1. Introduction
Drought is recognized as a global problem that threatens food and water security and affects people worldwide socially, economically, and environmentally (Taufik et al. 2020). Drought monitoring depends on the continuous collection and analysis of drought indicators that can provide information about the time of onset, end, and development of droughts to managers and planners in various sectors. The first small-scale drought monitoring systems operated based on the measurements of environmental variables such as precipitation, temperature, discharge, and groundwater level (Crocetti et al. 2020). Drought indices include the self-calibrating Palmer drought severity index (Palmer 1965), the surface water quality index (WQI; Shafer and Dezman 1982), the standardized precipitation index (SPI; McKee et al. 1993), and the standardized precipitation evapotranspiration index (SPEI; Vicente-Serrano et al. 2010a). Severe drought conditions have been found to mainly affect agriculture, the environment, and health, translating into severe socioeconomic consequences (Rahman and Lateh 2016; Mishra and Singh 2010; Dai 2013). In recent years, a lot of research has been done to predict droughts and assess drought vulnerability and population exposure in different parts of the world (e.g., Dabanli 2018; Wilhite 2019; Hagenlocher et al. 2019; Bokwa et al. 2021; Kubiak-Wójcicka et al. 2021; Jedd et al. 2021; Ai et al. 2021; Peltonen-Sainio et al. 2021; Byrareddy et al. 2021). Recently in Iran, Roshangar and Ghasempor (2021) during a study in Tabriz to model drought using SPI showed that the study of drought time series based on experimental methods leads to more accurate results. Also, climatic elements such as average monthly temperature and relative humidity as well as SPIs of previous months affect the prediction of drought, so modeling error increased due to their elimination. Their sensitivity analysis results showed that SPI is the most influential parameter in modeling. Speich (2019) showed that indices of calculating evaporation demand have better performance than rainfall indices and are more suitable than soil moisture storage indices. However, the comparison between drought and precipitation indices shows a significant divergence. Niu et al. (2020) applied SPEI, the soil water deficit index (SWDI), and the self-calibrating Palmer drought severity index (scPDSI) to evaluate the characteristics of agricultural and meteorological droughts in southwest China. The results indicated that the annual changes and growing seasons of three drought indices during 1982–2015 are equally consistent, which has important implications for monitoring drought and environmental sustainability in mountainous areas. In research in the Amazon and Mississippi River basins, Joetzjer et al. (2013) implemented widely operated drought criteria consisting of the standard runoff index (SRI) for river discharge and various meteorological drought indices (e.g., scPDSI, SPI, and SPEI). This study demonstrated that empirical meteorological drought indicators should be considered with caution, and statistical techniques of drought should be investigated to assess the characteristics of drought and climate change. Yan et al. (2020) utilized a PDSI model replacing potential evaporation with leaf area index–based total evapotranspiration (PDSIARTS), the Palmer drought severity index, and satellite surface indicators with monthly weather data from 2000 high-density stations in China during 1982–2016. This research shows an increase in temperature by 0.35°C decade−1 (significance level p < 0.001) in China; meanwhile, northern China became wetter over the same period. In East Africa, Haile et al. (2020) employed SPI and SPEI to measure drought severity. The results revealed that future droughts would be more pronounced due to temperature influences. They also indicated that the regions affected by extreme drought increased more rapidly than those affected by severe and moderate droughts. Mirgol et al. (2021), in a study to investigate the drought tendency of Lake Urmia utilizing scenarios RCP2.6, RCP4.5, and RCP8.5, standardized precipitation, and standardized precipitation–evapotranspiration indices, discovered that SPEI for drought assessment has more acceptable results than SPI. Investigating the drought of Darab City in Iran using the Coopras model, Attaee et al. (2022) indicated that the most destructive effects of this phenomenon were observed in surface and groundwater sources, the reduction of the per capita income of farmers, the abandonment of agricultural land, and the immigration rate of villagers to cities. Wang et al. (2022) applied the temperature–vegetation–precipitation drought index (TVPDI), the temperature–vegetation–soil moisture–precipitation drought index (TVMPDI), SPEI, and some of the remote sensing indices, such as the vegetation health index (VHI) and the scaled drought condition index (SDCI), to investigate the drought in southern Tibet. The TVMPDI trend showed that the drought rate decreased from 2003 to 2018. Moreover, the gross primary productivity (GPP) illustrated a negative correlation with SPEI (r = −0.4) and the soil moisture (SM) drought index. Recently, Das et al. (2023) used four common socioeconomic scenarios [shared socioeconomic pathway 1-2.6 (SSP1-2.6), SSP2-4.5, SSP3-7.0, and SSP5-8.5] along with 13 global climate models to investigate drought conditions in India. The results indicate that extreme drought (in 56%–72% of the area) and severe drought (in 99% of the area) are likely to increase under all scenarios for 3-month scale conditions. Population exposure to drought severity is predicted to increase for both conditions. A combined multivariate drought index (CMDI) was applied to assess drought in China during 2003–20 by Yang et al. (2023). The results indicated that the developed CMDI integrates hydrological and meteorological droughts using a combination of SPI, SRI, and the water storage deficit index (WSDI). CMDI also captures the overall drought more frequently than traditional drought indices.
