the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 24 Aug 2021
Hydrological Drought across Peninsular Malaysia: Implication of Drought Index
Abstract. Drought is considered a damaging natural disaster causing significant economic, social, and environmental impacts. The challenge of drought is to determine the characteristics of drought, its frequency, duration, and severity, which are critical for controlling the effects of drought and mitigation strategies. The objective of this study is to identify the drought characteristics and temporal assessment of drought using Streamflow Drought Index (SDI) and theory of runs (ToR). It also highlights the need and methods for selecting the most appropriate time scale for drought assessment, especially in tropical countries. Malaysia experiences tropical weather and monsoon seasons throughout the year with typically humid temperatures ranging from 20 °C to 30 °C. The different spatial patterns of SDI for three-, six-, nine-, and 12-month were adopted throughout Peninsular Malaysia, using 40 years of daily streamflow data from 42 gauging stations. The area under drought stress at different time scales during the study period is stable and accounts for about 24 % of the total area. The years 1997–1999, 2002 and 2016–2018 mark the most critical drought years, when more than 48 % of the total area of the basin was affected by hydrological drought. Spatial evaluation of drought characteristics shows that short-term droughts are common in most regions, with relatively high severity and frequency in the northeast and southeast of Peninsular Malaysia, where the maximum frequency reached 35.7 % and 42.8 %, respectively. The shortest scale (3-month) recorded more mild and moderate events. Since the most extensive time scale (12-month) includes more dry and wet periods, its high value may lead to misleading information for the early warning system. Using the results of this multi-scale SDI analysis, hydrologists, water managers, and policy makers can better understand the time scale selected for hydrological drought analysis. Short-term drought conditions show high interannual variability with the predominant pattern. It was shown that among the SDI time scales, the SDI for 3-month is the most suitable for effectively tracking hydrological droughts in tropical regions.
- Preprint
(1314 KB) - Metadata XML
- BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on nhess-2021-249', Anonymous Referee #1, 14 Sep 2021
Authors try to do their best for improving this manuscript. However, the authors were not able to positively address some of my previous comments.
The authors used only one index [Streamflow Drought Index (SDI)] for monitoring drought across Peninsular Malaysia. Also, the specific manuscript was presented with no innovative point of view regarding the advantages in drought and water scarcity monitoring, modeling, and forecasting.
One merit point of this manuscript could be the determination of a method to select the most appropriate time scale for drought assessment, especially in tropical countries, as the authors mentioned in the end of the introduction section. Specifically, the authors conclude that “Among the SDI time scales, SDI-3 is the most suitable for effectively tracking hydrological drought. For tropical regions, this is the scale that is most sensitive to changes in streamflow” However, I cannot find in this manuscript a robust scientific method / technique which could prove this result, on the contrary the authors mentioned in many different parts of this paper controversial aspects concerning this point.
For instance:
“SDI-12 is more suitable for water management applications” (PAGE 14, LINE 432)
“The spatial and temporal SDI analysis revealed that the SDI-3 and SDI-6 could be misleading in the regions that are normally dry for six months. The SDI-3 can be used to determine when the dry season begins and ends. However, a drought index for longer periods is essential. For example, a three-month drought may occur in the middle of a prolonged drought, but this would only be noticeable over longer periods such as 12 months” (PAGE 11, LINES 321-324)
Citation: https://doi.org/10.5194/nhess-2021-249-RC1 -
AC2: 'Reply on RC1', Hasrul Hazman Hasan, 21 Oct 2021
Comment on nhess-2021-249
Anonymous Referee #1
Authors try to do their best for improving this manuscript. However, the authors were not able to positively address some of my previous comments.
The authors used only one index [Streamflow Drought Index (SDI)] for monitoring drought across Peninsular Malaysia. Also, the specific manuscript was presented with no innovative point of view regarding the advantages in drought and water scarcity monitoring, modeling, and forecasting.
Response: We are grateful to the reviewer for their time and suggestions in helping to improve the manuscript.
