In this study, the prediction technology of Hydrological Quantitative Precipitation Forecast (HQPF) was improved by optimizing the weather predictors used as input data for machine learning. Results comparison was conducted using bias and Root Mean Square Error (RMSE), which are predictive accuracy verification indicators, based on the heavy rain case on August 21, 2021. By comparing the rainfall simulated using the improved HQPF and the observed accumulated rainfall, it was revealed that all HQPFs (conventional HQPF and improved HQPF 1 and HQPF 2) showed a decrease in rainfall as the lead time increased for the entire grid region. Hence, the difference from the observed rainfall increased. In the accumulated rainfall evaluation due to the reduction of input factors, compared to the existing HQPF, improved HQPF 1 and 2 predicted a larger accumulated rainfall. Furthermore, HQPF 2 used the lowest number of input factors and simulated more accumulated rainfall than that projected by conventional HQPF and HQPF 1. By improving the performance of conventional machine learning despite using lesser variables, the preprocessing period and model execution time can be reduced, thereby contributing to model optimization. As an additional advanced method of HQPF 1 and 2 mentioned above, a simulated analysis of the Local ENsemble prediction System (LENS) ensemble member and low pressure, one of the observed meteorological factors, was analyzed. Based on the results of this study, if we select for the positively performing ensemble members based on the heavy rain characteristics of Korea or apply additional weights differently for each ensemble member, the prediction accuracy is expected to increase.

The study aims to provide better understanding of sustainable earnings by a comprehensive analysis of earnings persistence of business firms in Vietnam as an example of developing economies in South-East Asia. Dataset of 1,278 publicly listed firms (excluding banking and financial services firms) on Vietnam Stock Exchange for the period from 2008 to 2017 was collected. By applying fixed effect regression model, the empirical results provided the basis to measure the persistence index (Pers index) and find low level of their earnings persistence. The literature of earnings quality analysis in developed countries suggests earnings persistence as a noteworthy determinant of future earnings forecast and stock valuation. However, research of sustainable earnings in developing countries is still highly underdeveloped. For Vietnamese listed firms, the average Pers index was estimated for the period from 2008 to 2010, indicating low level of earnings persistence. We also incorporated earnings persistence level into future earnings forecast by running the quintile regression model divided the data into four equal levels and conducted each section independently to see the difference in each percentile, thence assessed the factors’ influence on the specific model. The findings provide important information on the expected returns of firms, especially helping investors make sound decisions.

We have developed an algorithm for tracking coronal mass ejection (CME) propagation that allows us to estimate CME speed and its arrival time at Earth. The algorithm may be used either to forecast the CME’s arrival on the day of the forecast or to update the CME tracking information for the next day’s forecast. In our case study, we successfully tracked CME propagation using the algorithm based on g-values of interplanetary scintillation (IPS) observation provided by the Institute for Space- Earth Environmental Research (ISEE). We were able to forecast the arrival time (Δt = 0.30 h) and speed (Δv = 20 km/s) of a CME event on October 2, 2000. From the CME-interplanetary CME (ICME) pairs provided by Cane & Richardson (2003), we selected 50 events to evaluate the algorithm’s forecast capability. Average errors for arrival time and speed were 11.14 h and 310 km/s, respectively. Results demonstrated that g-values obtained continuously from any single station observation were able to be used as a proxy for CME speed. Therefore, our algorithm may give stable daily forecasts of CME position and speed during propagation in the region of 0.2–1 AU using the IPS g-values, even if IPS velocity observations are insufficient. We expect that this algorithm may be widely accepted for use in space weather forecasting in the near future.