Considering the continuation of the drought situation in Khorasan Razavi Province in recent years and the intensity of its effects on different economic and social sectors, the need to assess the vulnerability to drought doubles. Drought risk assessment is an essential measure to deal with drought and adjust its consequences to recognize and understand the vulnerability level of people in economic and social sectors and present an equitable drought risk management framework, which has not been investigated in previous studies in our study area. The major objectives of this study are 1) identifying the most severe temporal and spatial droughts, 2) assessing the risk of drought occurrence in different climatic regions of the province, 3) determining the socioeconomic vulnerability of the study area to drought and the population exposure, and 4) evaluating the population exposed to drought during the statistical period of 1950–2020. To this end, we applied various indices such as scPDSI, SPI, and stationary SPEI (sSPEI) and considered agricultural lands, water stress, population and livestock densities, and socioeconomic and infrastructural characteristics. The results of the utilized indices were compared to recognize the most effective indicator for drought monitoring and drought risk in different climatic subregions of the study area from 1950 to 2020. Our findings can provide more in-depth knowledge of drought change characteristics in the study area and be beneficial for policymakers as well as socioeconomic and environmental planners to manage and reduce drought damage in regional planning.
2. Materials and methods
a. Study area and data
Khorasan Razavi Province is an area of more than 127 000 km2 located between the geographical orbit of 33°52′ and 37°42′N and between 56°19′ and 61°16′E. According to the Köppen climate classification system, the province has a dry and semiarid climate because its drought coefficient values are classified into two categories, one and two. On one hand, the abnormal distribution of yearly precipitation during the cold season is one of the main characteristics of rainfall in this study region; on the other hand, torrential, short-term, and heavy rainfalls account for most of the annual precipitation ratio. Regarding temperature, it has hot summers and relatively cold winters, and the average annual temperature increases from north to south. The study area consists of 19 subbasins with different weather conditions based on the Köppen climate category system: the Khaf, Taybad, and Torbate Jam subregions with dry and cold climates (Bwk, Bwh); the Fariman, Sarakhs, Neishabour, and Torbate Heydariyeh subregions with moderate dry climates (Bsk); the subregions of Sabzevar and Rashtkhar with a cold and dry environment (Bwh); the subregions of Chenaran, Mashhad, and Kalat with dry and semimoderate weather (MR); the subarid regions of Bardaskan (CWA); the subarid areas of Khalilabad and Ferdows (PAK); and the arid subregions of Kashmar and Dargaz (CNW) and Gonabad (TIB) (Fig. 1). The drought situation of meteorology, hydrology, soil moisture, agriculture, or vegetation is determined by considering a single variable (rainfall only) or combining precipitation with other meteorological variables (evaporation, temperature, and relative humidity). In the present study, the monthly precipitation data of 176 rain gauges prepared by the Iranian Meteorological Organization (IRIMO) were used during 1950–2020. Satellite images include monthly normalized difference vegetation index (NDVI) and land surface temperature (LST) provided by MODIS satellite and monthly precipitation (mm h−1) with a spatial resolution of 0.25° from the TRMM satellite were employed. In addition, information on the population index, agricultural lands, assets, water resources, and infrastructure was provided by the Iranian Ministry of Agriculture–Jahad; also, population information from the National Organization for Civil Registration (Iranian Ministry of Interior) was used.
b. Overview of methods
Assessing the risk of an extreme event can be conducted by hazard, vulnerability, and risk analysis (Rajsekhar et al. 2015). The methodology of this study consists of several combined steps. First, meteorological data were analyzed by SPSS, and statistical deficiencies were reconstructed using the difference and ratio method. Second, scPDSI, SPI, and SPEI were used as proxies to quantify drought by Drought Indices Package (DIP) software; then, the Mann–Kendall (MK) test along with Sen’s slope estimator was employed to recognize the statistically significant trends of indicators; also, Spearman’s rank correlation was applied to check their correlation. A combination of indicators as well as population density (PDI), agricultural land (AL), livestock production (LPI), level of water stress (LWS), gross domestic product (GDP), poor population ratio (PHR), rural population ratio (RP), population over age 65 years (PA65), agricultural irrigation infrastructure (AIL), and per capita renewable freshwater resources (RIFW) is applied to study drought vulnerability and population exposure. Then, the population exposure to drought was considered by multiplying the drought hazard index (DHI) by PDI in each cell of the network; the values of PHR, RP, and PA65 are calculated similarly to drought-exposed indices. Last, drought risk and vulnerability zoning maps were provided using the Kriging method in ArcGIS.
c. scPDSI
This index introduces the K2 climate algorithm, which can correct the high frequency of events and perform automatic calibration of index behaviors at any location, focusing on hydrological calculations and the standardization process (Palmer 1965). The drought index calculated by scPDSI is generally more spatially compatible and, due to the consideration of rare event frequency, includes extreme wet and dry occasions (Zargar et al. 2011). The weakness of scPDSI is that it cannot detect the characteristics of several drought scales because it depends on a specific time scale (Vicente-Serrano et al. 2010b). Overall, the correlation between PDSI/scPDSI and SPEI at scales between 9 and 12 months is good and indicates that it can show water scarcity at these time scales (Chen and Sun 2015). In this research, we applied the meteorological dataset during the years 1950–2020, which is used by the initial version of the CRU time series (TS) 3.25 monthly dataset, and scPDSI ground surface data in the 0.5° × 0.5° spatial network. Big data (Big Earth 195) are used quantitatively, and then, their time series is fitted with a gamma distribution.
