Introducing a Novel Standardized Precipitation Evaporation Differential Index (SPEDI) for Improved Flash Drought Detection and Assessment: A Case Study in South Korea

Introducing a Novel Standardized Precipitation Evaporation Differential Index (SPEDI) for Improved Flash Drought Detection and Assessment: A Case Study in South Korea | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Introducing a Novel Standardized Precipitation Evaporation Differential Index (SPEDI) for Improved Flash Drought Detection and Assessment: A Case Study in South Korea Subin Kang, Ho-Jun Kim, Joo-Heon Lee, Hyeon-Cheol Yoon, Hyun-Han Kwon This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4405968/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 09 Sep, 2025 Read the published version in Stochastic Environmental Research and Risk Assessment → Version 1 posted 3 You are reading this latest preprint version Abstract This study introduces the Standardized Precipitation Evaporation Differential Index (SPEDI), a new composite drought index designed to better capture flash drought conditions by accounting for both precipitation deficits and evaporative demand. The performance of SPEDI is compared with three established indices—the Standardized Precipitation Index (SPI), the Standardized Precipitation Evapotranspiration Index (SPEI), and the Evaporative Demand Drought Index (EDDI)—in detecting and characterizing flash droughts across South Korea. To evaluate their effectiveness, we analyzed spatial and temporal patterns of flash droughts and compared outputs from each index with records of agricultural damage during four representative years (2017, 2018, 2019, and 2022). Results show that while SPI and SPEI often fail to capture the rapid onset and intensity of flash droughts, EDDI and SPEDI more reliably reflect observed impacts. In particular, SPEDI demonstrated the highest agreement with actual crop damage, offering early warnings in 12 out of 14 events and accurate timing in 10 out of 14 cases. A simplified model examining the relationship between precipitation and evaporative demand further supports SPEDI’s improved ability to represent water balance conditions, especially under high evaporative stress, where SPEI tends to fall short. These results suggest that SPEDI offers a more accurate and practical approach for identifying flash droughts, with clear potential for use in early warning systems and drought risk management. This study presents SPEDI as a valuable tool for supporting drought response planning and improving understanding of flash drought behavior in a changing climate. Flash Drought SPEDI Drought Detection EDDI SPEI and SPI Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 1. Introduction Drought is a recurring natural hazard that affects a wide range of climate zones (Naveendrakumar et al., 2019 ). Traditionally, droughts have been classified into four types based on their distinctive characteristics and underlying causes (Wilhite and Glantz, 1985 ): meteorological, agricultural, hydrological, and socioeconomic droughts. More recently, the recognition of “flash droughts”—events characterized by rapid onset and intensification—has led to refinements in this classification system (Otkin et al., 2018 ; Pendergrass et al., 2020 ). In contrast to the gradual development of typical drought events, flash drought can emerge in just two to six weeks, making timely detection and response particularly difficult. In particular, after a drought event in the US Central Plains in 2012, the sudden onset of the droughts during a short period was identified as a flash drought (Lisonbee et al., 2022 ). This recognition has led to a significant increase in studies that revisit several historical flash droughts. Flash droughts can develop even under varying initial moisture conditions (Christian et al., 2019 ). Despite the fundamental requirement of precipitation deficit for drought events, the emergence of flash drought is contingent upon additional climatic factors. The absence of rainfall, coupled with factors contributing to heightened evaporative demand ( \(\:{ET}_{0}\) ), such as elevated surface temperatures (i.e., severe heatwave), strong wind velocities, and clear skies persisting over several weeks, plays a crucial role in the rapid-onset and fast intensification of flash droughts (Otkin et al., 2018 ; Christian et al., 2019 ; Svoboda et al., 2002 ). Especially high temperatures can create conditions vulnerable to drought(Hu et al., 2023 ; Zhang et al., 2022 ; Won et al., 2020 ), making it more likely for flash droughts to occur. Under elevated \(\:{\text{E}\text{T}}_{0}\) conditions, a consequential depletion in soil moisture content occurs, leading to a simultaneous increase in evapotranspiration (ET) (Otkin et al., 2013 ; Anderson et al., 2013 ). If this combination persists for an extended duration (days to weeks), it has the potential to transform energy-limited conditions into water-limited conditions. This transition intensifies vegetation stress and eventually triggers a flash drought (Ford and Labosier, 2017 ; Ford et al., 2015 ; Mozny et al., 2012 ; Hunt et al., 2009 ; Hunt et al., 2014 ). Flash droughts can be further categorized into two types based on specific physical mechanisms. The first category, known as “heatwave flash drought,” is characterized by exceptionally high temperatures driving an increase in ET and a simultaneous decrease in soil moisture. Heatwave flash droughts tend to be more pronounced in densely vegetated areas with comparatively high humidity. The second category, termed “precipitation-deficit flash drought,” is associated with a deficiency in precipitation, leading to reduced soil moisture and a subsequent decrease in ET (Wang and Yuan, 2018 ; Mo and Lettenmaier, 2016 ; Zhang et al., 2022 ). Understanding both types requires tools that can account for the interaction between precipitation and evaporative demand. In this study, we focus on South Korea, a country particularly susceptible to flash droughts due to its climate and topography. Most of the annual rainfall is concentrated in the summer months (June to August), while the rest of the year, especially winter and early spring, tends to be extremely dry (Kim et al., 2014 ). In