We examine the effects of the complexity of tax-related information on the issuance of analyst’s pre-tax income forecast and its value relevance. If analysts respond adequately to the needs of investors, they are more likely to provide a pre-tax income forecast. The provision of a pre-tax income forecast may indicate analysts’ confidence in assessing the quality of earnings. Thus, investors, in turn, would be more confident in the analysts’ pre-tax income forecasts if analysts provide both pre-tax and earnings forecasts than only the latter. Using a sample of Korean listed companies for 2005–2014, we find that analysts are likely to provide an implicit tax forecast when the volatility of the effective tax rate is low and the book-tax differences are small. We also find that when analysts provide pretax and after tax income forecasts, the value relevance for unexpected earnings increases. These results indicate that analysts are likely to be interested in corporate tax information and the complexity of tax-related information affects the availability of implicit tax forecasts. Furthermore, this study provides empirical evidence that when analysts provide both pre-tax and after tax income forecasts, investors have more confidence in analysts’ earnings forecasts, which results in greater investors’ responses.

Dividend policy is an important business decision and is considered a channel to communicate a firm’s performance to shareholders. Given the empirical findings that earnings quality significantly affects financial analysts’ forecasting activities, it is predicted that higher earnings quality would positively influence forecast accuracy. Specifically, it is expected that financial analysts would forecast dividends more accurately for firms with higher earning quality. Unlike the research on financial analysts’ earnings forecasts was heavily conducted, there is little study about financial analysts’ dividend forecasts. This paper examines the effect of earnings quality on financial analysts’ dividend forecast accuracy. We use a sample of South Korean firms for the period of 2011–2015 for multivariate regression. Earnings quality is measured by accruals quality and performance-adjusted discretionary accruals followed by prior studies. We first compare the accuracy between dividend forecasts and earnings forecasts using t-test and Wilcoxon singed-rank test. It is confirmed that financial analysts’ dividend forecasts are more accurate than earnings forecasts in Korea. We find that financial analysts’ dividend forecasts are more accurate for firms with higher earnings quality. We also find that the result is still valid after controlling for the accuracy of financial analysts’ earnings forecasts. This confirms that earnings quality positively affects financial analysts’ dividend forecasts.

K패션으로 명칭되며 글로벌화를 시도하고 있는 국내 패션 브랜드가 늘어나는 추세다. 국내 브랜드의 해외 진출을 더욱 성공적으로 이끌기 위하여 해외 소비자들의 실질적 수요에 대한 정확한 분석이 필요하다. 이에 본 연구에서는 서울과 파리의 로컬 패션인 스트리트 패션을 비교하고 두 도시의 패션 스타일을 분석하여 공통점과 차이점을 파악, 이 결과와 트렌드 예측 정보와의 간극을 살펴봄으로써 트렌드 예측 정보가 스트리트 패션에 수용되는 국·내외 특성을 살펴보고자 한다. 본 연구의 내용은 첫째, 스트리트 패션과 패션 트렌드 예측 정보의 관점에서 연구 방법론을 모색하였다. 둘째, 2019년 S/S 패션 트렌드 관점에서 서울과 파리의 스트리트 패션을 파악하였다. 셋째, 패션 트렌드 예측 정보 수용의 관점에서 서울과 파리 스트리트 패션의 수용 정도 및 특성을 도출하였다. 연구 결과, 서울과 파리, 두 도시에서 일반적인 스트리트 스타일이 2019 S/S 트렌드로 예측된 스타일보다 높은 비율로 나타났다. 도시 별 스타일 선호에서도 차이가 나타났는데, 서울과 파리 두 도시에서 공통적으로는 클래식이 가장 높게 나타났고, 캐주얼과 페미닌, 키치 순으로 뒤를 이었다. 패션 트렌드 예측 정보에 대한 소비자 수용 현황을 볼 때, 두 도시의 소비자들은 패션 정보 업체가 예측한 트렌드 테마 중 한정된 몇가지 테마에 집중하는 양상을 보였다.

In this study, we compared the prediction performances according to the bias and dispersion of temperature using ensemble machine learning. Ensemble machine learning is meta-algorithm that combines several base learners into one prediction model in order to improve prediction. Multiple linear regression, ridge regression, LASSO (Least Absolute Shrinkage and Selection Operator; Tibshirani, 1996) and nonnegative ride and LASSO were used as base learners. Super learner (van der Lann et al ., 1997) was used to produce one optimal predictive model. The simulation and real data for temperature were used to compare the prediction skill of machine learning. The results showed that the prediction performances were different according to the characteristics of bias and dispersion and the prediction error was more improved in temperature with bias compared to dispersion. Also, ensemble machine learning method showed similar prediction performances in comparison to the base learners and showed better prediction skills than the ensemble mean.