d. SPI
SPI was utilized to observe and describe drought based on precipitation data over a 70-yr analysis period. The positive values of SPI indicate a normal up-to-wet situation, and its negative values demonstrate a standard till dry state (Shah et al. 2015). This index shows the probability of precipitation in a specific period in a region, and also, its data are used to model drought diversity (Huang et al. 2021). McKee et al. (1993) proposed the index in 1993 and used it to assess climate change and drought. The precipitation time series can be converted to a standardized normal index based on climatic distribution to check the monthly rainfall or longer time series. Therefore, SPI is suitable for quantifying types of drought events (Lloyd-Hughes and Saunders 2002). The advantage of SPI relative to scPDSI is its simplicity; it is only a rainfall-based index that is used to describe drought events at different time scales. However, the role of other variables such as temperature is ignored, and it is inappropriate to analyze the amount of dryness/humidity at shorter time scales (Stagge et al. 2017). In this study, to investigate SPI, the amount of precipitation in the time series of 3, 6, 9, and 12 months (SPI03, SPI06, SPI09, and SPI12, respectively) was calculated for accumulation periods, and the gamma coefficient was used as input data. The monthly precipitation observations with a spatial resolution of 1° × 1° were applied from 1950 to 2020. Then, to match the available data with the scPDSI spatial resolution, the SPI data are calculated on a spatial network of 0.5° × 0.5°. The interpolated original SPI datasets show no significant differences or biases in the compared analysis.
e. SPEI
The calculation process of SPEI is similar to that of SPI, but it shows the anomalies of the difference between the amounts of precipitation. SPEI provides a more comprehensive description of dry or wet conditions due to potential evapotranspiration being mainly caused by temperature (Kingston et al. 2015). The monthly SPEI dataset has a spatial resolution of 0.5° × 0.5° and covers the time series from 1950 to 2020. SPEI is an index that measures the difference between precipitation and evapotranspiration over time and can be calculated at various scales of 1–48 months. The range of changes in SPEI is the same as that of the standardized precipitation index, but it has approximately removed the limitations of SPI. This index uses the simple equation of water balance at various time scales, that is, the difference between precipitation and potential evapotranspiration based on the Thornthwaite method. By considering the potential evapotranspiration (PET), the difference between precipitation P and PET is calculated. SPI and CRU TS data were used to examine the scPDSI and SPEI of many stations. Then, to ensure logical stability and compare the drought characteristics between the three utilized indices in this study, data were generated in regular network cells. SPEI is calculated using the difference between monthly precipitation and potential evapotranspiration in SPI and takes into account the temperature factor. The water balance difference is normalized as the logistic probability distribution to estimate the SPEI value.
f. Vulnerability analysis and drought risk assessment
In this research, population density is obtained as one of the drought exposure indices at the lattice scale, which can facilitate the estimation of the probability of population exposure to drought for each network cell. To investigate the relative probability of drought occurrence in the study area, we historically calculated the meteorological drought hazards, and three different indices (SPI12, SPEI12, and scPDSI) were applied to compute DHI. Initially, SPI/SPEI equal to −1.0 and scPDSI equal to −2.0 were identified as drought thresholds; then the severity of annual drought events in each cell network was estimated during the study period; last, the mean of the normalized sum was utilized as DHI in this study. Assessment of DHI is a network-based estimate calculated for the study area by the mean weighted regional DHI lattice for each basin, in which values of −1 and 0 were used for the highest and lowest drought risks, respectively.
3. Results
According to the results of Spearman’s rank correlation for time series (Table 1), scPDSI shows a significant correlation with SPEI in time series of 9 and 12 months based on the Student’s t test approach; the time scale of scPDSI is not constant in the study area. On a 3-month time scale, the correlation of scPDSI and SPI indices is low, although the coefficients presented in most cells in the network have passed a significance test at the level of 5% (effective degrees of freedom are considered). Drought index values have increased over the past 70 years, and a statistically significant upward trend has been seen for SPEI and SPI during the statistical period of 1950–2020. The Sen slope and MK trend analysis results showed that the MD slopes of SPI, SPEI, and scPDSI are greater than 0. Accordingly, the time series of drought indices have a statistically significant upward trend. Based on the results of Table 1, the drought trend from 1950 to 2020 represented a dry to semidry tendency.
Drought percentage trend during the seven decades for three levels of drought in 19 subregions based on SPI12, SPE12, and scPDSI.
The results of the spatial distribution of SPI12, SPEI12, and scPDSI annual trends (Fig. 2) indicate that the increasing trend of drought intensity is more evident in high-altitude areas (more than 1500 m) than in southeastern areas with low altitude (500 m), while the trend of change in intensity and duration of humidity showed opposite outcomes. The precipitation and temperature variations play different functions in the occurrence of the drought/humidity characteristics process. Based on this, the effects of temperature change on this trend are greater than the impacts of precipitation change in the study region, so the increase in temperature affects the tendency toward change in duration and intensity of drought/humidity that is in good agreement with the results of Manzano et al. (2019), Cui et al. (2021), and Saharwardi and Kumar (2022). A slight increase in the average precipitation reduces the drought duration and severity, while a significant increase in the average temperature leads to an increase in the severity and duration of drought. In regions with the CWA climatic code, the trend of drought and reduced precipitation is visible. According to the comparison of drought indices (Fig. 2), changes in drought conditions and the amount of humidity are directly related to each other so that drought increases with decreasing humidity. The potential evapotranspiration difference in the trend between SPI and SPEI demonstrates this issue to some extent. The precipitation difference indicates in the southern and southwestern regions of the study area a significant positive trend (90% confidence level), as do some features such as annual temperature, number of hot days, the longest warm period, and the longest dry period (Fig. 3). The temperature of the whole study area, especially in the southern and southwestern parts of the region, showed a significant increase between 1980 and 2007. On one hand, the annual cold days, cold periods, wet days, and wet periods indicated negative drought trends; on the other, no significant tendency is observed in the average annual precipitation. The increasing temperature of the study area could be influenced by several factors, like the expansion of urbanization industries in Khorasan Razavi Province. According to the results, the number of rainy days decreases and consequently the region tends to have longer dry periods and shorter wet periods, so the dry and wet periods will be longer and shorter in the study area, respectively.
