainty analyses into a system uncertainty analysis. Our proposed decomposition‐based multicomponent uncertainty analysis approach is shown to be provably convergent in distribution under certain conditions. The proposed method is illustrated on quantification of uncertainty for a multidisciplinary gas turbine system and is compared to a traditional system‐level Monte Carlo uncertainty analysis approach. Copyright © 2014 John Wiley & Sons, Ltd.

A Bayesian approach to CO2 leakage detection at saline sequestration sites using pressure measurements

Diffusive leakage of CO2 through the sealing caprock is likely to occur at saline sequestration sites if the permeability of the caprock is too high. Pressure monitoring is a widely used technique for CO2 leakage detection. This study aims to characterize the CO2 leakage level in an idealized CO2 storage site through an assessment of the integrity and permeability of the caprock inferred from pressure measurements in the injection zone. Monte Carlo uncertainty analysis is conducted, allowing the detection power of pressure measurements to be estimated. Based on Bayesian classification theory, the probability of each caprock permeability class given pressure measurements with measurement error can be inferred. The influence of time and location of measurements, the relative magnitude of the measurement error, the CO2 injection rate, and the assumed uncertainty in reservoir properties on the inferred caprock permeability class and the detection power of the pressure monitoring is also evaluated. Pressure monitoring alone is not sufficient to detect a permeable caprock for the system considered in this study, unless the caprock permeability is very high. A significant improvement in the statistical detection power is evident in the case where the uncertainty in reservoir properties is reduced through further site characterization. Complementary monitoring techniques will likely be needed to improve the power of pressure-based CO2 leakage detection to acceptable levels.

Cryptosporidium and Giardia in tropical recreational marine waters contaminated with domestic sewage: Estimation of bathing-associated disease risks

Sewage is a major contributor to pollution problems involving human pathogens in tropical coastal areas. This study investigated the occurrence of intestinal protozoan parasites (Giardia and Cryptosporidium) in tropical recreational marine waters contaminated with sewage. The potential risks of Cryptosporidium and Giardia infection from recreational water exposure were estimated from the levels of viable (oo) cysts (DIC+, DAPI+, PI−) found in near-shore swimming areas using an exponential dose response model. A Monte Carlo uncertainty analysis was performed in order to determine the probability distribution of risks. Microbial indicators of recreational water quality (enterococci, Clostridium perfringens) and genetic markers of sewage pollution (human-specific Bacteroidales marker [HF183] and Clostridium coccoides) were simultaneously evaluated in order to estimate the extent of water quality deterioration associated with human wastes. The study revealed the potential risk of parasite infections via primary contact with tropical marine waters contaminated with sewage; higher risk estimates for Giardia than for Cryptosporidium were found. Mean risks estimated by Monte Carlo were below the U.S. EPA upper bound on recreational risk of 0.036 for cryptosporidiosis and giardiasis for both children and adults. However, 95th percentile estimates for giardiasis for children exceeded the 0.036 level. Environmental surveillance of microbial pathogens is crucial in order to control and eradicate the effects that increasing anthropogenic impacts have on marine ecosystems and human health.

Uncertainty analysis of estuarine hydrodynamic models: an evaluation of input data uncertainty in the weeks bay estuary, alabama

Uncertainty analyses are necessary to identify, evaluate, and report the main sources of errors in modeling studies and their impacts on the model predictions. Although uncertainty analysis has been a subject of increased interest by the water resources community during the last couple of decades, in estuarine hydrodynamic modeling this is an emerging topic that requires more research. Some of the most relevant problems remaining in the practice include the identification of the principal sources of errors affecting the model predictions, and the identification of effective and computationally feasible methodologies for their quantification. This investigation evaluates the impacts of input data errors on the predictions of a 3D hydrodynamic model of the Weeks Bay estuary, Alabama. The uncertainty analysis is performed using the First Order Variance Analysis (FOVA) and the results compared to a standard Monte Carlo Uncertainty Analysis (MCUA). A procedure to implement a skill assessment as a fundamental component of the FOVA method is presented. The uncertainty analyses are performed temporally as well as spatially distributed over the model domain. The results indicate that the uncertainty in a prognostic variable is not homogeneously distributed over the computational domain, and that there are areas prone to a higher or lower uncertainty. The identification of these areas is relevant for the design of data collection plans intended to improve the confidence in the model results. The comparison of the methods indicates that both are effective to provide uncertainty estimates, although FOVA tends to overestimate the predictions obtained by MCUA. In general this overestimation can be considered as a conservative estimation of the uncertainty given the existence of other sources of errors more complex to evaluate.

Reduction of costs for anemia-management drugs associated with the use of ferric citrate.