This study examines whether auditors restrain the analysts’ opportunistic behavior as reviewing the companies’ interim reports. Analysts' forecasts show a walkdown pattern in which their optimism has decreased as the earnings announcement date has approached. At the beginning of the year, there is a lack of high-quality benchmark information that enables information users to judge the accuracy of analyst’s earnings forecasts. Thus, early in the year, analysts are highly inspired to disseminate optimistic forecasts in order to gain manager’s favor. In this study, we examine adequate benchmarks prevent analysts from disclosing optimistically biased forecasts. We conjecture that auditors’ efforts might mitigate analysts’ walkdown pattern. To test this hypothesis, we use data from Korea, where it is mandatory to disclose auditor’s review hours. We find that the analyst forecast’s walkdown decreases with the ratio as well as the number of audit hours. It implies that an auditor's effort in reviewing interim financial information has a monitoring function that reduces analysts' opportunistic optimism at the beginning of the year. We conjecture that the tendency will be more pronounced when BIG4 auditors review the interim reports. Consistent with the prediction, BIG4 auditors’ interim review effort is more effective in suppressing the analysts’ walkdown.

Unit load factor, which is used for the quantification of non-point pollution in watersheds, has the limitation that it does not reflect spatial characteristics of soil, topography and temporal change due to the interannual or seasonal variability of precipitation. Therefore, we developed the method to estimate a watershed-scale non-point pollutant load using seasonal forecast data that forecast changes of precipitation up to 6 months from present time for watershed-scale water quality management. To establish a preemptive countermeasure against non-point pollution sources, it is possible to consider the unstructured management plan which is possible over several months timescale. Notably, it is possible to apply various management methods such as control of sowing and irrigation timing, control of irrigation through water management, and control of fertilizer through fertilization management. In this study, APEX-Paddy model, which can consider the farming method in field scale, was applied to evaluate the applicability of seasonal forecast data. It was confirmed that the rainfall amount during the growing season is an essential factor in the non-point pollution pollutant load. The APEX-Paddy model for quantifying non-point pollution according to various farming methods in paddy fields simulated similarly the annual variation tendency of TN and TP pollutant loads in rice paddies but showed a tendency to underestimate load quantitatively.

본 연구에서는 ANFIS 기반 GloSea5 앙상블 기상전망 개선 기법을 개발하고 평가하였다. 대상유역은 국내 주요 다목적댐인 충주댐 유역을 선정하였으며, 개선 기법은 ANFIS 기반의 전·후처리기법으로 구성된다. 전처리 기법에서 GloSea5의 앙상블 멤버에 가중치를 부여하며(OWM), 후처리 과정에서는 전처리결과를 편의보정 한다(MOS). 평가결과 편의보정된 GloSea5에 비해 예측성능이 개선되었으며, CASE3, CASE1, CASE2 순으로 모의성능이 우수하였다. 전처리 기법은 강수의 변동성이 큰 계절에 개선효과가 우수하였으며, 후처리 기법은 전처리로 개선하지 못한 오차를 줄일 수 있는 것으로 나타났다. 따라서 본 연구에서 개발한 ANFIS 기반 GloSea5 앙상블 기상전망 개선 기법은 전·후처리 기법을 함께 사용하는 것이 가장 좋으며, 특히 여름철과 같이 강수의 변동성이 큰 계절에 활용성이 높을 것으로 판단된다.