Based on the results of utilized indices, a trend of increasing humidity is observed throughout the northern, northwestern, and northeastern parts of the region. Meanwhile, the southern, southwestern, and southeastern parts consider a drought trend. The finding of SPEI12 and scPDSI indices indicate a significant trend existence for wet periods in the western part of the study area.
4. Discussion
The drought risk and spatial risk patterns in all subbasins between 1950 and 2020 are presented in Figs. 4 and 5, respectively. In most parts of the province, the estimated risk level using SPEI12 is higher than that using SPI12 and scPDSI. Based on the results, the subbasins with moderate dry climatic conditions (Bsk), including Fariman (basin 11), Sarakhs (basin 4), Neishabour (basin 19), and Torbat Heydariyeh (basin 9), are located in areas with low drought risk values, whereas the cities of Chenaran (basin 14) and Kalat (basin 13) and the Mashhad metropolis (basin 12), with dry and semimoderate climates (MR), are in a middle drought risk class. Meanwhile, the principal part of the study area, including Khaf (basin 1), Taybad (basin 6), and Torbat-e Jam (basin 5), with dry and cold climates (Bwk, Bwh); Bardaskan (CWA); the Khalilabad subdry basins (basins 7 and 8) and Ferdows (basin 16) (PAK); the Kashmar dry subarea (basin 18) and Dargaz (basin 15) (CNW); and the Gonabad basin (basin 3) (TIB), are at high risk of drought.
Exposure to drought at the regional scale is indicated in Fig. 4, such that the lowest and highest drought exposures are shown by the indicator values of 0 and 1, respectively. The subbasins with a high population density are located near 36°N (Fig. 6a). Our findings showed that many agricultural lands are exposed to drought in the sparsely populated areas, except the MR area (Fig. 6b). This study demonstrated that although smallholder farmers have a moderate resistance index to agricultural drought, they have little drought resistance in terms of natural resilience. Consequently, it is recommended that national planners and policymakers focus on building economic, social, human, and capital bases to improve the welfare of smallholder farmers. Water stress is the significant environmental stress that affects agricultural production worldwide, especially in arid and semiarid regions (Boutraa et al. 2010). The highest score of the livestock production index in terms of water stress exposure has appeared in CNW, PAK, and CWA regions. The livestock production index is shown based on the LPI-country (Fig. 6d). According to the drought exposure index (DEI) composite results (Fig. 6e), the highest drought rates have appeared in two subbasins (1 and 7). Subbasins 10, 3, and 18 and some of the central subbasins of the region also show high drought; however, we observed the most drought in basin 1, which may be due to the high density of agricultural land and water stress exposure. The results of this study are consistent with the studies by Bryan et al. (2019) and Marengo et al. (2021). Therefore, high population density and agricultural lands are the principal determinants of drought conditions in the northern and northwestern basins.
Figure 7 presents the drought vulnerability maps of the six indicators and the derived hybrid map. Two economic factors, namely, GDP and PHR-country, have an opposite relationship with drought; the study region that is more affected by drought will face significant economic damage. The increase in the country’s GDP possibly causes poverty and more drought vulnerability and reduces the ability of drought resistance. The GDP per capita and poverty census ratios (Figs. 7a,b) show relatively high values in subbasins 1, 3, 6, 7, 8, and 10. The two social factors of the rural population (RP-country) and people aged 65 years (PA65-country) show an inverse relationship with vulnerability. Low vulnerability exists in the subbasin (12) of the Mashhad basin, which is relatively well advanced in terms of RP-country. Generally, the rural population in the study area is more developing (Figs. 7c,d). Based on Fig. 7, vulnerability indices for the years 2000–20 are provided by 1) GDP per capita, 2) GDP (poverty census ratio), 3) PHR-country (rural population), 4) RP-country (population age ≥ 65 years), 5) PA65-country (agricultural irrigation lands), 6) AIL-country (renewable domestic freshwater), 7) RIFW-country (combined vulnerability), and 8) DVI-country. They are more vulnerable to natural hazards due to their relatively lower incomes. Our findings revealed that both infrastructure indices “AIL-country and RIFW-country” are negatively related with drought; these results are consistent with the outcomes of Ashraf et al. (2021), de Brito et al. (2020), Johnson et al. (2019), and Buzási et al. (2021). A high vulnerability in areas of irrigated agricultural land and areas with renewable inland freshwater was observed, while the southern, southeast, and southwest parts of the study area have lower scores in terms of RIFW-country (Figs. 7e,f). The vulnerability map extracted from the combination of all economic, social, and infrastructural factors is shown in Fig. 7g. In general, according to the vulnerability map, almost the entire area has a high vulnerability to drought risk, except for the northern and northeastern regions, where the vulnerability is low to moderate. Figures 7e and 7g also confirm that the northern, northwestern, northeastern, and western regions are at moderate risk. The results also indicate the extreme migration of the residents of the villages of the study area to urban regions, which can be due to the decrease in farming activities and animal husbandry, decrease in income, and lack of job opportunities due to frequent droughts. Accordingly, the population in the eastern and southern parts of the province is more exposed to drought. The continuation of this process will cause, on the one hand, the emptying of villages and, on the other hand, the excessive growth of the population in the cities and increase the informal settlements; as a result, social and economic damages will increase in both villages and cities of the province.
