Academia.eduAcademia.edu

Figure 4 – uploaded by Supasit Konyai

See full PDFdownloadDownload figure
Figure 4 Flow duration curve of Y.20 from 1988 to 2012 data

Figure 4 Flow duration curve of Y.20 from 1988 to 2012 data

Related Figures (12)

Figure 1 Chao Phraya Basin with Yom River as one of the main tributaries
Figure 1 Chao Phraya Basin with Yom River as one of the main tributaries
Figure 2 Hyetographs and hydrographs along the Yom River
Figure 2 Hyetographs and hydrographs along the Yom River
Among the two, only the deficit volume is preferable in drought mitigation analysis such as for storage design and operation. Therefore, we are focusing on how to analyze and predict streamflow deficit volume in this study. Table 1 Locations, elevations, and watershed areas of gauging stations
Among the two, only the deficit volume is preferable in drought mitigation analysis such as for storage design and operation. Therefore, we are focusing on how to analyze and predict streamflow deficit volume in this study. Table 1 Locations, elevations, and watershed areas of gauging stations
Figure 3 Example of deficit volumes of Y.20 with a fixed threshold level Streamflow drought happens when the discharge being lower than a specified value (threshold level) which can be fixed or varied with time [17], as in Figure 3 for a fixed value. Yevjevich [18] initiated the theory of run to be used with streamflow drought analysis. It is now known as threshold level method. The volume of flow during the time of streamflow being lower than a specified value is a deficit volume. Large deficit volume identifies severe drought event which needs a reasonable large storage for coping with the drought. Series of deficit volume data can be obtained from a series of daily discharge of streamflow at a gauging station. In this study we used the daily discharge data from the aforementioned eight stations (Figure 2). Threshold level is a subjective value, it might have been estimated from total water demand [17], from percentage of average streamflow [19], and from flow duration curve [20].
Figure 3 Example of deficit volumes of Y.20 with a fixed threshold level Streamflow drought happens when the discharge being lower than a specified value (threshold level) which can be fixed or varied with time [17], as in Figure 3 for a fixed value. Yevjevich [18] initiated the theory of run to be used with streamflow drought analysis. It is now known as threshold level method. The volume of flow during the time of streamflow being lower than a specified value is a deficit volume. Large deficit volume identifies severe drought event which needs a reasonable large storage for coping with the drought. Series of deficit volume data can be obtained from a series of daily discharge of streamflow at a gauging station. In this study we used the daily discharge data from the aforementioned eight stations (Figure 2). Threshold level is a subjective value, it might have been estimated from total water demand [17], from percentage of average streamflow [19], and from flow duration curve [20].
Figure 5 Sequential peak algorithm to overcome mutual drought events; (a) streamflow time series of Y.20, (b) showing fixed threshold level at 6.2 m?/s, and (c) deficit volumes and durations corresponding to (a) and (b).
Figure 5 Sequential peak algorithm to overcome mutual drought events; (a) streamflow time series of Y.20, (b) showing fixed threshold level at 6.2 m?/s, and (c) deficit volumes and durations corresponding to (a) and (b).
Table 2 Threshold levels Q70 from flow duration curves Table 3 Series of deficit volumes of Y.20 from water year 1998 to 2012
Table 2 Threshold levels Q70 from flow duration curves Table 3 Series of deficit volumes of Y.20 from water year 1998 to 2012
Table 4 Average numbers of drought events per year
Table 4 Average numbers of drought events per year
Figure 6 Comparison of 3 theoretical distributions with empirical Weibull for Y.20 Table 5 Comparison of R? for three theoretical models The series of deficit volumes, such as in Table 3, were used for calculating three theoretical probability distributions, i.e. lognormal, Weibull, and generalized Pareto. Then, they were compared to the empirical distribution of Weibull as in Figure 6 for Y.20. We applied the coefficient of determination R? to decide the best theoretical model as in Table 4. Figure 6 illustrates equally good matching of both Weibull and generalized Pareto with the empirical distribution. Lognormal distribution does not give a good result. The theoretical distribution of Weibull is preferable to generalized Pareto as shown in Table 5, therefore theoretical Weibull was chosen for frequency analyses in the study. The results of hydrological drought frequency analyses using theoretical Weibull distribution show, as in Figure 7, the distributions for almost all stations are arranged in an orderly manner, except Y.1C, with Y.17 with the most severe drought then Y.14, Y.6, Y.3A, Y.20, Y.4, and Y.5 the least. The result of Y.1C shows a strange distribution compared to others. It is almost smallest at 2-year return period, then climbs up to be more severe than Y.5 and Y.4 at 10-year, moreover at 20-year return period the drought severity of Y.1C is higher than Y.5, Y.4, and also Y.20 (Figure 7). The station Y.1C is at the city of Phrae with a weir across the
Figure 6 Comparison of 3 theoretical distributions with empirical Weibull for Y.20 Table 5 Comparison of R? for three theoretical models The series of deficit volumes, such as in Table 3, were used for calculating three theoretical probability distributions, i.e. lognormal, Weibull, and generalized Pareto. Then, they were compared to the empirical distribution of Weibull as in Figure 6 for Y.20. We applied the coefficient of determination R? to decide the best theoretical model as in Table 4. Figure 6 illustrates equally good matching of both Weibull and generalized Pareto with the empirical distribution. Lognormal distribution does not give a good result. The theoretical distribution of Weibull is preferable to generalized Pareto as shown in Table 5, therefore theoretical Weibull was chosen for frequency analyses in the study. The results of hydrological drought frequency analyses using theoretical Weibull distribution show, as in Figure 7, the distributions for almost all stations are arranged in an orderly manner, except Y.1C, with Y.17 with the most severe drought then Y.14, Y.6, Y.3A, Y.20, Y.4, and Y.5 the least. The result of Y.1C shows a strange distribution compared to others. It is almost smallest at 2-year return period, then climbs up to be more severe than Y.5 and Y.4 at 10-year, moreover at 20-year return period the drought severity of Y.1C is higher than Y.5, Y.4, and also Y.20 (Figure 7). The station Y.1C is at the city of Phrae with a weir across the
Figure 7 The deficit volumes at several return periods of all stations depend on risk acceptance and the suitability of the site, whether using surface or groundwater storages or both in conjunction [6],[30],[38]. The findings of this study can help planners and engineers quantify the amount of water to be reserved in the rainy season at each critical locations.
Figure 7 The deficit volumes at several return periods of all stations depend on risk acceptance and the suitability of the site, whether using surface or groundwater storages or both in conjunction [6],[30],[38]. The findings of this study can help planners and engineers quantify the amount of water to be reserved in the rainy season at each critical locations.
Figure 8 The deficit volumes at 20-year return period along the Yom, the numbers in brackets are deficit volumes (in Mm}+) at 20-year return period(modified from [36]).
Figure 8 The deficit volumes at 20-year return period along the Yom, the numbers in brackets are deficit volumes (in Mm}+) at 20-year return period(modified from [36]).
Related topics:
Environmental Science
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved