• Proceedings of the Korea Water Resources Association Conference
• /
• 2021.06a
• /
• pp.373-373
• /
• 2021

Rainfall forecasting is an important issue that is applied in many areas, such as agriculture, flood warning, and water resources management. In this context, this study proposed a statistical and machine learning-based forecasting model for monthly rainfall. The Bayesian Gaussian process was chosen to optimize the hyperparameters of the Stacked Long Short-term memory (SLSTM) model. The proposed SLSTM model was applied for predicting monthly precipitation of Seoul station, South Korea. Data were retrieved from the Korea Meteorological Administration (KMA) in the period between 1960 and 2019. Four schemes were examined in this study: (i) prediction with only rainfall; (ii) with deseasonalized rainfall; (iii) with rainfall and minimum temperature; (iv) with deseasonalized rainfall and minimum temperature. The error of predicted rainfall based on the root mean squared error (RMSE), 16-17 mm, is relatively small compared with the average monthly rainfall at Seoul station is 117mm. The results showed scheme (iv) gives the best prediction result. Therefore, this approach is more straightforward than the hydrological and hydraulic models, which request much more input data. The result indicated that a deep learning network could be applied successfully in the hydrology field. Overall, the proposed method is promising, given a good solution for rainfall prediction.

• Journal of Wetlands Research
• /
• v.16 no.1
• /
• pp.103-112
• /
• 2014

Area average rainfall estimation is important to determine the exact amount of the available water resources and the essential input data for rainfall-runoff analysis. Like that, the necessary criterion for accurate area average rainfall estimate is the uniform spatial distribution of raingauge network. In this study, we suggest the spatial distribution evaluation methodology of raingauge network to estimate better area average rainfall and after the suggested method is applied to Han River and Geum River basin. The spatial distribution of rainfall network can be quantified by the nearest neighbor index. In order to evaluate the effects of the spatial distribution of rainfall network by each basin, area average rainfall was estimated by arithmetic mean method, the Thiessen's weighting method and estimation theory for 2013's rainfall event, and evaluated the involved errors by each cases. As a result, it can be found that the estimation error at the best basin of spatial distribution was lower than the worst basin of spatial distribution.

• Magazine of the Korean Society of Agricultural Engineers
• /
• v.45 no.2
• /
• pp.45-57
• /
• 2003

An artificial neural network model was developed to analyze and forecast Short-term river runoff from the Naju watershed, in Korea. Error back propagation neural networks (EBPN) of hourly rainfall and runoff data were found to have a high performance In forecasting runoff. The number of hidden nodes were optimized using total error and Bayesian information criterion. Model forecasts are very accurate (i.e., relative error is less than 3% and $R^2$is greater than 0.99) for calibration and verification data sets. Increasing the time horizon for application data sets, thus mating the model suitable for flood forecasting. decreases the accuracy of the model. The resulting optimal EBPN models for forecasting hourly runoff consists of ten rainfall and four runoff data(ANN0410 model) and ten rainfall and ten runoff data(ANN1010 model). Performances of the ANN0410 and ANN1010 models remain satisfactory up to 6 hours (i.e., $R^2$is greater than 0.92).

• Proceedings of the KIEE Conference
• /
• 2005.07d
• /
• pp.2507-2509
• /
• 2005

Because the rain gauges of tipping bucket type can easily use the digital signal, the rain gauges are widely used for the meteorological observation. In general, the resolution of rain gauges of tipping bucket type can be categorized by the 0.1mm, 0.5mm, and 1.0mm classes. But, the error of the tipping bucket rain gauges is made by the intensity of rainfalls and is expected to make the standard calibration method for error measurement. Thus, we developed the hardware of standard calibration facility for rain gauges by weighting measurement method and proposed the standard procedure by rainfall intensity in this study. Also, we calculated the error for the rainfall intensity and obtained useful result through the proposed calibration method.

• Journal of Korea Water Resources Association
• /
• v.54 no.12
• /
• pp.1329-1338
• /
• 2021

Rain radar provides high spatio-temporal radar rainfall that can be used as input data to short-term precipitation forecasting models. Korea Institute of Civil Engineering and Building Technology (KICT) has developed a flash flood forecasting system that is providing flash flood forecasting based on short-term rainfall forecasts estimated by the radar rainfall. Accuracy of the radar rainfall as well as the short-term rainfall forecasts, however, can deteriorate when radar polarimetric variables have error. In this study, we develope real-time program that can correct the error inherent in the radar polarimetric variables. First, effect according to the correction of the error was verified using 363 rainfall events on non real-time. The accuracy (1-NE) of the radar rainfall was approximately 70% and correlation coefficient was higher than 0.8 after correcting the error on non real-time. The accuracy (1-NE) using the real-time program was also approximately 70% after correcting the error.