The results of SPI12 in Fig. 8a show the vital role of rainfall in the occurrence of drought risk. According to the results, the highest drought risk occurred in the southern, western, and central parts (Figs. 7 and 8), while the northern and eastern regions are at moderate drought risk. Drought risk assessment based on SPEI12 reveals a higher drought intensity relative to that based on the SPI12 and scPDSI indicators, which is in good agreement with the results of Kourakos et al. (2019), Li et al. (2020), MajidiRad and Rahimi (2021), Parchizadeh and Belant (2021), and Zengir and Sobhani (2020). A tendency toward reduced drought risk in the CNE, CNW, and Bwk regions was shown by SPI12. The frequency of dry and wet periods is high in short-term time scales, while with increasing time scales, dry/wet periods and their continuity decrease and increase, respectively. The results showed that although the standardized precipitation index has a significant correlation, the standardized precipitation evapotranspiration index has a faster response to drought. On the other hand, drought duration increases in the region due to an increasing time scale, which indicates the main changes in drought/humidity conditions. The studies conducted by Heydari Alamdarloo et al. (2020), Chiang et al. (2021), and Spinoni et al. (2020) achieved similar findings. As can be seen, the number of months calculated in various stations in severe and very severe drought classes is different; also, the average minimum and maximum values of station drought are greater. These can be attributed to the consideration of temperature factors in the determination of dry and wet periods and the importance of potential evapotranspiration calculated in the desired stations. Future drought events are likely to occur due to increased temperature and potential evapotranspiration in the study area, which is consistent with the findings of Pei et al. (2017) and Wu et al. (2015). The correlation degree between the determined months of wet and drought periods is different according to SPI and SPEI. Thus, the lowest correlation existed between the SPI wet period continuity and the highest correlation is related to the SPEI drought period continuity. SPEI was able to identify severe and very severe droughts at all the studied stations, which is in good agreement with the results of Zengir et al. (2020), Yazdani et al. (2021), and Sorkhabi (2021). Overall, the average monthly temperature of the province showed a negative relationship with latitude; on the other hand, the altitude factor also causes spatial anomalies of precipitation in the study area. Drought intensity is slowly worsening in the southern part of the province, and the decreased trends are shifted from northwest to northeast and eventually become lower in the northern part of the province. Drought severity and duration results indicate that the trend coefficient significantly increased from east to west. Based on the maps of spatial distribution trends of drought intensity and frequency in different periods, the central, eastern, western, and southern regions are susceptible to drought occurrence.
5. Conclusions
In the current research, we applied standard indices to analyze the drought in Khorasan Razavi Province and evaluated their relationships with different variables. The longest drought periods in most parts of the study area occurred from 1990 to 2015 and showed the highest correlation with SPI12. Increasing drought periods can be attributed to expanding agricultural water consumption in the region, which can influence crop yields. On the other hand, the study area is strongly affected by anthropogenic activities that can influence the treatment of climatic parameters. More drought conditions in the future were shown by SPEI than by SPI, which indicates an increase in evapotranspiration potential relative to precipitation. SPI and SPEI show a significant difference in drought characteristics in some parts of the study area; the results analysis showed that SPI decreases continuously during the statistical period; meanwhile, SPEI indicates an increasing trend since the year 1981. Based on the results of SPEI, drought is greater in the southern and central parts than in the eastern regions. Assessment of drought vulnerability and risk at different spatiotemporal scales (e.g., increased socioeconomic and environmental effects of drought) can increase the conception of future drought hazards. The drought increase in the study area is primarily due to the increase in evaporation because of high temperatures, so the use of SPEI can be beneficial for understanding future drought characteristics in the area. Due to the geographical diversity of drought risk, the spatial variability of factors affecting drought risk assessment should be considered as a significant component. The high-risk regions shown in Fig. 8 are exposed to a high risk of drought vulnerability, which leads to water scarcity. Therefore, the adaptation of the water resource infrastructure of the southern region, which has limited water resources, according to its climatic conditions is recommended. For instance, rainwater harvesting techniques, greenhouse cultivation development, modern irrigation methods, use of drought-resistant plants, artificial recharge methods, and so on could be beneficial in reducing drought effects. Greater percentages of subbasins 10, 11, 12, 13, 17, and 19 are surrounded by agricultural lands and have suffered a lot of drought damage over time, so controlling groundwater extraction, protecting the farmlands, selecting water-deficient plants, improving the agricultural insurance, and so on can significantly reduce drought effects in these subbasins. Although the different indices have a high correlation together, the maximum correlation of drought change has occurred between scPDSI and SPEI with the time scales of 9 and 12 months; also, the highest correlation between SPI and scPDSI/SPEI was observed at the 3-month time scale. The precipitation values, especially for short-term drought events, indicate that the eastern regions including Bwk, CWA, and Bwh were affected by a drying stage during the study period. According to the research results, significant moisture is observed in many northern and northwestern regions, and in the south and southeast, droughts have occurred with more frequency. In recent decades, subbasins 1 and 6 have experienced long-term drought events, while shorter-term drought events are more common in the north, northwest, and northeast of the region. Also, severe drought events have an increasing trend in the southern and central parts during the years 1991–2005, and high drought risks were observed for the southern parts of the province. Although the parameters of vulnerability and drought exposure cannot fully assess the unfavorable effects of drought risk, socioeconomic conditions can help drought monitoring. Drought exposure in densely populated basins, vast agricultural lands, and areas facing water stress are mainly located in the northern and western parts. However, the assessment of vulnerability reflected by the six factors shows a slight difference between the subbasins of the province in the results; most subbasins, except limited parts of the northern region, are facing high drought vulnerability. Since the drought risk in our study region is high and population density appears in the north, northwest, and northeast, economic development has not reduced the effects of drought on citizens. In addition, the risk assessment in this study includes the development of drought policy, climate change, and a variety of indicators of drought exposure and vulnerability. High-resolution socioeconomic data significantly improve the reliability of drought assessment. Although the present study focuses on one province, results can be extended to other arid regions of Iran that are climatically similar to our study area. Due to the existence of wet periods during the years 2018–20 in the study area, at the end of the twenty-first century, drought levels in the north and northwest of the studied area, which are located in the highlands and have monsoon rains, will probably decrease. The present research gave us useful insight to understand the possible conditions for future drought and identify activities that can be applied for drought risk reduction, for instance, establishing a drought monitoring and warning system, constructing non-water-related jobs, educating communities to reduce water use, applying modern irrigation techniques, and preventing the creation of water-intensive industries. The main limitations facing this study were the reconstruction of the missing data and the prolongation of the short-term statistical periods in some stations, which could be primary sources of error.