addition, nearly 70% of the country’s terrain is mountainous, which complicates precipitation patterns and hydrological analysis. These geographical characteristics were evident in 2022, when the country experienced severe drought in the south and flooding in central regions. Reports of flash drought impacts have become more frequent (KMA, 2017 ; Lee et al., 2021 ), highlighting the need for improved assessment tools. However, most drought-related studies in South Korea focus on seasonal or long-term trends, with flash droughts remaining underexamined, accounting for only 4.11% of global drought studies according to the Web of Science (as of April 2025). A key step toward better flash drought management is having a clear definition that distinguishes flash droughts from general droughts. The definition of flash drought includes rapid onset and intensification, and various studies have proposed different indicators for identification. For instance, Ford and Labosier ( 2017 ) utilized soil moisture observations at 0-40cm and quantified temporal differences in soil moisture to assess flash droughts in the United States. Conversely, Otkin et al. ( 2013 ) employed the evaporative stress index (ESI) as an alternative to soil moisture observations to assess flash drought. Pendergrass et al. ( 2020 ) proposed using the evaporative demand drought index (EDDI) based on standardized anomalies of \(\:{\text{E}\text{T}}_{0}\) , highlighting shifts toward moisture-stressed environments. Noguera et al. ( 2020 ) evaluated flash drought in Spain using the Standardized Precipitation Evapotranspiration Index (SPEI; Vicente-Serrano et al., 2010 ), considering the effects of both precipitation and ET. Parker et al. ( 2021 ) conducted a study in the Australian region, investigating flash drought detection criteria based on standardized precipitation index (SPI; McKee et al., 1993 ) and EDDI, ESI, and standardized soil-moisture index (SSI). In line with earlier efforts, our study evaluates the suitability of commonly used drought indices—SPI, SPEI, and EDDI—for detecting flash droughts. SPI, based solely on precipitation, remains a widely used benchmark for meteorological droughts (McKee et al., 1993 ; Kwon et al., 2019 ; Abu Arra and Şişman, 2024 ; Arra et al., 2024 ). Its simplicity, however, limits its ability to capture rapid drought development influenced by atmospheric demand. To address this, SPEI was developed by Vicente-Serrano et al. ( 2010 ) to incorporate both precipitation and potential evapotranspiration (PET), offering a more comprehensive perspective (Mozny et al., 2012 ; Qin et al., 2023 ). Yet, concerns remain over its reliability, especially when PET is overestimated. McEvoy et al. ( 2012 ) suggested using reference evapotranspiration (ET₀) as a more accurate input—a method adopted in this study. EDDI, unlike SPI and SPEI, is based entirely on ET₀ and represents atmospheric moisture demand through anomalies in water vapor flux. It was developed by Hobbins et al. ( 2016 ) and is particularly suited to identifying rapid drought developments. While these indices each capture important drought signals, SPI and EDDI rely on single variables and may fall short when applied independently. SPEI attempts to bridge this gap but has been shown to correlate more closely with SPI than with EDDI, suggesting it may not fully capture the interplay between precipitation and evaporative demand (Yao et al., 2018 ; Won et al., 2020 ). To address the limitations of existing drought indices, we propose a new composite index: the Standardized Precipitation Evaporation Differential Index (SPEDI). Designed to combine the benefits of SPI and EDDI, SPEDI highlights conditions in which low precipitation coincides with high evaporative demand—characteristics typical of flash droughts. By independently calculating and then differencing SPI and EDDI, SPEDI retains the key features of both, offering a more balanced and responsive measure. This study investigates flash drought behavior in South Korea by analyzing four indices—SPI, SPEI, EDDI, and SPEDI—across multiple seasonal timescales. Our objectives are threefold: First, we seek to map the spatiotemporal patterns of flash droughts to better understand their frequency and distribution across South Korea. Second, we evaluate the effectiveness of the SPI, SPEI, EDDI, and SPEDI indices in capturing seasonal differences in flash drought behavior. Third, we explore the key meteorological drivers that influence the seasonal and regional variations of flash droughts, with an emphasis on improving early detection and management. To evaluate the effectiveness of each drought index, we use the observed agricultural damage area as a proxy for flash drought severity. This allows us to assess how well each index captures drought conditions that result in actual impacts. In doing so, we aim to demonstrate the practical advantages of SPEDI compared to conventional indices. The study is structured around four key steps: (1) development of the SPEDI index, (2) definition of flash drought criteria, (3) identification of flash drought events in South Korea, and (4) comparative analysis of index performance, as summarized in Fig. 1 . [Insert Fig. 1 ] 2. Study Area and Meteorological Data 2.1 Study area South Korea spans a peninsular area of approximately 100,364 km 2 , with its northern border adjacent to the continent and all remaining borders surrounded by sea. The topography is approximately 30% flat terrain and 70% mountainous regions. Despite the abundance of mountains, only 15% of them surpass an elevation of 1,000 meters above sea level, and more than 65% peak below 500 meters. Of note is the distinctive topographical layout of the Korean Peninsula, which features prominent mountainous regions primarily lining the eastern border of the peninsula, as exemplified by the Taebaek Mountains. The key characteristic of the peninsula’s river systems is the geographical orientation of their main watersheds, with the Taebaek Mountains limiting the area for the watershed toward the East Sea. Consequently, major rivers predominantly flow southwestward, ultimately connecting to the Yellow or South Seas. In contrast, those rivers that discharge into the East Sea follow