• Journal of Environmental Science International
• /
• v.11 no.10
• /
• pp.1055-1060
• /
• 2002

A new method of automatic recording raingauge is developed to measure rainfall 1200mm full scale with high accuracy and resolution. The principle of new instrument is to detect a weight change of a buoyant weight according to a change in water level of raingauge measured by the use of a strain gauge load cell. This method has the advantage of increasing measurement accuracy, since no moving equipment is used. Laboratory test of the instrument was recorded 0.4% error of 190mm rainfall amount. The validity of new instrument was examined by comparing its measured values with values recorded by automatic weather station on June 24 to 25 2001 at Daegu Meteorological Station, when there is 148.3mm rainfall amount. In spite of much rainfall there is only 0.77mm difference of total rainfall amount. This instrument was accomplished high accuracy and resolution at field test in much rainy day.

• Atmosphere
• /
• v.22 no.3
• /
• pp.309-319
• /
• 2012

It is important to characterize and quantify the inherent error in the radar rainfall to make full use of the radar rainfall. This study verified the error structure of the reflectivity and corrected the range dependent error in the CAPPI using a VPR (vertical profile of reflectivity) model. The error of the CAPPI to display the reflectivity data becomes bigger for the range longer than 100 km. This range dependent error, however, is significantly improved by corrected the CAPPI data using the VPR model.

• Journal of Korea Water Resources Association
• /
• v.35 no.3
• /
• pp.307-319
• /
• 2002

This study used the WGR model to generate the rainfall input and the modified Clark method to estimate the runoff with the aim of investigating how the errors from the areal average rainfall propagates to runoff estimates. This was done for several cases of raingauge density and also by considering several storm directions. Summarizing the study results are as follows. (1) Rainfall and runoff errors decrease exponentially as the raingauge density increases. However, the error stagnates after a threshold density of raingauges. (2) Rainfall errors more affect to runoff estimates when the density of raingauges is relatively low. Generally, the ratio between estimation errors of rainfall and runoff volumes was found much less than one, which indicates that there is a smoothing effect of the basin. However, the ratio between estimation errors of rainfall to peak flow becomes greater than one to indicate the amplification of rainfall effect to peak flow. (3) For the study basin in this studs no significant effect of storm direction could be found. However, the runoff error becomes higher when the storm and drainage directions are identical. Also, the error was found higher for the peak flow than for the overall runoff hydrograph.

• Journal of Korea Water Resources Association
• /
• v.55 no.12
• /
• pp.1105-1114
• /
• 2022

A study on a method to overcome the limitations of the topographical and hydrological observation environment for estimating the QPE with high consistency with the ground rainfall by utilizing the spatiotemporal observation advantages of the rainfall radar for use in flood forecasting, and quantitative observations of localized rainfall due to these limiting conditions Uncertainty should be identified in terms of flood analysis. Against this background, in this study, 22 major heavy rain events in 2016 were analyzed for each of Mt. Biseul (BSL), Mt. Sobaek (SBS), Mt. Gari (GRS), Mt. Mohu (MHS), and Mt. Seodae (SDS) to determine the observation distance and altitude. The uncertainty of observation was quantified and an error map was derived. As a result of the analysis, it was found that, on average, the rainfall radar exceeded 10% up to 100 km and 30% over 150 km. Based on the average radar operating altitude angle, it was found that the error for the altitude was approximately 10% or less up to the second altitude angle, 20% at the third or higher altitude angle, and more than 50% at the fourth altitude angle or higher.

• Journal of Korea Water Resources Association
• /
• v.48 no.4
• /
• pp.257-268
• /
• 2015

The objective of this study is to assess the dual-polarization radar for flood forecasting. First, radar rainfall has temporal and spatial errors, so estimated radar rainfall was compared with ground observation rainfall to assess accuracy improvement, especially, considering the radar range of observation and increase of the rainfall intensity. The results of this study showed that the error for estimated dual-polarization radar rainfall was less than single-polarization radar rainfall. And in this study, dual-polarization radar rainfall for flood forecasting was assessed using MAP (Mean Areal Precipitation) and SURR (Sejong University Rainfall Runoff) model in Namkang dam watershed. The results of MAP are more accurate using dual-polarization radar. And the results of runoff using dual-polarization radar rainfall showed that peak flow error was reduced approximately 12~63%, runoff volumes error was reduced by approximately 30~42%, and also the root mean square error decreased compared to the result of runoff using single-polarization radar rainfall. The results revealed that dual-polarization radar will contribute to improving the accuracy of the flood forecasting.