Acknowledgments.
The authors appreciate the IRIMO and Iran Water Resources Company for providing the observational data for the study area. The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Data availability statement.
The authors do not have permission to share used data.
REFERENCES
Ai, P., B. Chen, D. Yuan, M. Hong, and H. Liu, 2021: Dynamic risk assessment of drought disaster: A case study of Jiangxi Province, China. J. Water Climate Change, 12, 1761–1777, https://doi.org/10.2166/wcc.2020.141.
Ashraf, S., A. Nazemi, and A. AghaKouchak, 2021: Anthropogenic drought dominates groundwater depletion in Iran. Sci. Rep., 11, 9135, https://doi.org/10.1038/s41598-021-88522-y.
Attaee, H., A. Bostany, R. Soltani, and G. Salahi, 2022: Assessing the vulnerability of rural settlements to drought using Coopras model (case study: Darab township). Phys. Soc. Plann., 9, 73–86, https://doi.org/10.30473/psp.2022.53248.2307.
Bokwa, A., M. Klimek, P. Krzaklewski, and W. Kukułka, 2021: Drought trends in the Polish Carpathian Mts. in the years 1991–2020. Atmosphere, 12, 1259, https://doi.org/10.3390/atmos12101259.
Boutraa, T., A. Akhkha, A. A. Al-Shoaibi, and A. M. Alhejeli, 2010: Effect of water stress on growth and water use efficiency (WUE) of some wheat cultivars (Triticum durum) grown in Saudi Arabia. J. Taibah Univ. Sci., 3, 39–48, https://doi.org/10.1016/S1658-3655(12)60019-3.
Bryan, K., S. Ward, S. Barr, and D. Butler, 2019: Coping with drought: Perceptions, intentions and decision-stages of south west England households. Water Resour. Manage., 33, 1185–1202, https://doi.org/10.1007/s11269-018-2175-2.
Buzási, A., T. Pálvölgyi, and D. Esses, 2021: Drought-related vulnerability and its policy implications in Hungary. Mitigation Adapt. Strategies Global Change, 26, 11, https://doi.org/10.1007/s11027-021-09943-8.
Byrareddy, V., L. Kouadio, S. Mushtaq, J. Kath, and R. Stone, 2021: Coping with drought: Lessons learned from robusta coffee growers in Vietnam. Climate Serv., 22, 100229, https://doi.org/10.1016/j.cliser.2021.100229.
Chen, H., and J. Sun, 2015: Changes in drought characteristics over China using the standardized precipitation evapotranspiration index. J. Climate, 28, 5430–5447, https://doi.org/10.1175/JCLI-D-14-00707.1.
Chiang, F., O. Mazdiyasni, and A. AghaKouchak, 2021: Evidence of anthropogenic impacts on global drought frequency, duration, and intensity. Nat. Commun., 12, 2754, https://doi.org/10.1038/s41467-021-22314-w.
Crocetti, L., and Coauthors, 2020: Earth observation for agricultural drought monitoring in the Pannonian basin (southeastern Europe): Current state and future directions. Reg. Environ. Change, 20, 123, https://doi.org/10.1007/s10113-020-01710-w.
Cui, Y., B. Zhang, H. Huang, J. Zeng, X. Wang, and W. Jiao, 2021: Spatiotemporal characteristics of drought in the North China Plain over the past 58 years. Atmosphere, 12, 844, https://doi.org/10.3390/atmos12070844.
Dabanli, I., 2018: Drought risk assessment by using drought hazard and vulnerability indexes. Nat. Hazards Earth Syst. Sci. Discuss, https://doi.org/10.5194/nhess-2018-129.
Dai, A., 2013: Increasing drought under global warming in observations and models. Nat. Climate Change, 3, 52–58, https://doi.org/10.1038/nclimate1633.
Das, J., S. Das, and N. V. Umamahesh, 2023: Population exposure to drought severities under shared socioeconomic pathways scenarios in India. Sci. Total Environ., 867, 161566, https://doi.org/10.1016/j.scitotenv.2023.161566.
de Brito, M. M., C. Kuhlicke, and A. Marx, 2020: Near-real-time drought impact assessment: A text mining approach on the 2018/19 drought in Germany. Environ. Res. Lett., 15, 1040a9, https://doi.org/10.1088/1748-9326/aba4ca.
Hagenlocher, M., I. Meza, C. C. Anderson, A. Min, F. G. Renaud, Y. Walz, S. Siebert, and Z. Sebesvari, 2019: Drought vulnerability and risk assessments: State of the art, persistent gaps and research agenda. Environ. Res. Lett., 14, 083002, https://doi.org/10.1088/1748-9326/ab225d.
Haile, G. G., and Coauthors, 2020: Projected impacts of climate change on drought patterns over East Africa. Earth’s Future, 8, e2020EF001502, https://doi.org/10.1029/2020EF001502.
Heydari Alamdarloo, E., H. Khosravi, S. Nasabpour, and A. Gholami, 2020: Assessment of drought hazard, vulnerability and risk in Iran using GIS techniques. J. Arid Land, 12, 984–1000, https://doi.org/10.1007/s40333-020-0096-4.