shorter courses. In addition to topography, substantial seasonal variations in precipitation and temperature shape hydrological characteristics. Located within the geographic coordinates of 33–38°N and 126–132°E, the region experiences significant influence from monsoon winds, leading to cold and dry conditions in winter and a hot and humid climate in summer. The annual precipitation averages around 1,306mm, accompanied by a mean temperature range of 7–15℃. Approximately 54% of the annual precipitation falls during the summer season, establishing it as the dominant period for rainfall (Kwon et al., 2016 ). The mean annual temperature remains consistently at 13℃, with seasonal temperature extremes ranging from 1℃ in winter to 24℃ during the summer. These climatic variations contribute to the dynamic hydrological patterns observed in South Korea. [Insert Fig. 2 ] 2.2 Meteorological Dataset This study utilized data from 59 Automated Synoptic Observing Service (ASOS) stations operated by the Korea Meteorological Administration. The ASOS network, initially established in the early 1900s, provides consistent and systematic records of meteorological variables, including precipitation, solar radiation, and air temperature, at both hourly and daily resolutions. For the primary drought analysis, we used daily meteorological data from 1980 to 2022 to ensure sufficient temporal coverage for retrospective evaluation. In addition, a subset of data from 2017 to 2022 was used to align with the available validation dataset on agricultural damage (2018–2022). Figure 2 presents the spatial distribution of the ASOS stations used in this study along with their associated Thiessen polygons. Based on the computed polygons, each station covers an average area with a diameter of approximately 40.05 km, ranging from a minimum of 8.76 km to a maximum of 61.18 km. These values reflect notable spatial variability in station density. While the overall coverage is generally acceptable for regional-scale analysis, some mountainous regions appear to be underrepresented, which may affect the accuracy in localized drought detection. Nevertheless, the primary aim of this study is to provide broad insights into flash drought detection across South Korea using multiple drought indices. Accordingly, we prioritized the use of observation data with high temporal and spatial resolution to support the core analysis. The dataset served as the foundation for generating the drought indices SPI, SPEI, EDDI, and SPEDI. Recognizing the absence of directly measured ET, this study intentionally derived \(\:{\text{E}\text{T}}_{0}\) using a multiscale surrogate model based on the Hargreaves-Samani equation proposed by Kim et al. ( 2023 ). To facilitate a comprehensive evaluation of flash droughts by region, point data from ASOS datasets were interpolated into an administrative district areal dataset. This spatial approach enhances the regional assessment of drought conditions, contributing to a more comprehensive understanding of the phenomenon under investigation. 2.3 Field Crop Damage Dataset During the Summer Season Flash drought can result in a variety of damages, including agricultural losses, wildfires, and ecological impacts. The government evaluates the damage to each field and crop by surveying the affected areas, subsequently calculating the damage ratio. The agricultural crop damage ratio represents the proportion of the area that cannot be harvested (determined by the damaged area multiplied by the damage ratio) in comparison to the entire cultivated area of crops. In this study, our primary emphasis is on agricultural damage incurred in field crops rather than rice paddies, which we have identified as the key indicator for assessing flash droughts in South Korea. Field crops in South Korea primarily depend on river or groundwater irrigation, whereas rice paddies benefit from irrigation supplied through agricultural reservoirs. This distinction in irrigation sources makes rice paddies less susceptible to the direct impacts of flash droughts compared to field crops. Moreover, most damage to field crops is caused by flash droughts that occur during the summer. In these contexts, we gathered a dataset on agricultural damage to field crops from the Ministry of Agriculture, Food, and Rural Affairs, as summarized in Table 1 . The dataset includes officially reported damage during the summer period (June to August) for the years 2017, 2018, 2019, and 2022, in which field-level drought damage was documented. In some administrative districts, no drought damage was reported or recorded during these years. A summary of the dataset is provided in Table 1 . Table 1 Total areas, field crop area, and flash drought-induced field crop damage by district from June to August for the years 2017, 2018, 2019, and 2022 (ha) (N/A indicates that no investigation was conducted or data was not collected for the respective district.) District Name Total Area (ha) Field Crop Land Use (ha) Field Crop Damage (ha) 2017 2018 2019 2022 Seoul 60,500 407 N/A N/A N/A N/A Busan 77,100 2,519 286 795 0 0 Daegu 149,900 3,873 2,620 1,605 560 55 Incheon 106,700 5,154 0 0 0 6 Gwangju 50,100 4,165 N/A N/A N/A N/A Daejeon 54,000 2,656 64 4,402 2,078 118 Ulsan 106,300 4,722 0 21 0 0 Sejong-si 46,500 3,401 2,447 630 0 133 Gyeonggi-do 1,020,000 70,288 1,086 1,805 0 13 Gangwon-do 1,683,100 70,037 279 0 0 1,216 Chungcheongbuk-do 740,700 60,128 0 850 384 0 Chungcheongnam-do 824,800 73,657 0 8,649 0 0 Jeollabuk-do 807,300 67,595 2,383 2,875 0 238 Jellanam-do 1,236,200 110,747 291 1,126 0 0 Gyeongsangbuk-do 1,842,400 143,350 0 0 0 0 Gyeongsangnam-do 1,054,300 69,864 N/A N/A N/A N/A Jeju-do 185,000 55,593 0 9 0 0 [Insert Table 1 ] 3. Methodology 3.1 Thiessen Polygon Method The meteorological dataset used in this study consists of point-based observations collected from individual stations, whereas the validation data dataset is aggregated by administrative districts. To bridge this mismatch in spatial format, we applied the Thiessen polygon method to convert point observations into area-weighted averages. This method assigns weights to each observation station based on its proximity to neighboring stations. This is achieved by constructing polygons through the perpendicular bisector between stations, forming Thiessen polygons that delineate each station’s zone of influence (Fig. 2 ). Using these polygons, spatially weighted averages of meteorological variables were computed for each administrative district used in the validation dataset. The weighted value \(\:{P}_{n}\) for each district is calculated as follows: $$\:{P}_{n}=\frac{\sum\:{P}_{i}{A}_{i}}{\sum\:{A}_{i}}$$ 1 where \(\:{P}_{n}\) is the meteorological value at station \(\:i\) , and \(\:{A}_{i}\) is the area of the Thiessen polygon within the target administrative district. 