One merit point of this manuscript could be the determination of a method to select the most appropriate time scale for drought assessment, especially in tropical countries, as the authors mentioned in the end of the introduction section. Specifically, the authors conclude that “Among the SDI time scales, SDI-3 is the most suitable for effectively tracking hydrological drought. For tropical regions, this is the scale that is most sensitive to changes in streamflow” However, I cannot find in this manuscript a robust scientific method / technique which could prove this result, on the contrary the authors mentioned in many different parts of this paper controversial aspects concerning this point.
For instance:
“SDI-12 is more suitable for water management applications” (PAGE 14, LINE 432)
“The spatial and temporal SDI analysis revealed that the SDI-3 and SDI-6 could be misleading in the regions that are normally dry for six months. The SDI-3 can be used to determine when the dry season begins and ends. However, a drought index for longer periods is essential. For example, a three-month drought may occur in the middle of a prolonged drought, but this would only be noticeable over longer periods such as 12 months” (PAGE 11, LINES 321-324)
Response:
For the next correction, the authors will test the performance of different probability distributions (assuming that each month fits different probability distributions) to calculate the streamflow drought index (SDI). It is well known that in hydrological studies based on frequency analysis, there are often uncertainties in sampling due to the limited data length and discontinuity of the observed streamflow series compared to meteorological data. The required procedure for estimating the parameters for the PDF implies that the calculation of the SDI from specific samples depends significantly on the characteristics of the sample and the size of the observed streamflow series.
This has enabled the authors of this paper to propose an accurate procedure to obtain a hydrological drought index useful for spatial and temporal comparisons over a wide range of flow regimes and flow characteristics.
Citation: https://doi.org/10.5194/nhess-2021-249-AC2 -
AC3: 'Reply on RC1', Hasrul Hazman Hasan, 22 Oct 2021
We are grateful to the reviewer for their time and suggestions in helping to improve the manuscript.
For the next correction, the authors will test the performance of different probability distributions (assuming that each month fits different probability distributions) to calculate the streamflow drought index (SDI). It is well known that in hydrological studies based on frequency analysis, there are often uncertainties in sampling due to the limited data length and discontinuity of the observed streamflow series compared to meteorological data. The required procedure for estimating the parameters for the PDF implies that the calculation of the SDI from specific samples depends significantly on the characteristics of the sample and the size of the observed streamflow series.
This has enabled the authors of this paper to propose an accurate procedure to obtain a hydrological drought index useful for spatial and temporal comparisons over a wide range of flow regimes and flow characteristics.
Citation: https://doi.org/10.5194/nhess-2021-249-AC3
-
AC2: 'Reply on RC1', Hasrul Hazman Hasan, 21 Oct 2021
-
RC2: 'Comment on nhess-2021-249', Anonymous Referee #2, 23 Sep 2021
The manuscript investigates drought conditions on the Peninsula Malaysia in between 1978 and 2018 based on monthly discharge data in 42 stations. The manuscript is well organized, even if it requires additional effort to improve the structure of the writing; it mainly presents a case study regional application, which might be improved. Indeed, the innovative contribution to the literature appears to be weak; the Authors should better motivate their work and describe how they are contributing in terms of new methods. Some specific comments follow; I hope that they will be useful for manuscript improvement.
The structure of the text needs some additional efforts from the Authors; specifically, in the Introduction Section there are many repetitions that can be avoided and the arguments presentation should be easier to follow.
L 170-172. I understand that the availability of discharge data is a positive, not so frequent condition. I was wondering if the use of rainfall data together with discharge data might provide a deeper analysis of drought condition with respect of using only the discharge data. I expected a comment on this from the Authors.
L 175, 279-282. The motivation to investigate different temporal resolution appears to be weak; I suppose that different temporal resolution could be of interest depending on the characteristic temporal scale of the storage/supply system. Since this is the main motivation of the proposed work, apart from analyzing the drought condition in Peninsula Malaysia, I believe that the Authors should better explain this issue and justify their choices.
L 476-477. The conclusion that the 3-months SDI better describes the variability of the process in time is almost expected, and does not depend on the results of the analysis.