Huang, W., J. Yang, Y. Liu, and E. Yu, 2021: Spatiotemporal variations of drought in the arid region of northwestern China during 1950–2012. Adv. Meteor., 2021, 6680067, https://doi.org/10.1155/2021/6680067.
Jedd, T., S. Russell Fragaszy, C. Knutson, M. J. Hayes, M. Belhaj Fraj, N. Wall, M. Svoboda, and R. McDonnell, 2021: Drought management norms: Is the Middle East and North Africa region managing risks or crises? J. Environ. Dev., 30, 3–40, https://doi.org/10.1177/1070496520960204.
Joetzjer, E., H. Douville, C. Delire, P. Ciais, B. Decharme, and S. Tyteca, 2013: Hydrologic benchmarking of meteorological drought indices at interannual to climate change timescales: A case study over the Amazon and Mississippi River basins. Hydrol. Earth Syst. Sci., 17, 4885–4895, https://doi.org/10.5194/hess-17-4885-2013.
Johnson, S. J., and Coauthors, 2019: SEAS5: The new ECMWF seasonal forecast system. Geosci. Model Dev., 12, 1087–1117, https://doi.org/10.5194/gmd-12-1087-2019.
Kingston, D. G., J. H. Stagge, L. M. Tallaksen, and D. M. Hannah, 2015: European-scale drought: Understanding connections between atmospheric circulation and meteorological drought indices. J. Climate, 28, 505–516, https://doi.org/10.1175/JCLI-D-14-00001.1.
Kourakos, G., H. E. Dahlke, and T. Harter, 2019: Increasing groundwater availability and seasonal base flow through agricultural managed aquifer recharge in an irrigated basin. Water Resour. Res., 55, 7464–7492, https://doi.org/10.1029/2018WR024019.
Kubiak-Wójcicka, K., A. Pilarska, and D. Kamiński, 2021: The analysis of long-term trends in the meteorological and hydrological drought occurrences using non-parametric methods—Case study of the catchment of the Upper Noteć River (central Poland). Atmosphere, 12, 1098, https://doi.org/10.3390/atmos12091098.
Li, J., Z. Wang, X. Wu, J. Chen, S. Guo, and Z. Zhang, 2020: A new framework for tracking flash drought events in space and time. Catena, 194, 104763, https://doi.org/10.1016/j.catena.2020.104763.
Lloyd-Hughes, B., and M. A. Saunders, 2002: A drought climatology for Europe. Int. J. Climatol., 22, 1571–1592, https://doi.org/10.1002/joc.846.
MajidiRad, N., and S. Rahimi, 2021: Hermeneutic and climatology hazards; investigation of temporal–spatial variations of high-altitude submarine and its impacts on drought occurrence. Geogr. Environ. Hazards, 10, 119–141, https://doi.org/10.22067/GEOEH.2021.67027.0.
Manzano, A., M. A. Clemente, A. Morata, M. Y. Luna, S. Beguería, S. M. Vicente-Serrano, M. L. Martín, 2019: Analysis of the atmospheric circulation pattern effects over SPEI drought index in Spain. Atmos. Res., 230, 104630, https://doi.org/10.1016/j.atmosres.2019.104630.
Marengo, J., and Coauthors, 2021: Extreme drought in the Brazilian Pantanal in 2019–2020: Characterization, causes, and impacts. Front. Water, 3, 639204, https://doi.org/10.3389/frwa.2021.639204.
McKee, T. B., N. J. Doesken, and J. Kleist, 1993: The relationship of drought frequency and duration to time scales. Eighth Conf. on Applied Climatology, Anaheim, CA, Amer. Meteor. Soc., 179–184.
Mirgol, B., M. Nazari, H. Ramezani Etedali, and K. Zamanian, 2021: Past and future drought trends, duration, and frequency in the semi-arid Urmia Lake basin under a changing climate. Meteor. Appl., 28, e2009, https://doi.org/10.1002/met.2009.
Mishra, A. K., and V. P. Singh, 2010: A review of drought concepts. J. Hydrol., 391, 202–216, https://doi.org/10.1016/j.jhydrol.2010.07.012.
Niu, Q., L. Liu, J. Heng, H. Li, and Z. Xu, 2020: A multi-index evaluation of drought characteristics in the Yarlung Zangbo River basin of Tibetan Plateau, southwest China. Front. Earth Sci., 8, 213, https://doi.org/10.3389/feart.2020.00213.
Palmer, W. C., 1965: Meteorological drought. U.S. Weather Bureau Research Paper 45, 58 pp., https://www.droughtmanagement.info/literature/USWB_Meteorological_Drought_1965.pdf.
Parchizadeh, J., and J. L. Belant, 2021: Iran: Drought must top new government’s agenda. Nature, 596, 189, https://doi.org/10.1038/d41586-021-02189-z.
Pei, W., Q. Fu, D. Liu, T. Li, K. Cheng, and S. Cui, 2017: Spatiotemporal analysis of the agricultural drought risk in Heilongjiang Province, China. Theor. Appl. Climatol., 133, 151–164, https://doi.org/10.1007/s00704-017-2182-x.
Peltonen-Sainio, P., J. Juvonen, N. Korhonen, P. Parkkila, J. Sorvali, and H. Gregow, 2021: Climate change, precipitation shifts and early summer drought: An irrigation tipping point for Finnish farmers? Climate Risk Manage., 33, 100334, https://doi.org/10.1016/j.crm.2021.100334.
Rahman, M. R., and H. Lateh, 2016: Meteorological drought in Bangladesh: Assessing, analyzing and hazard mapping using SPI, GIS and monthly rainfall data. Environ. Earth Sci., 75, 1026, https://doi.org/10.1007/s12665-016-5829-5.