3.2 Computation of Reference Evapotranspiration \(\:{\varvec{E}\varvec{T}}_{\mathbf{o}}\) In regions with limited hydrometeorological data, it is crucial to adopt surrogate equations for calculating evapotranspiration (ET) or reference evapotranspiration \(\:{\text{E}\text{T}}_{0}\) . Numerous empirical equations have been derived from extensive research to reflect the underlying physical relationships between ET and meteorological input variables. Various approaches for estimating \(\:{ET}_{0}\) have been widely applied in hydrology, including the Hargreaves method (Hargreaves and Samani, 1985 ), FAO56 Penman-Monteith method (Allen et al., 1998 ), and ASCE PM method (Allen et al., 2005 ). This study employs the temperature-based Hargreaves equation and physically-based FAO56 Penman-Monteith equation (FAO56 PM) to estimate \(\:{\text{E}\text{T}}_{0}\) : $$\:{ET}_{0\left(PM\right)}=\frac{0.408\varDelta\:\left({R}_{n}-G\right)+\gamma\:\frac{900}{\gamma\:+273}{u}_{2}({e}_{s}-{e}_{a})}{\varDelta\:+\gamma\:(1+0.34{u}_{2})}$$ 2 $$\:{ET}_{0\left(HS\right)}=0.0023\bullet\:RA\times\:(T℃+17.8)\times\:{({T}_{max}-{T}_{min})}^{0.50}$$ 3 Equation ( 1 ) represents the FAO56 PM equation, where \(\:{R}_{n}\) is the net radiation ( \(\:MJ/{m}^{2}/day\) ), \(\:G\) is the soil heat flux density ( \(\:MJ/{m}^{2}/day\) ), \(\:\text{T}\) is the average temperature at a 2m height ( \(\:℃\) ), \(\:{u}_{2}\) is the wind speed at a 2m height ( \(\:m/s\) ), \(\:{e}_{s}\) is the saturation vapor pressure ( \(\:kPa\) ), \(\:{e}_{a}\) is the actual vapor pressure ( \(\:kPa\) ), \(\:{e}_{s}-{e}_{a}\) is the saturation vapor pressure deficit ( \(\:kPa)\) , and \(\:\varDelta\:\) is the slope vapor pressure curve ( \(\:kPa/℃\) ) (Allen et al., 1998 ). Eq. ( 2 ) represents the HS equation, where \(\:RA\) is the extraterrestrial radiation in mm/day ; \(\:\text{T}\) is the average temperature; and \(\:{\text{T}}_{\text{m}\text{a}\text{x}}\) and \(\:{\text{T}}_{\text{m}\text{i}\text{n}}\) are the maximum and minimum temperature in °C, respectively (Hargreaves and Samani, 1985 ). The HS equation, initially formulated with fixed regression parameters, has the potential to produce biased estimates when applied in different climatic regimes (Valero et al., 2013 ). In response to this concern, Samani ( 2000 ) recommended the calibration of coefficients to better represent local climate conditions. Several publications have suggested improved parameters for the HS equation in the different climate zones (Gavilán et al., 2006 ; Martınez-Cob and Tejero-Juste, 2004 ; Raziei and Pereira, 2013 ; Yang et al., 2021 ). The general form of the HS equation can be represented by Eq. ( 3 ). $$\:{ET}_{0}=\alpha\:\bullet\:RA\times\:(T+\beta\:)\times\:{({T}_{max}-{T}_{min})}^{\gamma\:}$$ 4 In a recent study conducted by Kim et al. ( 2023 ), a Bayesian approach was developed within a multi-scale modeling framework to estimate parameters for the HS equation, maintaining accurate estimations of evapotranspiration across various temporal scales. More specifically, the proposed surrogate model offers a new perspective for extending the existing HS equation to encompass daily, monthly, and annual scales. The three parameters ( \(\:\alpha\:,\:\beta\:,\:\gamma\:\) ) of the multi-scale surrogate model were recalibrated and validated using \(\:{\text{E}\text{T}}_{0}\) derived from the FAO56 PM method. In a previous study, the performance of the multi-scale surrogate model was assessed by categorizing it based on different temporal scales (i.e., daily, monthly, annual). The findings demonstrated that the multi-scale model outperformed a single-scale model in minimizing systematic bias across various temporal scales. In our current study, we build upon the findings from that previous research (Kim et al., 2023 ), and the calibrated parameters were used for estimating \(\:{ET}_{0}\) , which can be used to estimate EDDI. The parameters used for this study are detailed in Appendix A. 3.3 Computation of Drought Indices In this study, we utilize three widely accepted drought indices—SPI, SPEI, and EDDI—with a one-month accumulation period and daily temporal resolution to effectively capture flash drought conditions, which are characterized by rapid onset and intensification within a few weeks. Computation of SPI The SPI computation process involves fitting the accumulated precipitation data to a probability distribution, typically the gamma distribution, computing cumulative probability for the corresponding precipitation deficits, and then mapping these probabilities into standardized scores using the inverse cumulative standard normal distribution (McKee et al., 1993 ). First, precipitation anomalies are quantified by fitting accumulated precipitation for a given time period (e.g., 1, 3, or 12 months) to a gamma distribution. The gamma probability density function is defined as: $$\:f\left(x\right)=\:\frac{1}{{\Gamma\:}\left(\alpha\:\right){\beta\:}^{\alpha\:}}{x}^{\alpha\:-1}\text{e}\text{x}\text{p}\left(-\frac{x}{\beta\:}\right)$$ 5 where \(\:x\) is the accumulated precipitation amount, \(\:\alpha\:\) is the shape parameter, \(\:\beta\:\) is the scale parameter, and \(\:{\Gamma\:}\left(\alpha\:\right)\) is the gamma function. These parameters \(\:\alpha\:\) and \(\:\beta\:\) are estimated using the maximum likelihood method. Because precipitation time series may include zero values, the cumulative distribution function (CDF) \(\:G\left(x\right)\) is adjusted to account for the probability of zero precipitation. The adjusted cumulative probability, \(\:H\left(x\right)\) , is computed as: $$\:H\left(x\right)=q+\left(1-q\right)G\left(x\right)$$ 6 where \(\:q\) is the probability of precipitation events in the time series., Finally, PI is obtained by transforming \(\:H\left(x\right)\) to a standard normal distribution with a mean value of 0 and a standard deviation of 1. Positive SPI values indicate the wet-than-average conditions, while negative values indicate drier-than-average conditions. Computation of SPEI The SPEI extends the SPI framework by incorporating precipitation and reference evapotranspiration ( \(\:{ET}_{0}\) ) to account for atmospheric moisture demand (Vicente-Serrano et al., 2010 ). It uses the climatic water balance, defined as \(\:D=P-{ET}_{0}\) , ​where \(\:P\) is precipitation and \(\:{ET}_{0}\) is reference evapotranspiration, as its input variable. This balance is accumulated over a specified time scale to capture short- to medium-term drought conditions. Unlike SPI, which typically uses the gamma distribution, SPEI is fitted to a log-logistic distribution, which can accommodate both positive and negative values. The log-logistic CDF is given by: $$\:f\left(D\right)=\frac{\beta\:}{\alpha\:}\times\:\frac{{\left(D/\alpha\:\right)}^{\beta\:-1}}{\alpha\:{\left[1+{\left(D/\alpha\:\right)}^{\beta\:}\right]}^{2}}$$ 7 where \(\:\alpha\:\) , \(\:\beta\:\) , and \(\:\gamma\:\) represent the scale, shape, and location parameters, respectively. The resulting probability is then transformed into a standard normal variate, consistent with the SPI procedure. Computation of EDDI EDDI, originally proposed by Hobbins et al. ( 2016 ), uses accumulated \(\:{ET}_{0}\) as the input to represent atmospheric moisture demand. EDDI reflects how strongly the atmosphere is “pulling” moisture from the surface, and is particularly sensitive to conditions such as heatwaves, persistent clear skies, and strong winds. While the original method relies on a non-parametric empirical approach using the Tukey plotting position formula, we adopted a parametric approach in this study for consistency with other indices. Specifically, we applied the gamma distribution to fit the accumulated \(\:{ET}_{0}\) data over a one-month time scale. This choice was based on Kolmogorov-Smirnov (K-S) goodness-of-fit tests conducted across multiple locations in South Korea, which confirmed the gamma distribution as a suitable model for \(\:{ET}_{0}\) ​. The computation procedure follows the general SPI framework, substituting precipitation with accumulated \(\:{ET}_{0}\) . The resulting values are then transformed to standard normal variates to produce standardized EDDI scores. Unlike SPI and SPEI, where negative values indicate drier conditions, EDDI conventionally defines positive values as indicative of dryness. To maintain consistency across all indices used in this study, we inverted the sign of the EDDI values so that negative values also represent dry conditions, aligning the interpretation of all indices. 3.4 Development of Composite Drought Index SPEDI Accurate assessment of flash drought requires accounting for both precipitation deficit and increased atmospheric evaporative demand. While SPI and EDDI each represent one of these components, their single-variable nature limits their ability to fully characterize flash drought conditions. SPEI was developed to integrate both precipitation and evapotranspiration; however, previous studies have shown that it tends to correlate more closely with SPI than with EDDI, suggesting it may not adequately capture the independent effects of evaporative demand (Yao et al., 2018 ; Won et al., 2020 ). Given these limitations, we propose a novel composite drought index, SPEDI (Standardized Precipitation Evaporation Differential Index), designed to integrate the distinct signals from SPI and EDDI. The index is intended to provide a more balanced representation of drought severity by combining independently derived anomalies from precipitation and atmospheric dryness. We define an intermediate variable Total Water Deficit ( \(\:TWD\) ) as the sum of SPI and sign-reversed EDDI: $$\:TWD=SPI+EDDI$$ 8 Note that in this study, we reversed the sign of EDDI values so that negative values indicate dry conditions, aligning with the interpretation of SPI and SPEI. If the original (non-reversed) EDDI values are used, \(\:TWD\) should instead be computed as: $$\:TWD=SPI-EDDI$$ 9 Higher value of \(\:TWD\) represent conditions where both precipitation is below normal and atmospheric moisture demand \(\:{ET}_{0}\) is elevated, hallmarks of flash drought. To standardize \(\:TWD\) and allow direct comparison with other drought indices, we apply Z-score normalization: $$\:SPEDI=\frac{TWD-\mu\:}{\sigma\:}$$ 10 where \(\:\mu\:\) and \(\:\sigma\:\) are the mean and standard deviation of TWD over the reference period. By explicitly combining two independently derived indicators, SPEDI retains the full signal of both variables and offers a more integrated and sensitive measure of flash drought conditions. 3.5 Flash Drought Identification (criteria) The criteria for identifying dry conditions are outlined in Table 2 , following the percentile-based classification system proposed by Svoboda et al. ( 2002 ). In this framework, drought severity is classified based on the percentile ranking of drought index values: severe drought corresponds to values below the 10th percentile, moderate drought to the 10th -20th percentile range, and mild drought to the 20th -40th percentile range. Values above the 40th percentile are considered normal. This percentile-based approach enables consistent detection of drought intensity across different regions and time periods. Table 2 Criteria for identifying dry conditions using drought indices (SPI, SPEI, EDDI, and SPEDI) based on the percentile ranking system by Svoboda et al. ( 2002 ) Percentile of Drought Indices Drought conditions Less than 10% Severe drought 10 ~ 20% Moderate drought 20 ~ 40% Mild drought Greater than 40% Normal Flash droughts, in contrast to seasonal or long-term droughts, are defined by two distinguishing features: (2) rapid onset, typically developing within 2 to 6 weeks. (2) rapid intensification, often resulting in substantial impacts before conventional monitoring systems can detect them. These characteristics distinguish flash drought from traditional drought events