the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 07 Feb 2025
A New Method Towards More Rational Drought Propagation Characterization in Karst Regions
Abstract. Drought propagation time (TP) and lag time (TL) from meteorological to hydrological droughts are two important indicators for characterizing drought propagation, and thus, reasonable estimates of these indicators are conducive to improve the prediction of hydrological droughts. However, traditional quantification methods are mostly based on moving correlations of whole hydro-meteorological series, including the misinformative non-drought periods. Particularly for the karst regions, the flashy regime of rainfall-runoff relationship during non-drought periods might strongly bias the estimations of the drought propagation indicators. In this study, we propose a new method that use only the informative drought-period data to better quantify the above indicators. Moreover, we compared the effectiveness between the new and traditional methods in regions with different karstification degrees. The results revealed that: (1) hydrological drought events generally exhibited reduced frequencies, longer durations, and smaller deficit volumes than the corresponding meteorological drought events; (2) compared to conventional methods, the TP and TL obtained by using the new method better meet the practical requirements for monitoring more hydrological drought deficit volume, especially in karst regions; (3) the karstification degree is a key factor influencing the difference between the results obtained using the new and traditional methods, and the superiority of the new method is more pronounced with stronger karstification degrees of the region. The above results can improve our understanding of the drought propagation features in karst regions and can provide an evidence-base for precautional drought-mitigation actions.
- Preprint
(2207 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on hess-2024-308', Anonymous Referee #1, 25 Mar 2025
This manuscript proposes a new method based on drought-period data for characterizing drought propagation in karst regions. This method avoids the interference of data from non-drought periods, especially addressing the applicability issue in karst regions, and contrasts it with the limitations of traditional methods affected by non-drought period data. The research findings have practical significance for drought early warning and water resources management, particularly in karst regions where it is difficult to accurately simulate using hydrological models. However, due to the following issues, the manuscript currently does not meet the publication requirements of the HESS journal.
The specific comments are as follows:- The groundwater hydrological characteristics in karst regions have not been specifically quantified. Only a general description of "karstification degree" is used, which weakens the depth of the mechanism explanation. It is recommended to supplement karst development parameters (such as karst porosity, groundwater level dynamic data) or cite existing geological survey results, and clarify the classification criteria for karstification degree (such as based on the proportion of karst area or hydrogeological zoning).
- The data time range (1957–2008) is relatively early. Why not use more recent data? The impact of climate change in the past decade or so has been more significant. It is recommended to discuss the potential impact of climate change (such as the increased frequency of extreme precipitation) on drought propagation characteristics.
- In Figure 3, "meteorological drought inde" is used for all indices, while "index" is used in the text. Is this a spelling error or does it have a specific meaning?
- The new method has obvious limitations as it is only compared between karst and non-karst basins. It is recommended to increase case verifications in different climate zones (such as arid and humid zones), or cite existing studies to illustrate the transfer potential of the method. In the "Discussion" section, supplement the analysis of the applicability to other types of droughts (such as agricultural drought) to clarify the boundaries of the method.
- The description of some statistical tests is vague (for example, the Tp value of the Kruskal-Wallis H test is not fully labeled), and the impact of the sample size (such as the number of drought events) on the results has not been explained.
- The mechanism of the differences in TP and TL in karst regions has not been fully explored. It is only attributed to "weak regulation and storage capacity", lacking quantitative support from hydrogeological processes. It is recommended to analyze how the interaction between rapid runoff and slow seepage in karst areas prolongs the drought propagation time by combining groundwater flow models or field observation data, to enhance the scientific nature of the conclusion.
- Some figure citations are incomplete (for example, the descriptions of Figure 1 and Figure 5 have insufficient correspondence with the text), and the variable definitions in Equation (1) have not been fully explained. It is recommended to ensure that all chart titles and axis labels are complete, and explain key charts one by one in the text.
- The text descriptions and explanations of the figures are too simple, making it difficult to understand the specific meaning of the figures, especially for Figure 3 and Figure 4. For example, in Figure 3, there are three sub - figures (upper, middle, and lower), and the upper sub - figure is further divided into left and right parts, but the differences between them are not explained. In Figure 4, the bottom sub - figure lacks the label (c).
- The most crucial factors affecting the propagation from meteorological drought to hydrological drought in karst regions are the heterogeneity and sensitivity of the karst spatial structure. In particular, the drought propagation in different karst compartments has obvious differences. However, this part of the discussion is missing in the paper.
- It is recommended to use some specific terms. For example, "karsification degree" should preferably be replaced with the more common "karst development intensity" or refer to international karst classification standards (such as Ford & Williams, 2007).
Citation: https://doi.org/10.5194/hess-2024-308-RC1 -
RC2: 'Comment on hess-2024-308', Anonymous Referee #2, 12 Jun 2025
Referee report on: A New Method Towards More Rational Drought Propagation Characterization in Karst Regions
The present manuscript (MS) presents a new methodology that concentrates on drought periods instead of the entire time series to estimate drought propagation from meteorological to hydrological droughts. The basic idea is appealing and the basic topic is relevant for HESS.
However, the MS suffers from deficits in structure and language, it should be more precise and in many parts shortened. Also, existing knowledge should be considered in more detail in the introduction to arrive at a more convincing starting point. And most important, more prove is needed to justify why this new method, which is by far more complicated than the traditional one, is superior, because this is a single case study and differences in estimated drought deficit volumes appear to be relatively marginal. I detail my general concerns (A-D) first, followed by line-to line remarks. I also give ideas for improvement and hope those will be helpful for the authors.