BackgroundFerric citrate is a novel phosphate binder which has the potential to reduce usage of erythropoietin-stimulating agents (ESAs) and intravenous (IV) iron used for anemia management during hemodialysis (HD) among patients with end-stage renal disease (ESRD). Currently, the potential health care cost savings on a national scale due to the use of ferric citrate in ESRD are undetermined.MethodsPer-patient-per-year costs of ESAs (Epogen® and Aranesp® [Amgen Inc., CA, USA]) and IV iron (Venofer® [American Regent, Inc., NY, USA] and Ferrlecit® [Sanofi US, Bridgewater, NJ, USA]) were based on RED BOOK™ (Truven Health Analytics New York, NY, USA) costs combined with the Centers for Medicare and Medicaid Services (CMS) base rate and actual usage in 2011 for the four drugs. The annual number of outpatients undergoing HD in the US was based on frequencies reported by the USRDS (United States Renal Data System). Monte Carlo uncertainty analysis was performed to determine total annual costs and cost reduction based on ferric citrate usage.ResultsTotal annual cost of ESAs and IV iron for anemia management in ESRD determined by Monte Carlo analysis assuming CMS base rate value was 5.127 (3.664–6.260) billion USD. For actual utilization in 2011, total annual cost of ESAs and IV iron was 3.981 (2.780–4.930) billion USD. If ferric citrate usage reduced ESA utilization by 20% and IV iron by 40%, then total cost would be reduced by 21.2% to 4.038 (2.868–4.914) billion USD for the CMS base rate, and by 21.8% to 3.111 (2.148–3.845) billion USD, based on 2011 actual utilization.ConclusionIt is likely that US health care costs for anemia-management drugs associated with ESRD among HD patients can be reduced by using ferric citrate as a phosphate binder.

Smoking-attributable burden of lung cancer in the Philippines

BackgroundIn the Philippines, smoking is highly prevalent and tobacco control policies fail to fully implement the WHO Framework Convention on Tobacco Control provisions. To aid in policy change, intervention implementation,...

Uncertainty in annual rankings from NOAA's global temperature time series

[1] Annual rankings of global temperature are an important component of climate monitoring. However, there is some degree of uncertainty for every yearly value in the global temperature time series, which leads to uncertainty in annual rankings as well. This study applies a Monte Carlo uncertainty analysis to the National Oceanic and Atmospheric Administration (NOAA) National Climatic Data Center's global land-ocean surface temperature (NOAATMP) time series. Accounting for persistence between years does not materially affect the results versus presuming statistical independence. The highest probabilities for the warmest year analysis (1880–2012) are associated with the years 2010 (~36%), 2005 (~28%), and 1998 (~11%). The current separation among the warmest observed years is relatively small compared to the standard errors of the NOAATMP time series. However, each year between 1997 and 2012 was warmer than the vast majority of all other years since 1880 at the 95% confidence level.

Application of TREECS Modeling System to Strontium-90 for Borschi Watershed near Chernobyl, Ukraine

The Training Range Environmental Evaluation and Characterization System (TREECS™) (http://el.erdc.usace.army.mil/treecs/) is being developed by the U.S. Army Engineer Research and Development Center (ERDC) for the U.S. Army to forecast the fate of munitions constituents (MC) (such as high explosives (HE) and metals) found on firing/training ranges, as well as those subsequently transported to surface water and groundwater. The overall purpose of TREECS™ is to provide environmental specialists with tools to assess the potential for MC migration into surface water and groundwater systems and to assess range management strategies to ensure protection of human health and the environment. The multimedia fate/transport models within TREECS™ are mathematical models of reduced form (e.g., reduced dimensionality) that allow rapid application with less input data requirements compared with more complicated models. Although TREECS™ was developed for the fate of MC from military ranges, it has general applicability to many other situations requiring prediction of contaminant (including radionuclide) fate in multi-media environmental systems.TREECS™ was applied to the Borschi watershed near the Chernobyl Nuclear Power Plant, Ukraine. At this site, TREECS™ demonstrated its use as a modeling tool to predict the fate of strontium 90 (90Sr). The most sensitive and uncertain input for this application was the soil-water partitioning distribution coefficient (Kd) for 90Sr. The TREECS™ soil model provided reasonable estimates of the surface water export flux of 90Sr from the Borschi watershed when using a Kd for 90Sr of 200 L/kg. The computed export for the year 2000 was 0.18% of the watershed inventory of 90Sr compared to the estimated export flux of 0.14% based on field data collected during 1999–2001. The model indicated that assumptions regarding the form of the inventory, whether dissolved or in solid phase form, did not appreciably affect export rates. Also, the percentage of non-exchangeable adsorbed 90Sr, which is uncertain and affects the amount of 90Sr available for export, was fixed at 20% based on field data measurements. A Monte Carlo uncertainty analysis was conducted treating Kd as an uncertain input variable with a range of 100–300 L/kg. This analysis resulted in a range of 0.13–0.27% of inventory exported to surface water compared to 0.14% based on measured field data.Based on this model application, it was concluded that the export of 90Sr from the Borschi watershed to surface water is predominantly a result of soil pore water containing dissolved 90Sr being diverted to surface waters that eventually flow out of the watershed. The percentage of non-exchangeable adsorbed 90Sr and the soil-water Kd are the two most sensitive and uncertain factors affecting the amount of export. The 200-year projections of the model showed an exponential decline in 90Sr export fluxes from the watershed that should drop by a factor of 10 by the year 2100.This presentation will focus on TREECS capabilities and the case study done for the Borschi Watershed.