Citation: https://doi.org/10.5194/nhess-2021-249-RC2 -
AC1: 'Reply on RC2', Hasrul Hazman Hasan, 21 Oct 2021
Comment on nhess-2021-249
Anonymous Referee #2
The manuscript investigates drought conditions on the Peninsula Malaysia in between 1978 and 2018 based on monthly discharge data in 42 stations. The manuscript is well organized, even if it requires additional effort to improve the structure of the writing; it mainly presents a case study regional application, which might be improved. Indeed, the innovative contribution to the literature appears to be weak; the Authors should better motivate their work and describe how they are contributing in terms of new methods. Some specific comments follow; I hope that they will be useful for manuscript improvement.
The structure of the text needs some additional efforts from the Authors; specifically, in the Introduction Section there are many repetitions that can be avoided and the arguments presentation should be easier to follow.
L 170-172. I understand that the availability of discharge data is a positive, not so frequent condition. I was wondering if the use of rainfall data together with discharge data might provide a deeper analysis of drought condition with respect of using only the discharge data. I expected a comment on this from the Authors.
L 175, 279-282. The motivation to investigate different temporal resolution appears to be weak; I suppose that different temporal resolution could be of interest depending on the characteristic temporal scale of the storage/supply system. Since this is the main motivation of the proposed work, apart from analyzing the drought condition in Peninsula Malaysia, I believe that the Authors should better explain this issue and justify their choices.
L 476-477. The conclusion that the 3-months SDI better describes the variability of the process in time is almost expected, and does not depend on the results of the analysis.
Response: We thank the reviewers for their time and suggestions, which helped to improve the manuscript.
The origin of hydrological droughts is usually a climatic drought, but the quantification of hydrological droughts as an independent phenomenon has also received much attention in the scientific community. This is because there is usually no direct spatial or temporal relationship between climate and hydrological droughts (Vidal et al. 2010). Moreover, the analysis of hydrological droughts allows for a direct quantification of deficits in usable water sources.
However, a reduction in flows during high flow periods can have negative impacts on natural systems adapted to a specific flow regime. For example, the unusually low flows during high flow duration can reduce the storage of downstream reservoirs and affect the availability of water resources for certain uses a few months later. For these reasons, in addition to using low-flow analysis through run theory (see reviews in Smakhtin 2001; Tallaksen and Van Lanen 2004), it would be beneficial to develop a standardised hydrological drought indicator that would allow comparisons of drought severity over time and space, including in catchments with different characteristics in terms of regime, flow variability and magnitude. Such an indicator could be calculated using the same theoretical approach as the climatic drought indices.
In contrast to climatic drought, the quantification of hydrological drought is usually not based on indices but on the theory of runs. These indices have the same theoretical background, as they derive the hydrological drought index by converting monthly streamflow into z-scores. The problem with this approach is that the selection of the most appropriate probability distribution for calculating the index and the impact of the selection on the final series have not been thoroughly tested.
River flow tends to have greater spatial variability than the climate variables used to derive drought indicators. This is due to the influence of a number of factors, including topography, lithology, vegetation and human management. It is also a consequence of the spatial aggregation of runoff, which alters the statistical properties of the series downstream (Mudelsee 2007). Therefore, there is a high degree of spatial variability in the probability distributions that best fit the monthly streamflow data, making it difficult to select the most appropriate distribution for calculating a drought index for a large area.
To further corrected paper will investigate the statistical properties of observed samples of hydrological variables, the relationship between the sampling uncertainty resulting from limited observed flow series and the corresponding sample size based on the bootstrap method. By reconstructing a large number of bootstrap samples from the original flow series, the effects of different data lengths on the estimation of the parameters of PDF and SDI for Peninsular Malaysia were analysed.
Citation: https://doi.org/10.5194/nhess-2021-249-AC1 -
AC4: 'Reply on RC2', Hasrul Hazman Hasan, 22 Oct 2021
We thank the reviewers for their time and suggestions, which helped to improve the manuscript.
The origin of hydrological droughts is usually a climatic drought, but the quantification of hydrological droughts as an independent phenomenon has also received much attention in the scientific community. This is because there is usually no direct spatial or temporal relationship between climate and hydrological droughts (Vidal et al. 2010). Moreover, the analysis of hydrological droughts allows for a direct quantification of deficits in usable water sources.
However, a reduction in flows during high flow periods can have negative impacts on natural systems adapted to a specific flow regime. For example, the unusually low flows during high flow duration can reduce the storage of downstream reservoirs and affect the availability of water resources for certain uses a few months later. For these reasons, in addition to using low-flow analysis through run theory (see reviews in Smakhtin 2001; Tallaksen and Van Lanen 2004), it would be beneficial to develop a standardised hydrological drought indicator that would allow comparisons of drought severity over time and space, including in catchments with different characteristics in terms of regime, flow variability and magnitude. Such an indicator could be calculated using the same theoretical approach as the climatic drought indices.
In contrast to climatic drought, the quantification of hydrological drought is usually not based on indices but on the theory of runs. These indices have the same theoretical background, as they derive the hydrological drought index by converting monthly streamflow into z-scores. The problem with this approach is that the selection of the most appropriate probability distribution for calculating the index and the impact of the selection on the final series have not been thoroughly tested.
River flow tends to have greater spatial variability than the climate variables used to derive drought indicators. This is due to the influence of a number of factors, including topography, lithology, vegetation and human management. It is also a consequence of the spatial aggregation of runoff, which alters the statistical properties of the series downstream (Mudelsee 2007). Therefore, there is a high degree of spatial variability in the probability distributions that best fit the monthly streamflow data, making it difficult to select the most appropriate distribution for calculating a drought index for a large area.
To further corrected paper will investigate the statistical properties of observed samples of hydrological variables, the relationship between the sampling uncertainty resulting from limited observed flow series and the corresponding sample size based on the bootstrap method. By reconstructing a large number of bootstrap samples from the original flow series, the effects of different data lengths on the estimation of the parameters of PDF and SDI for Peninsular Malaysia were analysed.
References
Mudelsee, M. (2007). “Long memory of rivers from spatial aggregation.” Water Resour. Res., 43, W01202.
Smakhtin, V. U. (2001). “Low flow hydrology: A review.” J. Hydrol. (Amsterdam), 240(3–4), 147–186.
Tallaksen, L. M., and van Lanen, H. A. J. (2004). Hydrological drought— Processes and estimation methods for streamflow and groundwater, Elsevier, Amsterdam, Netherlands.
Vidal, J. P., et al. (2010). “Multilevel and multiscale drought reanalysis over France with the Safran-Isba-Modcou hydrometeorological suite.” Hydrol. Earth Syst. Sci., 14(3), 459–478.
Citation: https://doi.org/10.5194/nhess-2021-249-AC4
-
AC1: 'Reply on RC2', Hasrul Hazman Hasan, 21 Oct 2021
Status: closed
-
RC1: 'Comment on nhess-2021-249', Anonymous Referee #1, 14 Sep 2021
Authors try to do their best for improving this manuscript. However, the authors were not able to positively address some of my previous comments.
The authors used only one index [Streamflow Drought Index (SDI)] for monitoring drought across Peninsular Malaysia. Also, the specific manuscript was presented with no innovative point of view regarding the advantages in drought and water scarcity monitoring, modeling, and forecasting.
One merit point of this manuscript could be the determination of a method to select the most appropriate time scale for drought assessment, especially in tropical countries, as the authors mentioned in the end of the introduction section. Specifically, the authors conclude that “Among the SDI time scales, SDI-3 is the most suitable for effectively tracking hydrological drought. For tropical regions, this is the scale that is most sensitive to changes in streamflow” However, I cannot find in this manuscript a robust scientific method / technique which could prove this result, on the contrary the authors mentioned in many different parts of this paper controversial aspects concerning this point.
For instance:
“SDI-12 is more suitable for water management applications” (PAGE 14, LINE 432)
“The spatial and temporal SDI analysis revealed that the SDI-3 and SDI-6 could be misleading in the regions that are normally dry for six months. The SDI-3 can be used to determine when the dry season begins and ends. However, a drought index for longer periods is essential. For example, a three-month drought may occur in the middle of a prolonged drought, but this would only be noticeable over longer periods such as 12 months” (PAGE 11, LINES 321-324)
Citation: https://doi.org/10.5194/nhess-2021-249-RC1 -
AC2: 'Reply on RC1', Hasrul Hazman Hasan, 21 Oct 2021
Comment on nhess-2021-249
Anonymous Referee #1
Authors try to do their best for improving this manuscript. However, the authors were not able to positively address some of my previous comments.
The authors used only one index [Streamflow Drought Index (SDI)] for monitoring drought across Peninsular Malaysia. Also, the specific manuscript was presented with no innovative point of view regarding the advantages in drought and water scarcity monitoring, modeling, and forecasting.
Response: We are grateful to the reviewer for their time and suggestions in helping to improve the manuscript.
One merit point of this manuscript could be the determination of a method to select the most appropriate time scale for drought assessment, especially in tropical countries, as the authors mentioned in the end of the introduction section. Specifically, the authors conclude that “Among the SDI time scales, SDI-3 is the most suitable for effectively tracking hydrological drought. For tropical regions, this is the scale that is most sensitive to changes in streamflow” However, I cannot find in this manuscript a robust scientific method / technique which could prove this result, on the contrary the authors mentioned in many different parts of this paper controversial aspects concerning this point.
For instance:
“SDI-12 is more suitable for water management applications” (PAGE 14, LINE 432)
“The spatial and temporal SDI analysis revealed that the SDI-3 and SDI-6 could be misleading in the regions that are normally dry for six months. The SDI-3 can be used to determine when the dry season begins and ends. However, a drought index for longer periods is essential. For example, a three-month drought may occur in the middle of a prolonged drought, but this would only be noticeable over longer periods such as 12 months” (PAGE 11, LINES 321-324)
Response:
For the next correction, the authors will test the performance of different probability distributions (assuming that each month fits different probability distributions) to calculate the streamflow drought index (SDI). It is well known that in hydrological studies based on frequency analysis, there are often uncertainties in sampling due to the limited data length and discontinuity of the observed streamflow series compared to meteorological data. The required procedure for estimating the parameters for the PDF implies that the calculation of the SDI from specific samples depends significantly on the characteristics of the sample and the size of the observed streamflow series.
This has enabled the authors of this paper to propose an accurate procedure to obtain a hydrological drought index useful for spatial and temporal comparisons over a wide range of flow regimes and flow characteristics.
Citation: https://doi.org/10.5194/nhess-2021-249-AC2 -
AC3: 'Reply on RC1', Hasrul Hazman Hasan, 22 Oct 2021
We are grateful to the reviewer for their time and suggestions in helping to improve the manuscript.
For the next correction, the authors will test the performance of different probability distributions (assuming that each month fits different probability distributions) to calculate the streamflow drought index (SDI). It is well known that in hydrological studies based on frequency analysis, there are often uncertainties in sampling due to the limited data length and discontinuity of the observed streamflow series compared to meteorological data. The required procedure for estimating the parameters for the PDF implies that the calculation of the SDI from specific samples depends significantly on the characteristics of the sample and the size of the observed streamflow series.
This has enabled the authors of this paper to propose an accurate procedure to obtain a hydrological drought index useful for spatial and temporal comparisons over a wide range of flow regimes and flow characteristics.
Citation: https://doi.org/10.5194/nhess-2021-249-AC3
-
AC2: 'Reply on RC1', Hasrul Hazman Hasan, 21 Oct 2021
-
RC2: 'Comment on nhess-2021-249', Anonymous Referee #2, 23 Sep 2021
The manuscript investigates drought conditions on the Peninsula Malaysia in between 1978 and 2018 based on monthly discharge data in 42 stations. The manuscript is well organized, even if it requires additional effort to improve the structure of the writing; it mainly presents a case study regional application, which might be improved. Indeed, the innovative contribution to the literature appears to be weak; the Authors should better motivate their work and describe how they are contributing in terms of new methods. Some specific comments follow; I hope that they will be useful for manuscript improvement.
The structure of the text needs some additional efforts from the Authors; specifically, in the Introduction Section there are many repetitions that can be avoided and the arguments presentation should be easier to follow.
L 170-172. I understand that the availability of discharge data is a positive, not so frequent condition. I was wondering if the use of rainfall data together with discharge data might provide a deeper analysis of drought condition with respect of using only the discharge data. I expected a comment on this from the Authors.
L 175, 279-282. The motivation to investigate different temporal resolution appears to be weak; I suppose that different temporal resolution could be of interest depending on the characteristic temporal scale of the storage/supply system. Since this is the main motivation of the proposed work, apart from analyzing the drought condition in Peninsula Malaysia, I believe that the Authors should better explain this issue and justify their choices.
L 476-477. The conclusion that the 3-months SDI better describes the variability of the process in time is almost expected, and does not depend on the results of the analysis.
Citation: https://doi.org/10.5194/nhess-2021-249-RC2 -
AC1: 'Reply on RC2', Hasrul Hazman Hasan, 21 Oct 2021
Comment on nhess-2021-249
Anonymous Referee #2
The manuscript investigates drought conditions on the Peninsula Malaysia in between 1978 and 2018 based on monthly discharge data in 42 stations. The manuscript is well organized, even if it requires additional effort to improve the structure of the writing; it mainly presents a case study regional application, which might be improved. Indeed, the innovative contribution to the literature appears to be weak; the Authors should better motivate their work and describe how they are contributing in terms of new methods. Some specific comments follow; I hope that they will be useful for manuscript improvement.
The structure of the text needs some additional efforts from the Authors; specifically, in the Introduction Section there are many repetitions that can be avoided and the arguments presentation should be easier to follow.
L 170-172. I understand that the availability of discharge data is a positive, not so frequent condition. I was wondering if the use of rainfall data together with discharge data might provide a deeper analysis of drought condition with respect of using only the discharge data. I expected a comment on this from the Authors.
L 175, 279-282. The motivation to investigate different temporal resolution appears to be weak; I suppose that different temporal resolution could be of interest depending on the characteristic temporal scale of the storage/supply system. Since this is the main motivation of the proposed work, apart from analyzing the drought condition in Peninsula Malaysia, I believe that the Authors should better explain this issue and justify their choices.
L 476-477. The conclusion that the 3-months SDI better describes the variability of the process in time is almost expected, and does not depend on the results of the analysis.
Response: We thank the reviewers for their time and suggestions, which helped to improve the manuscript.
The origin of hydrological droughts is usually a climatic drought, but the quantification of hydrological droughts as an independent phenomenon has also received much attention in the scientific community. This is because there is usually no direct spatial or temporal relationship between climate and hydrological droughts (Vidal et al. 2010). Moreover, the analysis of hydrological droughts allows for a direct quantification of deficits in usable water sources.
However, a reduction in flows during high flow periods can have negative impacts on natural systems adapted to a specific flow regime. For example, the unusually low flows during high flow duration can reduce the storage of downstream reservoirs and affect the availability of water resources for certain uses a few months later. For these reasons, in addition to using low-flow analysis through run theory (see reviews in Smakhtin 2001; Tallaksen and Van Lanen 2004), it would be beneficial to develop a standardised hydrological drought indicator that would allow comparisons of drought severity over time and space, including in catchments with different characteristics in terms of regime, flow variability and magnitude. Such an indicator could be calculated using the same theoretical approach as the climatic drought indices.
In contrast to climatic drought, the quantification of hydrological drought is usually not based on indices but on the theory of runs. These indices have the same theoretical background, as they derive the hydrological drought index by converting monthly streamflow into z-scores. The problem with this approach is that the selection of the most appropriate probability distribution for calculating the index and the impact of the selection on the final series have not been thoroughly tested.