A. Introduction and existing knowledge
In the introduction, a straightforward summary of existing knowledge is required to show the need for a new method to characterize drought propagation, particularly in karst regions. In its present form, the introduction largely summarizes the methods that have been applied in drought research but does not concisely summarize the results of these studies. This should be done in general terms, in karst regions, and in the study area. And this concerns three main topics:
- Drought propagation through the entire hydrological cycle also including streamflow and groundwater droughts. How are streamflow and groundwater droughts related? What are the main results by traditional methods to characterize drought propagation? One example here is that droughts propagate more quickly in catchments with higher annual rainfall. Therefore one would expect longer lag times when only drought periods are included in the analysis.
- Catchment characteristics and baseflow: High and stable baseflow in a catchment can be interpreted as a synonym for resilience to streamflow drought. There are many studies in literature that address the relation between baseflow and catchment characteristics, also including geology. This needs to be summarized.
- Karst hydrology: Since the study addresses the difference between karst and non-karst: The large bunch of knowledge on hydrological studies in karst regions must be summarized that are relevant for drought propagation: e.g. Flashiness of system responses, young and mature karst, etc. Here also the question should be answered how the degree of karstification can be assessed.
B. Rainfall and runoff data
The authors used the Thiessen Polygon method to calculate catchment rainfall. According to Tab. 1 they arrive at very accurate numbers ranging from 886.2 to 1517.3 mm per year. First, this high accuracy is not supported by this estimation method and second, Thiessen Polygons do not account for topography effects of rainfall that might be present. The DEM in Fig. 2 shows altitudes between 20 and 2870 m. I would assume topography effects, e.g. shadowing effects or more rainfall falling in higher altitudes, etc. This calls for more sophisticated methods of rainfall interpolation. Those are available and should be used, at least their results should be compared to the simple Thiessen method.
Moreover, the authors correct runoff data by a simple factor based on average mean values (equation I do not think that this is adequate here, because runoff signals still propagate from upstream catchments to the downstream gauge. And just these signals are decisive to drought definition (e.g. deficit volumes, pooling, termination, etc.). Hence the correction should be performed in a different way (e.g. by subtraction including routing) and the analysis should be re-done.
C. Make the paper more concise
Figures 2,3,4 are not needed and could be omitted to safe space. Figure 2 more or less explains textbook knowledge on drought definition. Figures 3 and 4 are too complicated and not helpful to understand the methods that are better explained in the text.
D. Is the new method really superior?
This is the main point of criticism. The new method needs a thorough, statistically sound prove that it is really superior compared to existing methods of drought propagation. The authors define their own efficiency criterion delta-E (equations 3,4,5) that is based on the correct estimation of the drought deficit volumes in their catchments. Hence it is difficult to judge, how powerful the new method really is, because traditional and well-proven efficiency measures are not applied and there is no comparison to studies outside the present region. On a first glance, the presented differences in deficit volumes seem rather small (fig. 10) and also the invalid modes of droughts, particularly mode 4, are larger for the new method compared to the traditional one. Here, better ways of method testing that are established in the scientific community are strongly recommended.
Line-by-Line
L29: what is meant by “wide ranges” (extent?)
L30: it is generally known that droughts can occur anywhere!
L32-33: term: what is a hydrological cycling system?
L37: what is accumulative time?
L39-40: How is this statement justified?
L41: What is meant by “relatively stable”?
L42: Language: Studies have been shown?
L44: what is traditional physical drought prediction?
L52: “In contrast” not “in contrasts”
L58: if you write "mostly" are there exceptions?
L61: sentence not complete, verb is missing.
L62: “examine” not “exam”
L77: the basin is characterized by basins?
L82: how do you define karstification degree?
L86ff (Table 1): how do the different sizes of your catchments affect your study, is there a scale issue?
L86ff (Table 1): Can you exclude anthropogenic effects on your streamflow during droughts (i.e. water abstraction, dams, artificial wastewater inflow, etc.
L97: Qi,r is used for two different things. Yet the entire approach needs to be changed, see above.
L105: Why do you choose 30 days?
L122: what ia the "accumulative period"? I would assume that this is the time lag between the the atmospheric and the hydrologic drought indices?
L130: not “sliding” but “moving” average.
L195-196: This statement is based on a very simple Thiessen polygon method!
L198: what is meant by “human activities”? Can you exclude them in your case?, see above.
L199: “that” not “than”
L201: “on the one hand” is missing
L220: this is not new (rainfall is generally more variable than streamflow) and results from other studies should be mentioned here.
L231: what do you mean by “regulation and storage capacity” this sounds artificial, use the correct terms.
L 245-248: This is not specific to China and not a new result: If they are defined by the same percentile, meteorological droughts occur more often, have higher deficit but are shorter. This is largely due to the balancing effect of (groundwater) reservoirs on streamflow. And if it comes to differences in catchments, the existing knowledge on baseflow characteristics should be discussed and included. Here also karst-non karst studies exist.
L258-260: What is known and what is new here?
L300-303: The general message here is: If only drought periods are selected to specify drought propagation from atmospheric to hydrological droughts, you will get longer lag times. But this can be expected, because drought propagation is quicker in catchments with higher annual rainfall. This obvious connection needs to be discussed.
L357: “modes” not “models”!
L405ff: The efficiency of this method should be proved first, before it is used to other types of drought. I propose to totally omit chapter 4.4
Citation: https://doi.org/10.5194/hess-2024-308-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 180 | 39 | 15 | 234 | 11 | 15 |
- HTML: 180
- PDF: 39
- XML: 15
- Total: 234
- BibTeX: 11
- EndNote: 15
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1