the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 26 Feb 2024
Future drought over Urmia Lake Basin under SSP scenarios: the relevance of snow melt
Abstract. Snow melt is one of the sources of freshwater supply in the late spring and summer in the mountainous regions of Iran, especially in the Urmia Lake Basin (ULB). In this study, past and future droughts of Urmia Lake Basin (ULB) have been studied by analyzing three types of droughts: (i) precipitation-deficit based characterized by the Standardized Precipitation Index (SPI), (ii) precipitation-evapotranspiration based droughts characterized by the Standardized Precipitation and Evapotranspiration Index (SPEI) and (iii) those droughts forced additionally by snow melt using the Snowmelt and Rain Index (SMRI). While reanalysis data ERA5-land describes the past climate, bias-corrected CMIP6 ensemble serves as the data for the future climate. Contrary to the SPI drought index, an increasing trend has been projected both in snowmelt-based (SMRI trend -0.068 units/year) and evapotranspiration-based (SPEI trend -0.079 units/year) drought indices, both for the period 1995–2014 and significant at the 5 % level. This indicates that summer droughts in the ULB will increase in the future, particularly because of increasing evapotranspiration and less snowmelt, while precipitation changes play a minor role.
Drought severity will generally increase from the near future (2021–2040) to the far future (2081–2100), particularly forced by snowmelt deficit under the SSP5-8.5 scenario for the far future. Under the present climate, the extent of drought-affected areas is similar for all three types of droughts. However, under future climate drought-affected areas forced by snowmelt deficit will increase from about 20 % in the near future (2021–2040) to 60 % in the far future (2081–2100), showing that snow melt plays a vital role in aggravating the drought over the Basin. A decrease in the Basin's drought trend in the 2080s and later can be seen both for SMRI and SPEI indices under SSP1-2.6, which may be due to the temperature effect on snowmelt and evapotranspiration from the reduction of greenhouse gas emissions in SSP1-2.6 scenario at the end of 21st century. Such a decrease in SMRI and SPEI drought indices can also be seen around the 2090s under the SSP2-4.5 scenario. Results also reveal that the mountainous areas of the Basin will experience much less drought compared to the lowlands (including the lake) and foothills.
- Preprint
(2322 KB) - Metadata XML
-
Supplement
(227 KB) - BibTeX
- EndNote
Status: open (extended)
-
RC1: 'Comment on hess-2024-48', Mauricio Zambrano-Bigiarini, 03 Apr 2024
reply
This manuscript investigates the evolution of historical and future droughts in The Urmia Lake Basin (ULB) in Iran.ERA5-Land is used as reference dataset for precipitation, air temperature and snow depth in the historical period 1995-2014, while seven CMIP6 climate models are used to provide the same data for the future period 2021-2100, considering three different shared socioeconomic pathways (SSP2.6, SSP4.5, SSP8.5). Different bias-correction methods are applied to the historical ERA5-Land dataset to compute three different drought indices (SPI, SPEI, SMRI) and analyse the corresponding severity, duration, and frequency of drought events.
This manuscript address an important topic for the drought community. Unfortunately, this manuscript is poorly written and difficult to read; the methodology has some technical errors, is not well described, and it lacks several references, which prevents to follow the theoretical basis behind the selected methods. In summary, the manuscript does not represent a substantial contribution to scientific progress within the scope of Hydrology and Earth System Sciences; the scientific approach and applied methods can not be considered valid; and the scientific results and conclusions are far from being presented in a clear, concise, and well-structured way. Therefore, this manuscript should be rejected in it current form.
Major comments:
-
MC1. Equations are not numbered sequentially with Arabic numerals in parentheses on the right-hand side, as requested by HESS (https://www.hydrology-and-earth-system-sciences.net/submission.html#math), which hampers the elaboration of specific comments related to equations.
-
MC2. The manuscript does not explain why “snow depth” (instantaneous grib-box average of the snow thickness on the ground, excluding snow on canopy) was selected instead of “snow depth water equivalent” (depth of snow from the snow-covered area of a grid box, i.e, the depth the water would have if the snow melted and was spread evenly over the whole grid box) from the ERA5-Land dataset to be used in this study. Also, how the snow depth was used in the computation of the SMRI is not clear at all (see MC6 below).
-
MC3. Incorrect formulation of drought indices. The equations for SPI (L231-233) and SPEI (L238-240) arenot correct, because they omit the transformation of the accumulated variable into a standard normal variable with zero mean and variance equal to one, before any fitting of a probability distribution.
