• Journal of Korean Society of Water and Wastewater
• /
• v.22 no.1
• /
• pp.15-22
• /
• 2008

This study collected the latest 30-year (1976~2005) continuous rainfall data hourly recorded at 61 meterological observatories in Korea, and the continuous rainfall data was divided into individual rainfall events. In addition, distribution charts of average rainfall event-depth were created to facilitate the application to the overflow risk-based design of detention storage basin. This study shows that 4 hour is appropriate for SST (storm separation time) to separate individual rainfall events from the continuous rainfall data, and the one-parameter exponential distribution is suitable for the frequency distribution of rainfall event depths for the domestic rainfall data. The analysis of the domestic rainfall data using SST of 4 hour showed that the individual rainfall event was 1380 to 2031 times, the average rainfall event-depth was 19.1 to 32.4mm, and ranged between 0.877 and 0.926. Distribution charts of average rainfall event-depth were created for 4hour and 6 hour of SST, respectively. The inland Gyeongsangbuk-do, Western coastal area and inland of Jeollabuk-do had relatively lower average rainfall event-depth, whereas Southern coastal area, such as Namhae, Yeosu, and Jeju-do had relatively higher average rainfall event-depth.

• Journal of Korean Society for Atmospheric Environment
• /
• v.34 no.5
• /
• pp.735-747
• /
• 2018

Rainfall depth is an important meteorological information. Generally, high spatial resolution rainfall data such as road-level rainfall data are more beneficial. However, it is expensive to set up sufficient Automatic Weather Systems to get the road-level rainfall data. In this paper, we propose to use deep learning to recognize rainfall depth from road surveillance videos. To achieve this goal, we collect a new video dataset and propose a procedure to calculate refined rainfall depth from the original meteorological data. We also propose to utilize the differential frame as well as the optical flow image for better recognition of rainfall depth. Under the Temporal Segment Networks framework, the experimental results show that the combination of the video frame and the differential frame is a superior solution for the rainfall depth recognition. The final model is able to achieve high performance in the single-location low sensitivity classification task and reasonable accuracy in the higher sensitivity classification task for both the single-location and the multi-location case.

• Journal of Korean Society of Water and Wastewater
• /
• v.22 no.1
• /
• pp.65-71
• /
• 2008

This study collected the latest 30-year (1976~2005) continuous rainfall data hourly recorded at 61 meterological observatories in Korea, and the continuous rainfall data was divided into individual rainfall events. In addition, distribution charts of average rainfall event-depth were created to facilitate the application to the overflow risk-based design of detention storage basin. This study shows that 4 hour is appropriate for SST (storm separation time) to separate individual rainfall events from the continuous rainfall data, and the one-parameter exponential distribution is suitable for the frequency distribution of rainfall event depths for the domestic rainfall data. The analysis of the domestic rainfall data using SST of 4 hour showed that the individual rainfall event was 1380 to 2031 times, the average rainfall event-depth was 19.1 to 32.4mm, and ranged between 0.877 and 0.926. Distribution charts of average rainfall event-depth were created for 4hour and 6 hour of SST, respectively. The inland Gyeongsangbuk-do, Western coastal area and inland of Jeollabuk-do had relatively lower average rainfall event-depth, whereas Southern coastal area, such as Namhae, Yeosu, and Jeju-do had relatively higher average rainfall event-depth.

• Journal of The Korean Society of Agricultural Engineers
• /
• v.59 no.6
• /
• pp.73-80
• /
• 2017

The objective of this study is to investigate whether the daily rainfall depth derived from daily data represents the event rainfall depth derived from hourly data. For analysis, the 85th, 90th, and 95th percentile daily rainfall depths were first computed using daily rainfall data (1986~2015) collected at 63 weather stations. In addition, the storm event was separated by the interevent time definition (IETD) of 6, 12, 18, and 24 hr using hourly rainfall data. Based on the separated storm events, the 85th, 90th, and 95th percentile event rainfall depths were calculated and compared with the using hourly rainfall data with the 85th, 90th, and 95th percentile daily rainfall depths. The event rainfall depths computed using the IETD were greater than the daily rainfall depths. The difference between the event rainfall depth and the daily rainfall depth affects the design and size of the facility for controlling the stormwater. Therefore, the designer and policy decision-maker in designing the stormwater best management practices need to take into account the difference generated by the difference of the used rainfall data and the selected IETD.

• Water for future
• /
• v.20 no.1
• /
• pp.55-61
• /
• 1987

This Study is to estimate depth-length ratios by the frequency analysis of simultaneous rainfall data on fixed points in the Han -River basin. The depth-length ratios are based on examination of spatial and temporal correlation structures of the Han-River basin's rainfall. This Study reveals that the depth-length ratios are decreasing as the distance of the basin increases and that they are increasing as the duration of rainfall increases. This study also shows that the ratios tend to decrease as the return period increase. The results present general equations and characteristic equations of depth-length ratios, figures and characteristic parameter tables which explain the temporal and spatial structures of rainfall in the Han-River basin.

• The Journal of Korean Institute of Next Generation Computing
• /
• v.14 no.6
• /
• pp.87-97
• /
• 2018

Rainfall depth is an important meteorological information. Generally, high spatial resolution rainfall data such as road-level rainfall data are more beneficial. However, it is expensive to set up sufficient Automatic Weather Systems to get the road-level rainfall data. In this paper, we proposed to use deep learning to recognize rainfall depth from road surveillance videos. To achieve this goal, we collected two new video datasets, and proposed a new deep learning architecture named Temporal and Spatial Segment Networks (TSSN) for rainfall depth recognition. Under TSSN, the experimental results show that the combination of the video frame and the differential frame is a superior solution for the rainfall depth recognition. Also, the proposed TSSN architecture outperforms other architectures implemented in this paper.

• Journal of Korean Society of Water and Wastewater
• /
• v.30 no.1
• /
• pp.51-58
• /
• 2016

On account of the increase in water demand and climate change, droughts are in great concern for water resources planning and management. In this study, rainfall characteristics with stationary and non-stationary perspectives were analyzed using Weibull distribution model with 40-year records of annual minimum rainfall depth collected in major cities of Korea. As a result, the non-stationary minimum probable rainfall was expected to decrease, compared with the stationary probable rainfall. The reliability of ${\xi}_1$, a variable reflecting the decrease of the minimum rainfall depth due to climate change, in Wonju, Daegu, and Busan was over 90%, indicating the probability that the minimal rainfall depths in those city decrease is high.

• Proceedings of the Korean Society of Agricultural Engineers Conference
• /
• 2001.10a
• /
• pp.93-98
• /
• 2001

This study was performed to analyse the rainfall and the rainfall-runoff characteristics of a rural watershed. The Sangwha basin($105.9km^{2}$) in the Geum river system was selected for this study. The arithmetic mean method, the Thiessen's weighing method, and the isohyetal method were used to analyse areal rainfall distribution and the Huff's quartile method was used to analyse temporal rainfall distribution. In addition, daily runoff analyses were peformed using the DAWAST and tank model. In the model calibration, the data from June through November, 1999 were used. In the model calibration, the observed runoff depth was 513.7mm and runoff rate was 45.2%, and the DAWAST model simulated runoff depth was 608.6mm and runoff rate was 53.5%, and the tank model runoff depth was 596.5mm and runoff rate was 52.5%, respectively. In the model test, the data from June through November, 2000 were used. In the model test, the observed runoff depth was 1032.3mm and runoff rate was 72.5%, and the DAWAST model simulated runoff depth was 871.6mm and runoff rate was 61.3%, and the tank model runoff depth was 825.4mm and runoff rate was 58%, respectively. The DAWAST and tank model's $R^{2}$ and RMSE were 0.85, 3.61mm, and 0.85, 2.77mm in 1999, and 0.83, 5.73mm, and 0.87, 5.39mm in 2000, respectively. Both models predicted low flow runoff better than flood runoff.

• Journal of the Korean Society of Hazard Mitigation
• /
• v.1 no.1 s.1
• /
• pp.103-115
• /
• 2001

The current procedure to design hydraulic structures in a small basin area is to estimate the probable rainfall depth using rainfall intensity formula. The estimation of probable rainfall depth has many uncertainties inherent with it. However, it has been inevitable to simplify the nonlinearity if the rainfall in practice. This study attend to address a method which can model the nonlinearity in order to derive better rainfall intensity formula for the estimation of probable rainfall depth. The results show that genetic algorithm is more reliable and accurate than trial-and-error method or nonlinear programming technique(Powell's method) in the derivation of the rainfall intensity formula.

• Journal of the Korean Geosynthetics Society
• /
• v.11 no.4
• /
• pp.63-69
• /
• 2012

Climate change, according to the country to increase locality of slope collapse of heavy disaster, such as increasing the likelihood and prior in order to prevent these disasters, "Slope construction design standards (Ministry of Land, 2011)," is prescribed in the relevant guidelines. In recent years, guidelines Slope Stability Analysis of the existing methods when the rainy season infiltration of rainfall, taking into account have been revised to perform more realistic. In this study, according these trends to the analysis of saturation depth by rainfall intensity and soil conditions. Results as a whole, the larger the saturated hydraulic conductivity and depth of rainfall intensity also showed a tendency to rise in proportion but MH, CL did not occur in the saturation region. Analysis of antecedent rainfall case also reflects an overall increase of depth in the saturated, rainfall in many cases is less than the growth rate was higher in the saturation region.