• Communications for Statistical Applications and Methods
• /
• v.29 no.1
• /
• pp.41-51
• /
• 2022

We forecast the US oil consumption level taking advantage of google trends. The google trends are the search volumes of the specific search terms that people search on google. We focus on whether proper selection of google trend terms leads to an improvement in forecast performance for oil consumption. As the forecast models, we consider the least absolute shrinkage and selection operator (LASSO) regression and the structured regularization method for large vector autoregressive (VAR-L) model of Nicholson et al. (2017), which select automatically the google trend terms and the lags of the predictors. An out-of-sample forecast comparison reveals that reducing the high dimensional google trend data set to a low-dimensional data set by the LASSO and the VAR-L models produces better forecast performance for oil consumption compared to the frequently-used forecast models such as the autoregressive model, the autoregressive distributed lag model and the vector error correction model.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.25 no.4
• /
• pp.368-375
• /
• 2023

In this study, we examined the error characteristic and bias correction method for one-month temperature forecast data produced through joint development between the Rural Development Administration and the H ong Kong University of Science and Technology. For this purpose, hindcast data from 2013 to 2021, weather observation data, and various environmental information were collected and error characteristics under various environmental conditions were analyzed. In the case of maximum and minimum temperatures, the higher the elevation and latitude, the larger the forecast error. On average, the RMSE of the forecast data corrected by the linear regression model and the XGBoost decreased by 0.203, 0.438 (maximum temperature) and 0.069, 0.390 (minimum temperature), respectively, compared to the uncorrected forecast data. Overall, XGBoost showed better error improvement than the linear regression model. Through this study, it was found that errors in prediction data are affected by topographical conditions, and that machine learning methods such as XGBoost can effectively improve errors by considering various environmental factors.

• Atmosphere
• /
• v.27 no.1
• /
• pp.93-104
• /
• 2017

The effect of Advanced Microwave Sounding Unit-A (AMSU-A) observations on the short-range forecast in East Asia (EA) was investigated for the Northern Hemispheric (NH) summer and winter months, using the Forecast Sensitivity to Observations (FSO) method. For both periods, the contribution of radiosonde (TEMP) to the EA forecast was largest, followed by AIRCRAFT, AMSU-A, Infrared Atmospheric Sounding Interferometer (IASI), and the atmospheric motion vector of Communication, Ocean and Meteorological Satellite (COMS) or Multi-functional Transport Satellite (MTSAT). The contribution of AMSU-A sensor was largely originated from the NOAA 19, NOAA 18, and MetOp-A (NOAA 19 and 18) satellites in the NH summer (winter). The contribution of AMSU-A sensor on the MetOp-A (NOAA 18 and 19) satellites was large at 00 and 12 UTC (06 and 18 UTC) analysis times, which was associated with the scanning track of four satellites. The MetOp-A provided the radiance data over the Korea Peninsula in the morning (08:00~11:30 LST), which was important to the morning forecast. In the NH summer, the channel 5 observations on MetOp-A, NOAA 18, 19 along the seaside (along the ridge of the subtropical high) increased (decreased) the forecast error slightly (largely). In the NH winter, the channel 8 observations on NOAA 18 (NOAA 15 and MetOp-A) over the Eastern China (Tibetan Plateau) decreased (increased) the forecast error. The FSO provides useful information on the effect of each AMSU-A sensor on the EA forecasts, which leads guidance to better use of AMSU-A observations for EA regional numerical weather prediction.

• Atmosphere
• /
• v.19 no.2
• /
• pp.213-226
• /
• 2009

In order to investigate the characteristics of Changma over the Korean peninsula, KEOP-2007 IOP (Intensive Observing Period) was conducted from 15 June 2007 to 15 July 2007. KEOP-2007 IOP is high spatial and temporal radiosonde observations (RAOB) which consisted of three special stations (Munsan, Haenam, and Ieodo) from National Institute of Meteorological Research, five operational stations (Sokcho, Baengnyeongdo, Pohang, Heuksando, and Gosan) from Korea Meteorological Administration (KMA), and two operational stations (Osan and Gwangju) from Korean Air Force (KAF) using four different types of radiosonde sensors. The error statistics of the sensor of radiosonde were investigated using quality control check. The minimum and maximum error frequency appears at the sensor of RS92-SGP and RS1524L respectively. The error frequency of DFM-06 tends to increase below 200 hPa but RS80-15L and RS1524L show vice versa. Especially, the error frequency of RS1524L tends to increase rapidly over 200 hPa. Systematic biases of radiosonde show warm biases in case of temperature and dry biases in case of relative humidity compared with ECMWF (European Center for Medium-Range Weather Forecast) analysis data and precipitable water vapor from GPS. The maximum and minimum values of systematic bias appear at the sensor of DFM-06 and RS92-SGP in case of temperature and RS80-15L and DFM-06 in case of relative humidity. The systematic warm and dry biases at all sensors tend to increase during daytime than nighttime because air temperature around sensor increases from the solar heating during daytime. Systematic biases of radiosonde are affected by the sensor type and the height of the sun but random errors are more correlated with the moisture conditions at each observation station.

