• Journal of Environmental Impact Assessment
• /
• v.29 no.4
• /
• pp.272-285
• /
• 2020

We investigated statistical evaluation parameters for 3D meteorological and air quality models and selected several quantitative indicator references, and summarized the reference values of the statistical parameters for domestic air quality modeling researcher. The finally selected 9 statistical parameters are MB (Mean Bias), ME (Mean Error), MNB (Mean Normalized Bias Error), MNE (Mean Absolute Gross Error), RMSE (Root Mean Square Error), IOA (Index of Agreement), R (Correlation Coefficient), FE (Fractional Error), FB (Fractional Bias), and the associated reference values are summarized. The results showed that MB and ME have been widely used in evaluating the meteorological model output, and NMB and NME are most frequently used for air quality model results. In addition, discussed are the presentation diagrams such as Soccer Plot, Taylor diagram, and Q-Q (Quantile-Quantile) diagram. The current results from our study is expected to be effectively used as the statistical evaluation parameters suitable for situation in Korea considering various characteristics such as including the mountainous surface areas.

• Journal of Korean Society for Atmospheric Environment
• /
• v.33 no.1
• /
• pp.11-18
• /
• 2017

The performance of a modeling system consisting of WRF model v3.3 and CMAQ model v4.7.1 for forecasting $PM_{2.5}$ concentrations were evaluated during the period May 2012 through December 2014. Twenty-four hour averages of $PM_{2.5}$ and its major components obtained through filter sampling at the Bulgwang intensive measurement station were used for comparison. The mean predicted $PM_{2.5}$ concentration over the entire period was 68% of the mean measured value. Predicted concentrations for major components were underestimated except for $NO_3{^-}$. The model performance for $PM_{2.5}$ generally tended to degrade with increasing the concentration level. However, the mean fractional bias (MFB) for high concentration above the $80^{th}$ percentile fell within the criteria, the level of accuracy acceptable for standard model applications. Among three bias correction methods, the ratio adjustment was generally most effective in improving the performance. Albeit for limited test conditions, this analysis demonstrated that the effects of bias correction were larger when using the data with a larger bias of predicted values from measurement values.

• Macromolecular Research
• /
• v.15 no.6
• /
• pp.491-497
• /
• 2007

NPT Monte Carlo simulations were performed to calculate the molecular properties of syndiotactic poly(vinyl chloride) (PVC) and syndiotactic poly(vinyl alcohol) (PVA) melts using the configurational bias Monte Carlo move, concerted rotation, reptation, and volume fluctuation. The density, mean square backbone end-to-end distance, mean square radius of gyration, fractional free-volume distribution, distribution of torsional angles, small molecule solubility constant, and radial distribution function of PVC at 0.1 MPa and above the glass transition temperature were calculated/measured, and those of PVA were calculated. The calculated results were compared with the corresponding experimental data and discussed. The calculated densities of PVC and PVA were smaller than the experimental values, probably due to the very low molecular weight of the model polymer used in the simulation. The fractional free-volume distribution and radial distribution function for PVC and PVA were nearly independent of temperature.

• Environmental Engineering Research
• /
• v.24 no.4
• /
• pp.572-581
• /
• 2019

The emission of hydrogen sulfide (H₂S) from the Kahrizak landfill was studied. Firstly, the field measurements were conducted in the summer and winter seasons; and the samples were analyzed using Jacob method. We predicted the H₂S concentrations in the downwind using AERMOD and ISCST3. According to the AERMOD, the maximum concentration of H₂S in the summer and winter were 117 ㎍/㎥ and 205 ㎍/㎥ respectively. The downwind concentrations reached zero at the distance of 35 km from the leachate treatment plant. The Geometric mean bias, Geometric variance, Fractional bias, Fraction of predictions within a factor of two of the observations and Normalized mean square error for the AERMOD were 0.58, 1.35, -0.12, 1.91 and 0.042, respectively in the summer and 1.39, 1.35, -0.05, 1.46 and 0.027 in the winter; and for the ISCST3, were 0.85, 1.03, 0.02, 1.45 and 0.04 in the summer and 1.18, 1.03, 0.15, 1.16 and 0.04 in the winter. The results of the AERMOD were compared with the ISCST3 and indicated that the AERMOD performance was more suitable than the ISCST3.