본 연구의 목적은 기상자료(강수량, 최고기온, 최저기온, 평균기온, 평균풍속) 기반의 다중선형 회귀모형을 개발하여 농업용저수지 저수율을 예측 하는 것이다. 나이브 베이즈 분류를 활용하여 전국 1,559개의 저수지를 지리형태학적 제원(유효저수량, 수혜면적, 유역면적, 위도, 경도 및 한발빈도)을 기준으로 30개 군집으로 분류하였다. 각 군집별로, 기상청 기상자료와 한국농어촌공사 저수지 저수율의 13년(2002~2014) 자료를 활용하여 월별 회귀모형을 유도하였다. 저수율의 회귀모형은 결정계수(R2)가 0.76, Nash-Sutcliffe efficiency (NSE)가 0.73, 평균제곱근오차가 8.33%로 나타났다. 회귀모형은 2년(2015~2016) 기간의 기상청 3개월 기상전망자료인 GloSea5 (GS5)를 사용하여 평가되었다. 현재저수율과 평년저수율에 의해 산정되는 저수지 가뭄지수(Reservoir Drought Index, RDI)에 의한 ROC (Receiver Operating Characteristics) 분석의 적중률은 관측값을 이용한 회귀식에서 0.80과 GS5를 이용한 회귀식에서 0.73으로 나타났다. 본 연구의 결과를 이용해 미래 저수율을 전망하여 안정적인 미래 농업용수 공급에 대한 의사결정 자료로 사용할 수 있을 것이다.

In this study, to investigate an optimal configuration method for the modeling system, we performed an optimization experiment by controlling the types of compilers and libraries, and the number of CPU cores because it was important to provide reliable model data very quickly for the national air quality forecast. We were made up the optimization experiment of twelve according to compilers (PGI and Intel), MPIs (mvapich-2.0, mvapich-2.2, and mpich-3.2) and NetCDF (NetCDF-3.6.3 and NetCDF-4.1.3) and performed wall clock time measurement for the WRF and CMAQ models based on the built computing resources. In the result of the experiment according to the compiler and library type, the performance of the WRF (30 min 30 s) and CMAQ (47 min 22 s) was best when the combination of Intel complier, mavapich-2.0, and NetCDF-3.6.3 was applied. Additionally, in a result of optimization by the number of CPU cores, the WRF model was best performed with 140 cores (five calculation servers), and the CMAQ model with 120 cores ( five calculation servers). While the WRF model demonstrated obvious differences depending on the number of CPU cores rather than the types of compilers and libraries, CMAQ model demonstrated the biggest differences on the combination of compilers and libraries.

최근 이상기후로 인한 집중호우 발생빈도와 이로 인한 국지적인 홍수 피해가 증가하고 있다. 이러한 점에서 홍수피해 예방측면에서 수치예보 정보 활용이 요구되고 있다. 그러나 수치예보모델은 초기 조건 및 지형적 요인으로 인해 시공간적 편의가 존재하며 실시간 예측정보로 활용하기 전에 모 형결과에 대한 편의보정이 요구된다. 본 연구에서는 관측지점 기준으로 편의 보정계수를 산정하는 과정에서 모든 관측소간의 상관성을 거리의 함 수로 고려하여 미계측지점의 편의 보정계수를 공간적으로 확장할 수 있는 Bayesian Kriging 기반 MFBC 기법을 개발하였다. 본 연구에서 개발한 방법은 미계측 유역에 대해서도 보정계수를 효과적으로 추정하는 것이 확인되었으며, 비교적 고해상도로 72시간(3일) 정도까지 예측강우 정보를 활용하는 것이 가능할 것으로 판단된다.

This study used a quantile regression model and a non-homogeneous regression model to calibrate probabilistic forecasts of wind speed. These techniques were applied to the forecasts of wind speed over Pyeongchang area using 51-member European Centre for Medium-Range Weather Forecast (ECMWF). Reliability analysis was carried out by using rank histogram to identify the statistical consistency of ensemble forecasts and corresponding observations. The performances were evaluated by rank histogram, mean absolute error, root mean square error and continuous ranked probability score. The results showed that the forecasts of quantile regression and non-homogeneous regression models performed better than the raw ensemble forecasts. However, the differences of prediction skills between quantile regression and nonhomogeneous regression models were insignificant.