-
MC4. Missing details of drought indices computation. The temporal scale used for computing SPI, SPEI and SMRI is not described in the manuscript, which is a very important missing technical detail, which prevents to properly understand the drought characteristics (duration, severity and frequency) derived from those indices. Also, the choice of the probability distribution and the observational window used to fit the selected distribution to data is not discussed at all in the manuscript, notwithstanding they have been recently identified by Laimighofer & Laaha (2022) as the most important sources of uncertainty in the computation of SPI and SPEI.
-) Laimighofer, J.; Laaha, G. (2022). How standard are standardized drought indices? Uncertainty components for the SPI & SPEI case. Journal of Hydrology, 613, 128385. doi:10.1016/j.jhydrol.2022.128385.
-
MC5. Computation of potential evapotranspiration. The equation used for computing the potential evapotranspiration (L244-246) is not named and not referenced, without any discussion about the suitability of the selected formula for snow-related climate change impact studies. This is a critical missing point, since Kingston et al. (2009) mentioned that “for certain regions and GCMs, choice of PET method can actually determine the direction of projections of future water resources”.
-) Kingston, D. G.; Todd, M. C.; Taylor, R. G.; Thompson, J. R.; Arnell, N. W. (2009). Uncertainty in the estimation of potential evapotranspiration under climate change. Geophysical Research Letters, 36(20). doi:10.1029/2009GL040267.
-
MC6. Computation of SMRI drought index. Given the fact that this article mentions “snow accumulation” in its title, all the snow-related aspects of the manuscript should be carefully explained, and that is not the case. In particular, the computation of SMRI described in the manuscript involves only snow accumulation and snowmelt, while the original formulation of Staudinger et al. (2014) includes also refreezing, with an upper bound for the liquid water content given by the water holding capacity. None of these aspects is discussed in detail in this manuscript. Also, does the snow depth collected from ERA5-Land play any role in the computation or verification of SMRI?. If not, why is snow depth collected from ERA5-Land?
-
MC7. Confounding description of drought duration and severity. L277-278 mention that “The duration (DD) of drought is the period in which the SPEI/SPI/SMRI value is continuously negative. It starts when the indices values are equal to -1 and ends when values become positive”. However, then in L278 it mentions that “It starts when the indices values are equal to -1 and ends when values become positive”. The first description implies that a drought event starts whenever a drought index value is lower than zero, while the second definition explicitly mention that an event starts when a drought index is equal or lower than -1. Not having a clear definition of a drought event prevents any confidence in the drought characteristics presented in the Results and Discussion sections.
-
MC8. Missing description of drought frequency. The computation of drought frequency is completely absent from the Methodology section, which prevents a proper understanding of all the frequency-related paragraphs in the Results section.
-
MC9. Bias correction. Section 2.3 (“Bias correction”) mentions that seven bias correction methods (Delta, EQM, PTF, QUANT, RQUANT, SSPLIN) were used “to enhance the accuracy of our climatological and hydrological models” (Were there any climatological or hydrological models developed by the authors in the manuscript?). However, L354 mention that only four models were corrected with additive and multiplicative delta methods, without explaining why the other six methods were not applied to those 4 models or clarifying if the seven bias correction methods were actually applied to all the climate models. Moreover, it seems that afterwards, only one method was used for the whole Results section, without explicitly mentioning which method was the one selected and why.
However, even more important than the previous confounding description of the bias correction methods used in this study, was the omission of two methods widely used to bias correct climate model data for impact studies in mountain environments: i) Quantile Delta Mapping (QDM; Cannon et al., 2015), which is a variant of quantile mapping that was designed to correct systematic distributional biases relative to historical observations and topreserve model-projected relative changes, avoiding the artificial deterioration of trends arising as a statistical artefact of standard quantile mapping; ii) Multivariate Bias Correction (MBC; Cannon, 2018), which combines QDM and random orthogonal rotations to match the multivariate distributions of climate model data and observed data, to allow an explicit consideration of the relation between air temperature and precipitation inmodelling the impacts of climate change in snow-dominated catchments. The omission of the previous two methods was not discussed at all in the manuscript.
-) Cannon, A. J.; Sobie, S. R.; Murdock, T. Q. (2015). Bias Correction of GCM Precipitation by Quantile Mapping: How Well Do Methods Preserve Changes in Quantiles and Extremes?, J. Climate, 28, 6938–6959, doi:10.1175/JCLI-D-14-00754.1.
-) Cannon, A. J. (2018). Multivariate quantile mapping bias correction: An N-dimensional probability density function transform for climate model simulations of multiple variables, Clim. Dynam., 50, 31–49,doi:10.1007/s00382-017-3580-6.