that develop gradually over months or seasons. Building on these characteristics, numerous studies have proposed criteria for flash drought identification (Ford and Labosier, 2017 ; Pendergrass et al., 2020 ; Parker et al., 2021 ). In this study, we developed a region-specific definition tailored to South Koreans. A flash drought event is defined as a rapid decline in drought index values, where the percentile shifts from above the 50 th percentile to below the 20 th percentile within a 14-day span. This indicates a rapid shift from normal to moderate drought within two weeks. This threshold captures a transition from normal to at least moderate drought within two weeks. This criterion provides a quantifiable and operational definition of flash drought onset, facilitating both early detection and more responsive drought management in rapidly evolving scenarios. [Insert Table 2 ] 4. Results 4.1 Spatial Distribution of Flash Drought In this study, it is assumed that the extent of agricultural damage areas to field crops can serve as an indicator of flash drought severity, enabling their use in assessing flash drought indices. The distribution of agricultural damage areas is presented visually for four specific years (2017, 2018, 2019, and 2022) in each district, as illustrated in Fig. 2 . Metrics related to severity and duration were aggregated for the flash drought events that occurred between May and August, as presented in Figs. 3 and 4 . Here, the flash drought events are defined as described in Section 3.2 . The flash drought assessment data used in this study rely exclusively on agricultural damage to field crops, which only accounts for agricultural areas associated with field crops. Therefore, this study does not account for other types of damage that may result from flash drought events, primarily due to limitations associated with the available dataset. In 2017, SPI and SPEI exhibited limitations in accurately representing the extent of damage caused by flash drought. In contrast, both EDDI and SPEDI effectively identified and delineated the drought-affected areas. Indeed, SPEDI offered a notably more comprehensive and extensive representation compared to other drought indices. In 2018, a concentration of drought was evident on the western side of the continent, with all drought indices indicating flash droughts throughout the country. The Chungcheongnam-do district, in particular, emerged as a focal point due to its substantial agricultural damage during the validation period. Significantly, SPEDI exhibited a similar spatial trend, highlighting its effectiveness in accurately representing the observed patterns of agricultural damage. In 2019, a year marked by limited agricultural damage, both EDDI and SPEDI revealed a distribution of flash droughts across the entire country. In contrast, both SPI and SPEI suggested that these events were concentrated in the northern regions, failing to capture the drought event in the Daegu district. This emphasizes the capability of EDDI and SPEDI to offer a more comprehensive perspective of flash drought occurrences, particularly in regions that may not be identified by other indices. In 2022, the Gangwon-do district experienced significant agricultural damage due to drought. The SPI and EDDI indices effectively identified these drought events. Conversely, the SPEI and SPEDI indices failed to detect drought events in this district during the same period. This discrepancy emphasizes the varied sensitivities of these indices to the regional patterns of drought, indicating the importance of drought indices for specific drought characteristics and locations. [Insert Figs. 3 –5] 4.2 Temporal Variation of Drought Indices This chapter’s primary objective is to assess the effectiveness of drought indices in detecting and characterizing flash drought events. The criteria for identifying these events, as previously outlined, were applied to recognize these events. Subsequently, validation was carried out using the dataset listed in Table 1 . Figure 6 presents the temporal evolution of drought indices—SPI, SPEI, EDDI, and SPEDI—from 2017 to 2022 and highlights the instances of flash drought, as determined through agricultural damage data. [Insert Fig. 6 ] During periods marked by significant flash drought occurrences, such as in 2018, the four drought indices exhibited consistent results, except in the Jeju-do district, as illustrated in Fig. 5(g). The drought event in Gyeonggi-do in 2017 was also detected effectively by all four indices. SPI, SPEI, and SPEDI proactively detected the drought, while EDDI precisely aligned with the validation period, as shown in Fig. 6 (a). Some drought events, particularly those categorized as “precipitation-deficit flash drought”, were not detected by EDDI. For example, the 2018 event in Jeju-do district (Fig. 6 (g)) and the 2019 event in Chungcheongbuk-do district (Fig. 6 (d)) were captured by SPI, SPEI, and SPEDI but missed by EDDI. This highlights the importance of precipitation-based indices in identifying droughts driven primarily by rainfall deficiency. On the other hand, there were cases where SPEI failed to detect drought, while other indices responded effectively. A notable example is the 2022 drought in Gangwon-do (Fig. 6 (b)), which was identified by SPI, EDDI, and SPEDI, but not by SPEI. This inconsistency further supports findings that SPEI may underrepresent evaporative demand under certain conditions. Events classified as “heatwave flash drought” were also observed. For the 2022 drought in the Gyeonggi-do district, only EDDI and SPEDI detected the event in advance, as illustrated in Fig. 6 (a). In contrast, SPI and SPEI indices did not indicate a drought, suggesting that precipitation levels alone were not abnormally low. In this case, elevated temperatures and resulting high evaporative demand elevated temperatures and resulting high evaporative demand played a key role in the development of drought. This reinforces the importance of incorporating both precipitation deficits and atmospheric evaporative demand when identifying flash droughts. Last, in some instances, drought events were identified by a single drought index. Specifically, the 2017 drought in the Jeollanam-do district