Rajsekhar, D., V. P. Singh, and A. K. Mishra, 2015: Integrated drought causality, hazard, and vulnerability assessment for future socioeconomic scenarios: An information theory perspective. J. Geophys. Res. Atmos., 120, 6346–6378, https://doi.org/10.1002/2014JD022670.
Roshangar, K., and R. Ghasempor, 2021: Drought modeling based on SPI index using terrestrial and satellite data using the integrated GPR-CEEMD model. J. Irrig. Water Eng., 11, 295–315, https://doi.org/10.22125/IWE.2021.133765.
Saharwardi, M. S., and P. Kumar, 2021: Future drought changes and associated uncertainty over the homogenous regions of India: A multimodel approach. Int. J. Climatol., 42, 652–670, https://doi.org/10.1002/joc.7265.
Shafer, B. A., and L. E. Dezman, 1982: Development of a surface water supply index (SWSI) to assess the severity of drought conditions in snowpack runoff areas. Proc. 50th Annual Western Snow Conf., Reno, NV, Western Snow Conference, 164–175, https://westernsnowconference.org/node/932.
Shah, R., N. Bharadiya, and V. Manekar, 2015: Drought index computation using standardized precipitation index (SPI) method for Surat District, Gujarat. Aquat. Procedia, 4, 1243–1249, https://doi.org/10.1016/j.aqpro.2015.02.162.
Sorkhabi, O. M., 2021: Iran drought monitoring in April 2021. Research Square, 7 pp., https://doi.org/10.21203/rs.3.rs-452837/v1.
Speich, M. J. R., 2019: Quantifying and modeling water availability in temperate forests: A review of drought and aridity indices. IForest Biogeosci. For., 12, 1–16, https://doi.org/10.3832/ifor2934-011.
Spinoni, J., and Coauthors, 2020: Future global meteorological drought hot spots: A study based on CORDEX data. J. Climate, 33, 3635–3661, https://doi.org/10.1175/JCLI-D-19-0084.1.
Stagge, J. H., D. G. Kingston, L. M. Tallaksen, and D. M. Hannah, 2017: Observed drought indices show increasing divergence across Europe. Sci. Rep., 7, 14045, https://doi.org/10.1038/s41598-017-14283-2.
Taufik, M., B. Minasny, A. B. McBratney, J. C. Van Dam, P. D. Jones, and H. A. J. Van Lanen, 2020: Human-induced changes in Indonesian peatlands increase drought severity. Environ. Res. Lett., 15, 084013, https://doi.org/10.1088/1748-9326/ab96d4.
Vicente-Serrano, S. M., S. Beguería, and J. I. López-Moreno, 2010a: A multiscalar drought index sensitive to global warming: The standardized precipitation evapotranspiration index. J. Climate, 23, 1696–1718, https://doi.org/10.1175/2009JCLI2909.1.
Vicente-Serrano, S. M., S. Beguería, J. I. López-Morenon, M. Angulo, and A. El Kenawy, 2010b: A new global 0.5 gridded dataset (1901–2006) of a multiscalar drought index: Comparison with current drought index datasets based on the Palmer drought severity index. J. Hydrometeor., 11, 1033–1043, https://doi.org/10.1175/2010JHM1224.1.
Wang, Z., Z. Wang, J. Xiong, W. He, Z. Yong, and X. Wang, 2022: Responses of the remote sensing drought index with soil information to meteorological and agricultural droughts in southeastern Tibet. Remote Sens., 14, 6125, https://doi.org/10.3390/rs14236125.
Wilhite, D. A., 2019: Integrated drought management: Moving from managing disasters to managing risk in the Mediterranean region. Euro-Mediterr. J. Environ. Integr., 4, 42, https://doi.org/10.1007/s41207-019-0131-z.
Wu, D., X. Zhao, S. Liang, T. Zhou, K. Huang, B. Tang, and W. Zhao, 2015: Time-lag effects of global vegetation responses to climate change. Global Change Biol., 21, 3520–3531, https://doi.org/10.1111/gcb.12945.
Yan, H., S.-Q. Wang, J.-B. Wang, A.-H. Guo, Z.-C. Zhu, R. B. Myneni, and H. H. Shugart, 2020: Recent wetting trend in China from 1982 to 2016 and the impacts of extreme El Niño events. Int. J. Climatol., 40, 5485–5501, https://doi.org/10.1002/joc.6530.
Yang, B., and Coauthors, 2023: Combined multivariate drought index for drought assessment in China from 2003 to 2020. Agric. Water Manage., 281, 108241, https://doi.org/10.1016/j.agwat.2023.108241.
Yazdani, M. H., K. Amininia, V. Safarianzengir, N. Soltani, and H. Parhizkar, 2021: Analyzing climate change and its effects on drought and water scarcity (case study: Ardabil, northwestern province of Iran, Iran). Sustainable Water Resour. Manage., 7, 16, https://doi.org/10.1007/s40899-021-00494-z.
Zargar, A., R. Sadiq, B. Naser, and F. Khan, 2011: A review of drought indices. Environ. Rev., 19, 333–349, https://doi.org/10.1139/a11-013.
Zengir, V. S., and B. Sobhani, 2020: Simulation and analysis of natural hazard phenomenon, drought in southwest of the Caspian Sea, Iran. Carpathian J. Earth Environ. Sci., 15, 127–136, https://doi.org/10.26471/cjees/2020/015/115.
Zengir, V. S., B. Sobhani, and S. Asghari, 2020: Monitoring and investigating the possibility of forecasting drought in the western part of Iran. Arabian J. Geosci., 13, 493, https://doi.org/10.1007/s12517-020-05555-9.