-
MC10. Uncertainty analysis. The Methodology section lacks a description of the uncertainty analysis used in this manuscript to evaluate the evolution of drought characteristics in the future period, which prevents a proper understanding of all the uncertainty-related paragraphs in the Results section.
-
MC11. Trends and statistical significance. Several parts of the manuscript describes trends (e.g., L374-391, L410, L460, L503) and statistical significance (e.g., L42, L103, L117, L122), without providing any description in the Methodology section about how these trends were computed and what type of statistical test was applied to compute the statistical significance.
-
MC12. Missing references. There are several references cited in the text but absent from the References section, which prevents a proper understanding of the theoretical background leading this manuscript. Some of them are:Separated & Nafung, 2021; Huo-Po et al., 2013; Hunting & Hauschka, 2020; Parka et al., 2016; Rhoades et al., 2022; Van Loon and Van Lanen, 2012; Ahmad Tahir et al., 2011; Habibi et al., 2021; Laimighofer and Laaha, 2022; Sam et al., 2022; Hersbach, 2020; Kaviani Malayeri et al., 2021; Bentsen et al., 2019; Yi-Chi et al., 2022; Yukimoto et al., 2019a, 2019b, 2019c; Guo et al., 2018a, 2018b, 2018c; McKee et al., 1993; Vicente-Serrano et al., 2010.
Minor comments for the Introduction section, only for illustrating the poor writing used in the manuscript:
-
L67. Incorrect citation: (Cowherd et a., 2023)
-
L68. Incorrect name of the hydrological variable “streamflow”: SSI (Standard stream flow index),
-
L68: Undefined acronym: SMRI
-
L77-79. Inconsistent tense: “If only the five most severe drought events of each catchment are considered, a shift towards more rain-to-snow-season shortages, warm snow-season droughts, and composite droughts was found”
-
L82: Past tense should be used for “show”
-
L83. Past tense should be used for “are”
-
L85-86. Not clear. Please correct: “a higher fraction of precipitation falls as snow is associated with higher mean streamflow compared to catchments with marginal or no snowfall”
-
L88: why “snow meltwater supply" is within quotation marks?
-
L89. Please add a reference for the statement “dropped by an average of 16%”
-
L92. Where PRECIS data “were employed”? In this study or somewhere else?
-
L94-95. Please add a reference for the statement “may result in an accumulation of 35–40% in Gilgit River flows”
-
L95. Remove the capital letter from “River”
-
L95, “indicate a doubling of summer runoff”. Where?
-
L97-98. To avoid likely misunderstandings, please remove the minus sign when indicating a decrease in any amount
-
L103. What is “oC/yr”?
-
L103. Urmia has not been introduced yet.
-
L104. What is “oC”?
-
L112. The acronym “ULB” has not been defined yet (since L1 onwards)
-
L115. Past tense should be used for “show”.
-
L116. Remove “usually” if the sentence refers to the work of Habibi et al. (2021).
-
L118. Remove capital letter from “Basin”
-
L118-120. “Above-average temperatures in the year result in earlier snowmelt and, therefore, an earlier streamflow peak”. Is the previous sentence a statement, a hypothesis or a finding of some previous work?
-
L120-123. If Saboor and Mir Mousavi (2014) found a decline in snowfall, their finding do not necessarily confirm the occurrence of streamflow droughts later in the year. Please re-phrase.
-
L125. In “mountain’s region”, did you mean “mountainous region”?
-
L131-132. Please provide better references.
-
L132-134. Please provide -at least- one reference to support the statement about “higher degree of modeling success”
-
L138. Laimighofer and Laaha (2022) is not in the References section.
-
L147-148. What is the reference for the “precipitation deficit droughts” and “precipitation-evaporation droughts”?
-
L151. Change “modeling frame” by “modeling framework”.
Some final minor comments in other sections:
-
Figure 2 has a very low quality and it is not legible.
-
L375. Table 4 is not present in the manuscript.
-
Figure 3 does not have the units of measurement for the numeric values of the trends.
-
Figures S1 to S4 in the Supplementary material have a very low quality, with illegible labels.
-
Figure 9. The 3D graphs do not depict a clear picture of the temporal evolution of drought characteristics.
Citation: https://doi.org/10.5194/hess-2024-48-RC1 -
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 208 | 40 | 9 | 257 | 16 | 13 | 14 |
- HTML: 208
- PDF: 40
- XML: 9
- Total: 257
- Supplement: 16
- BibTeX: 13
- EndNote: 14
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1