was detected by the SPEI, as shown in Fig. 6 (e). Similarly, the drought in the Gangwon-do district in the same year was only identified during the period highlighted by the EDDI, as displayed in Fig. 6 (b). Again, this suggests that each index has strengths and limitations in capturing specific drought patterns and events. Our comparative analysis of drought indices in evaluating flash drought events revealed distinct patterns. EDDI consistently detected a higher number of flash drought events compared to the other drought indices. The SPI and SPEI exhibited similar tendencies, often underestimating the duration of drought events in the historical period. SPEDI effectively addressed the issues of overestimation and underestimation that were evident in other indices. Throughout the study period, across the 14 flash drought events identified in South Korea, the SPEDI index demonstrated better performance in accurately detecting the periods of agricultural damage (see Table 3 ). Table 3 Comparative performance of drought indices (EDDI, SPI, SPEI, SPEDI) in detecting flash drought events based on agricultural damage periods in South Korea SPI SPEI EDDI SPEDI Total Early Detection 10 11 7 12 14 Matching Period 7 8 9 10 14 [Insert Table 3 ] 4.3 Assessment of SPEDI Performance Compared to SPEI In our study, we categorized the complex interactions between precipitation and evaporation into a simplified model, outlining four distinct states that characterize the water balance dynamics between the atmosphere and the land surface, as summarized in Table 4 . Table 4 Simplified model of water balance dynamics: categorization of atmospheric and land surface interactions into four distinct states State Atmosphere Condition Land Condition Evaporative Demand Actual Evaporative Drought Risk 1 Wet Wet Low Low Very low 2 Wet Dry Low Low Low 3 Dry Wet High High Moderate 4 Dry Dry High Low High State 1 is defined by very humid atmospheric conditions, where there is no additional evaporation as the atmosphere is saturated, even if the land surface is moist. State 2 also represents a humid atmosphere but with dry land conditions. Here, both the demand for evaporation and the actual evaporation rate are low because the dry land surface tends to absorb moisture from the air rather than release it. In contrast, State 3 marks the onset of a precipitation deficit characterized by a dry atmosphere and a still-wet land surface. Under these conditions, the atmosphere exerts a higher evaporative demand, and the land surface, having adequate moisture availability, responds with increased evaporation. Although this state may be considered the onset of a drought, it is described more accurately as a transitional (or normal condition), pending further development. Finally, State 4 emerges if the precipitation deficit continues, marking a progression from the conditions described in State 3. State 4 can be classified as a severe drought, marked by a dry atmosphere in need of moisture recovery but with a dry land surface unable to provide it. This condition sustains a high level of evaporative demand that exceeds the land’s capacity to supply moisture, leading to a failure in atmospheric recovery and an ongoing pattern of moisture demand. Consequently, the atmosphere continues to demand greater evaporation, resulting in a progressively drier land surface until soil moisture is depleted. Unlike State 3, where actual evaporation may be significant, in State 4, evaporative demand continues to increase while actual evaporation decreases due to the deficiency of available moisture. This state of severe drought can rapidly intensify under abnormal weather conditions, further exacerbating the moisture imbalance between the atmosphere and the land surface. [Insert Table 4 ] Theoretically, States 2 and 3 can be regarded as neutral in the context of drought analysis as precursors to the onset of drought, which is clearly marked by State 4. Drought indices that account for both precipitation and evaporation, particularly for evaporative demand, such as SPEI, are anticipated to classify these states effectively. However, some studies have pointed out that SPEI may be insufficient in a full consideration of the complexities of evaporation dynamics within this framework (Yao et al., 2018 ; Won et al., 2020 ), leading to higher correlation with SPI compared to EDDI (Xiao et al., 2024 ). To evaluate the relationships among drought indices, we aggregated index values from all weather stations across South Korea into single time series vectors for each index. Pearson correlation coefficients were then computed between these vectors to assess the linear association among SPI, SPEI, EDDI, and SPEDI on a national scale, as shown in Fig. 7 . Notably, SPI and SPEI demonstrate a strong correlation, whereas EDDI exhibits a lower correlation with SPI and SPEI. Interestingly, SPEDI exhibits relatively high correlations with all drought indices. These suggest that SPEI may not fully account for evaporative trends, leading to its lower correlation with EDDI. Additionally, the high correlations near one between SPEI and SPI indicate that SPEI may not effectively complement the limitation of SPI. [Insert Fig. 7 ] Considering the matter of insufficient consideration of evaporation in SPEI, we conducted a comparative analysis of the two drought indices, SPEI and SPEDI. While the two indices attempt to render the same information on the land-atmosphere water balance, they employ different computation methods. In the computation of SPEI, the difference between precipitation and potential evapotranspiration ( \(\:{ET}_{o}\) ) is calculated and then fitted to a probability distribution. Conversely, SPEDI computes the precipitation to potential evapotranspiration difference after individually fitting each component into its respective probability distributions. In fitting each factor to the probability distributions, we hypothesized that there would be a difference in the contributions of these factors to the resulting drought index. To validate this hypothesis, we conducted an experimental comparison using annual time series data in Gangwon-do. This region was selected due to its unique precipitation and evaporation patterns that