• Atmosphere
• /
• v.17 no.4
• /
• pp.327-337
• /
• 2007

In this study, we examined the new ensemble training approach to reduce the systematic error and improve prediction skill of wind by using the Short-range Ensemble prediction system (SENSE), which is the mesoscale multi-model ensemble prediction system. The SENSE has 16 ensemble members based on the MM5, WRF ARW, and WRF NMM. We evaluated the skill of surface wind prediction compared with AWS (Automatic Weather Station) observation during the summer season (June - August, 2006). At first stage, the correction of initial state for each member was performed with respect to the observed values, and the corrected members get the training stage to find out an adaptive weight function, which is formulated by Root Mean Square Vector Error (RMSVE). It was found that the optimal training period was 1-day through the experiments of sensitivity to the training interval. We obtained the weighted ensemble average which reveals smaller errors of the spatial and temporal pattern of wind speed than those of the simple ensemble average.

• Korean Journal of Organic Agriculture
• /
• v.25 no.2
• /
• pp.279-293
• /
• 2017

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.

• Ocean and Polar Research
• /
• v.34 no.2
• /
• pp.185-195
• /
• 2012

This paper focused on forecasting a short-term production of oysters, which have been farmed in Korea, with distinct periodicity of production by year, and different production level by month. To forecast a short-term oyster production, this paper uses monthly data (260 observations) from January 1990 to August 2011, and also adopts several econometrics methods, such as Multiple Regression Analysis Model (MRAM), Seasonal Autoregressive Integrated Moving Average (SARIMA) Model, and Vector Error Correction Model (VECM). As a result, first, the amount of short-term oyster production forecasted by the multiple regression analysis model was 1,337 ton with prediction error of 246 ton. Secondly, the amount of oyster production of the SARIMA I and II models was forecasted as 12,423 ton and 12,442 ton with prediction error of 11,404 ton and 11,423 ton, respectively. Thirdly, the amount of oyster production based on the VECM was estimated as 10,425 ton with prediction errors of 9,406 ton. In conclusion, based on Theil inequality coefficient criterion, short-term prediction of oyster by the VECM exhibited a better fit than ones by the SARIMA I and II models and Multiple Regression Analysis Model.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.58 no.9
• /
• pp.1700-1705
• /
• 2009

In general, short term load forecasting is based on the periodical load pattern during a day or a week. Therefore, the conventional methods do not expose stable performance to every day during a year. Especially for anomalous weather conditions such as typhoons, the methods have a tendency to show the conspicuous accuracy deterioration. Furthermore, the tendency raises the reliability and stability problems of the conventional load forecast. In this study, a new load forecasting method is proposed in order to increase the accuracy of the forecast result in case of anomalous weather conditions such as typhoons. For irregular weather conditions, the sensitivity between temperature and daily load is used to improve the accuracy of the load forecast. The proposed method was tested with the actual load profiles during 14 years, which shows that the suggested scheme considerably improves the accuracy of the load forecast results.

• Journal of the Korean Solar Energy Society
• /
• v.39 no.3
• /
• pp.59-66
• /
• 2019

New and renewable energy forecasts are key technology to reduce the annual operating cost of new and renewable facilities, and accuracy of forecasts is paramount. In this study, we intend to build a model for the prediction of short-term solar power generation for 1 hour to 3 hours. To this end, this study applied two time series technique, ARIMA model without considering seasonality and SARIMA model with considering seasonality, comparing which technique has better predictive accuracy. Comparing predicted errors by MAE measures of solar power generation for 1 hour to 3 hours at four locations, the solar power forecast model using ARIMA was better in terms of predictive accuracy than the solar power forecast model using SARIMA. On the other hand, a comparison of predicted error by RMSE measures resulted in a solar power forecast model using SARIMA being better in terms of predictive accuracy than a solar power forecast model using ARIMA.

• Atmosphere
• /
• v.28 no.3
• /
• pp.291-303
• /
• 2018

In this study, a new consensus technique for predicting tropical cyclone (TC) intensity in the western North Pacific was developed. The most important feature of the present consensus model is to select and combine the guidance numerical models with the best performance in the previous years based on various evaluation criteria and averaging methods. Specifically, the performance of the guidance models was evaluated using both the mean absolute error and the correlation coefficient for each forecast lead time, and the number of the numerical models used for the consensus model was not fixed. In averaging multiple models, both simple and weighted methods are used. These approaches are important because that the performance of the available guidance models differs according to forecast lead time and is changing every year. In particular, this study develops both a multi-consensus model (M-CON), which constructs the best consensus models with the lowest error for each forecast lead time, and a single best consensus model (S-CON) having the lowest 72-hour cumulative mean error, through on training process. The evaluation results of the selected consensus models for the training and forecast periods reveal that the M-CON and S-CON outperform the individual best-performance guidance models. In particular, the M-CON showed the best overall performance, having advantages in the early stages of prediction. This study finally suggests that forecaster needs to use the latest evaluation results of the guidance models every year rather than rely on the well-known accuracy of models for a long time to reduce prediction error.