River flow tends to have greater spatial variability than the climate variables used to derive drought indicators. This is due to the influence of a number of factors, including topography, lithology, vegetation and human management. It is also a consequence of the spatial aggregation of runoff, which alters the statistical properties of the series downstream (Mudelsee 2007). Therefore, there is a high degree of spatial variability in the probability distributions that best fit the monthly streamflow data, making it difficult to select the most appropriate distribution for calculating a drought index for a large area.
To further corrected paper will investigate the statistical properties of observed samples of hydrological variables, the relationship between the sampling uncertainty resulting from limited observed flow series and the corresponding sample size based on the bootstrap method. By reconstructing a large number of bootstrap samples from the original flow series, the effects of different data lengths on the estimation of the parameters of PDF and SDI for Peninsular Malaysia were analysed.
Citation: https://doi.org/10.5194/nhess-2021-249-AC1 -
AC4: 'Reply on RC2', Hasrul Hazman Hasan, 22 Oct 2021
We thank the reviewers for their time and suggestions, which helped to improve the manuscript.
The origin of hydrological droughts is usually a climatic drought, but the quantification of hydrological droughts as an independent phenomenon has also received much attention in the scientific community. This is because there is usually no direct spatial or temporal relationship between climate and hydrological droughts (Vidal et al. 2010). Moreover, the analysis of hydrological droughts allows for a direct quantification of deficits in usable water sources.
However, a reduction in flows during high flow periods can have negative impacts on natural systems adapted to a specific flow regime. For example, the unusually low flows during high flow duration can reduce the storage of downstream reservoirs and affect the availability of water resources for certain uses a few months later. For these reasons, in addition to using low-flow analysis through run theory (see reviews in Smakhtin 2001; Tallaksen and Van Lanen 2004), it would be beneficial to develop a standardised hydrological drought indicator that would allow comparisons of drought severity over time and space, including in catchments with different characteristics in terms of regime, flow variability and magnitude. Such an indicator could be calculated using the same theoretical approach as the climatic drought indices.
In contrast to climatic drought, the quantification of hydrological drought is usually not based on indices but on the theory of runs. These indices have the same theoretical background, as they derive the hydrological drought index by converting monthly streamflow into z-scores. The problem with this approach is that the selection of the most appropriate probability distribution for calculating the index and the impact of the selection on the final series have not been thoroughly tested.
River flow tends to have greater spatial variability than the climate variables used to derive drought indicators. This is due to the influence of a number of factors, including topography, lithology, vegetation and human management. It is also a consequence of the spatial aggregation of runoff, which alters the statistical properties of the series downstream (Mudelsee 2007). Therefore, there is a high degree of spatial variability in the probability distributions that best fit the monthly streamflow data, making it difficult to select the most appropriate distribution for calculating a drought index for a large area.
To further corrected paper will investigate the statistical properties of observed samples of hydrological variables, the relationship between the sampling uncertainty resulting from limited observed flow series and the corresponding sample size based on the bootstrap method. By reconstructing a large number of bootstrap samples from the original flow series, the effects of different data lengths on the estimation of the parameters of PDF and SDI for Peninsular Malaysia were analysed.
References
Mudelsee, M. (2007). “Long memory of rivers from spatial aggregation.” Water Resour. Res., 43, W01202.
Smakhtin, V. U. (2001). “Low flow hydrology: A review.” J. Hydrol. (Amsterdam), 240(3–4), 147–186.
Tallaksen, L. M., and van Lanen, H. A. J. (2004). Hydrological drought— Processes and estimation methods for streamflow and groundwater, Elsevier, Amsterdam, Netherlands.
Vidal, J. P., et al. (2010). “Multilevel and multiscale drought reanalysis over France with the Safran-Isba-Modcou hydrometeorological suite.” Hydrol. Earth Syst. Sci., 14(3), 459–478.
Citation: https://doi.org/10.5194/nhess-2021-249-AC4
-
AC1: 'Reply on RC2', Hasrul Hazman Hasan, 21 Oct 2021
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 925 | 387 | 58 | 1,370 | 60 | 52 |
- HTML: 925
- PDF: 387
- XML: 58
- Total: 1,370
- BibTeX: 60
- EndNote: 52
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