• The Korean Journal of Applied Statistics
• /
• v.15 no.2
• /
• pp.395-403
• /
• 2002

In the third phase of the response surface methods, the first-order model is assumed and the curvature of the response surface is checked with a fractional factorial design augmented by centre runs. We further assume that a true model is a quadratic polynomial. To choose an optimal design, Box and Draper(1959) suggested the use of an average mean squared error (AMSE), an average of MSE of y(x) over the region of interest R. The AMSE can be partitioned into the average prediction variance (APV) and average squared bias (ASB). Since AMSE is a function of design moments, region moments and a standardized vector of parameters, it is not possible to select the design that minimizes AMSE. As a practical alternative, Box and Draper(1959) proposed minimum bias design which minimize ASB and showed that factorial design points are shrunk toward the origin for a minimum bias design. In this paper we propose a robust AMSE design which maximizes the minimum efficiency of the design with respect to a standardized vector of parameters.

• Journal of Environmental Science International
• /
• v.12 no.7
• /
• pp.809-816
• /
• 2003

The short-term version of Industrial Source Complex Model(ISCST3) was evaluated for estimating short-term concentrations using criteria pollutant(SO$_2$, NO$_2$, CO, PM10) data from emission inventory of Young Tong area in Suwon for the year 2002. The contribution of pollutant concentration from point, line, area sources was found 21.8, 76.5 and 1.6%. Statistical parameters, such as correlation coefficient, index of agreement(IA), normalized mean square error(NMSE) and fractional bias(FB) were calculated for each pollutants. The model performance were found good for PM10(82%) and NO$_2$(69%), but poor for SO$_2$(34%) and CO(13%).

• Asian-Australasian Journal of Animal Sciences
• /
• v.25 no.9
• /
• pp.1237-1247
• /
• 2012

The objective of this study was to evaluate the predictions of dry matter intake (DMI) and average daily gain (ADG) of Vietnamese Yellow (Vang) purebred and crossbred (Vang with Red Sindhi or Brahman) bulls fed under Vietnamese conditions using two levels of solution (1 and 2) of the large ruminant nutrition system (LRNS) model. Animal information and feed chemical characterization were obtained from five studies. The initial mean body weight (BW) of the animals was 186, with standard deviation ${\pm}33.2$ kg. Animals were fed ad libitum commonly available feedstuffs, including cassava powder, corn grain, Napier grass, rice straw and bran, and minerals and vitamins, for 50 to 80 d. Adequacy of the predictions was assessed with the Model Evaluation System using the root of mean square error of prediction (RMSEP), accuracy (Cb), coefficient of determination ($r^2$), and mean bias (MB). When all treatment means were used, both levels of solution predicted DMI similarly with low precision ($r^2$ of 0.389 and 0.45 for level 1 and 2, respectively) and medium accuracy (Cb of 0.827 and 0.859, respectively). The LRNS clearly over-predicted the intake of one study. When this study was removed from the comparison, the precision and accuracy considerably increased for the level 1 solution. Metabolisable protein was limiting ADG for more than 68% of the treatment averages. Both levels differed regarding precision and accuracy. While level 1 solution had the least MB compared with level 2 (0.058 and 0.159 kg/d, respectively), the precision was greater for level 2 than level 1 (0.89 and 0.70, respectively). The accuracy (Cb) was similar between level 1 and level 2 (p = 0.8997; 0.977 and 0.871, respectively). The RMSEP indicated that both levels were on average under-or over-predicted by about 190 g/d, suggesting that even though the accuracy (Cb) was greater for level 1 compared to level 2, both levels are likely to wrongly predict ADG by the same amount. Our analyses indicated that the level 1 solution can predict DMI reasonably well for this type of animal, but it was not entirely clear if animals consumed at their voluntary intake and/or if the roughness of the diet decreased DMI. A deficit of ruminally-undegradable protein and/or a lack of microbial protein may have limited the performance of these animals. Based on these evaluations, the LRNS level 1 solution may be an alternative to predict animal performance when, under specific circumstances, the fractional degradation rates of the carbohydrate and protein fractions are not known.