In this paper, we used a nonhomogeneous Gaussian regression model (NGR) as the postprocessing techniques to calibrate probabilistic forecasts that take the form of probability density functions for temperature. We also performed the alternative implementation techniques of NGR, which are stationspecific ensemble model output statistics (EMOS) model. These techniques were applied to forecast temperature over Pyeongchang area using 24-member Ensemble Prediction System for Global (EPSG). The results showed that the station-specific EMOS model performed better than the raw ensemble and EMOS model.

This study suggests the yield forecast model for chilli pepper using artificial neural network. For this, we select the most suitable network models for chilli pepper’s yield and compare the predictive power with adaptive expectation model and panel model. The results show that the predictive power of artificial neural network with 5 weather input variables (temperature, precipitation, temperature range, humidity, sunshine amount) is higher than the alternative models. Implications for forecasting of yields are suggested at the end of this study.

2014년부터 기상청에서 현업으로 활용하고 있는 전지구 계절예측시스템 GloSea5의 최대 6개월 예측 강수량을 수자원 및 여러 응용분야에 활 용하기 위해서는 예측모델이 가지는 관측자료와의 정량적인 편의를 보정할 필요가 있다. 본 연구에서는 GloSea5의 예측 강수량에서 나타나는 편 의를 보정하기 위해 확률분포형을 활용한 편의보정기법, 매개변수 및 비매개변수적 편의보정기법 등 총 11개의 기법을 활용하여 계절예측모델의 적용성을 평가하고 최적의 편의보정기법을 선정하고자 하였다. 과거재현기간에 대한 편의보정 결과, 비매개변수적 편의보정기법이 다른 기법에 비해 가장 관측자료와 유사하게 보정하는 것으로 분석되었으나 예측기간에 대해서는 상대적으로 많은 이상치를 발생시켰다. 이와는 대조적으로 매개변수적 편의보정기법은 과거재현기간 및 예측기간 모두 안정된 결과를 보여주고 있음을 확인할 수 있었다. 본 연구의 결과는 수자원운영 및 관 리, 수력, 농업 등 계절예측모델을 활용한 여러 응용분야에 적용이 가능할 것으로 기대된다.

교량 바닥판은 대형 차량 및 제설제와 같은 다양한 환경 요인으로 인해 급속히 악화되는 부재이다. 한국에 건설된 교량의 수명이 길 어짐에 따라 교량 바닥판의 교체 수요가 증가 할 것으로 예상된다. 다른 국가에서는 프리 캐스트 바닥판을 이용한 급속 교량 건설 기술이 열화 된 교량 바닥판의 교체 수요 대응을 위해 적극적으로 사용되고 있다. 본 연구에서는 국내 교량 바닥판의 상태평가 데이터를 수집 및 분석하여 교량 바닥판 열화 모델을 제안 하였다. 또한 교량 규모의 관점에서 열화된 교량 바닥판의 미래 대체 수요를 예측하였다.

This study suggests the yield forecast models for autumn chinese cabbage and radish using crop growth and development information. For this, we construct 24 alternative yield forecast models and compare the predictive power using root mean square percentage errors. The results shows that the predictive power of model including crop growth and development informations is better than model which does not include those informations. But the forecast errors of best forecast models exceeds 5%. Thus it is important to establish reliable data and improve forecast models.

A statistical forecast model for early spring (March and April) precipitation over South Korea is developed by using multiple linear regression method. Predictors are selected among the forty five large-scale atmospheric and oceanic indices. Because the model is meant to use for real-time forecast, the predictors are chosen from the indices that have statistically significant lag correlation with observed early spring precipitation. The selected predictors of early spring precipitation are North Pacific Pattern with 6-month lead, Siberian High Index with 5-month lead and Indian Ocean Basin Mode Index with 3-month lead from March, and they are statistically independent. We applied leave-two-out cross validation. According to the regression map between these indices and synoptic circulations around Korean peninsula, these indices represent the induction of early spring rainfall by controlling East Asian jet and low level moisture flux. The regression coefficients for each training period show that three indices affects evenly at every forecast year and they show stable variability, indicating that the influence of each index does not depend on training period. The developed statistical model significantly predicted early spring precipitation over South Korea (r=0.63, p-value<0.01). Also it marks 61% of hit rate according to the three-category deterministic forecast.