give rise to a clear distinction between the states identified by SPI and EDDI. More importantly, Gangwon-do is a key region for field crop production in South Korea, particularly well-suited for evaluating flash droughts. The frequency and temporal distribution of each state in the Gangwon-do region from 1980 to 2022 are shown in Figs. 8 and 9 . [Insert Figs. 8 – 9 ] To assess the performance of SPEI and SPEDI across the categorized states, Fig. 10 displays the time series of drought indices for selected states. State 1 was excluded from the Figure due to the absence of drought. In the majority of seasons categorized as States 2 and 3, SPEDI consistently indicated neutral conditions, while the SPEI tracked more closely with the SPI. For instance, in March and April of 2018, which were classified as State 3, SPEI did not account adequately for the increased evaporation effects characteristic of this period. On the other hand, SPEDI accurately reflected these neutral conditions. Furthermore, during periods identified as State 4, SPEI exhibited a reduction in the severity of drought conditions. In contrast, SPEDI more effectively captured instances of extreme drought, offering a more comprehensive representation that considered both precipitation and evaporation conditions. This comparative analysis of SPEI and SPEDI allows us to better understand the distinct contributions of each component to these drought indices. The findings emphasize that SPEI does not fully describe the influence of evaporation. In contrast, SPEDI offers a more comprehensive representation, effectively accounting for the interaction between precipitation and evaporation in the assessment of drought conditions. [Insert Fig. 10 ] 5. Discussion This study demonstrates that traditional precipitation-based drought indices, such as SPI and SPEI, are limited in their ability to capture the rapid onset of flash droughts (Zhang et al., 2017 ; Parker et al., 2021 ). While EDDI accounts for atmospheric moisture demand via reference evapotranspiration ( \(\:{ET}_{0}\) ​), it has been shown to generate false positives by ignoring the precipitation component (Hobbins et al., 2016 ; Parker et al., 2021 ). These single-variable indices often fail to represent the complex interaction between elevated atmospheric demand and concurrent precipitation deficits that drives flash drought development (Hoell et al., 2020 ). Although SPEI attempts to integrate both precipitation and evaporative demand​ ( \(\:{ET}_{0}\) ), previous studies have found that it tends to underestimate drought severity under conditions of high evaporative demand (Yuan et al., 2019 , Hoffmann et al., 2021 ). To address this limitation, we proposed a new composite index, SPEDI, that preserves the distinct statistical properties of precipitation and \(\:{ET}_{0}\) , ​by standardizing them separately before integration. This formulation enables SPEDI to better capture the transition from normal to extreme drought conditions, particularly under rapidly intensifying scenarios. Our findings show that SPEDI more effectively delineates flash drought-affected areas, as evidenced by its strong agreement with reported agricultural damage data. Despite these promising results, several limitations of this study should be acknowledged. First, the validation of flash drought events relied exclusively on agricultural damage records, which were limited to the summer season and to specific years. This restricts the ability to analyze seasonal variation and long-term trends. Second, the study focused solely on agricultural impacts and did not consider flash drought effects on ecosystems, water resources, or urban areas. As a result, our findings and the performance of SPEDI are most applicable to agricultural drought, and caution is advised when generalizing to other sectors without further validation. Another limitation concerns the spatial resolution of the input meteorological data and drought indices. While the current resolution is suitable for regional-scale assessment, it may not adequately capture localized drought events or microclimatic effects, especially in mountainous or coastal regions. Using finer-resolution datasets could help detect localized flash droughts more accurately and support more targeted management. Looking ahead, several directions for future research can strengthen and expand upon this study. First, improving the validation dataset is essential. Future work should incorporate drought impact data from other sectors, such as ecosystems, water resources, and urban areas, and extend the analysis to different seasons. This would allow a more complete evaluation of index performance across a broader range of conditions. It would also be valuable to compare results across various regions and climate zones to assess how the index performs in different settings. Additionally, using higher-resolution meteorological data could improve the precision of flash drought detection, especially at the local scale. This would support the identification of smaller or localized events and help track drought development in areas with complex terrain, including mountainous and coastal regions. Finally, given the increasing recognition of flash drought as a significant hazard, efforts should continue to bridge research with real-time monitoring. Integrating indices like SPEDI into operational early warning systems, validated against observed impacts, has the potential to significantly enhance drought preparedness and response (Otkin et al., 2018 ; Otkin et al., 2022 ) 6. Conclusion In this study, we aimed to enhance our understanding of flash drought patterns by conducting a comprehensive analysis of four drought indices: the Standardized Precipitation Index (SPI), the Standardized Precipitation Evapotranspiration Index (SPEI), the Evaporative Demand Drought Index (EDDI), and a Standardized Precipitation Evaporation Differential Index (SPEDI). Specifically, this study investigated the occurrence of flash droughts in South Korea, with an emphasis on spatiotemporal patterns, assessing the effectiveness of the mentioned drought indices, and identifying the meteorological factors that influence their development. Our systematic analysis within the South Korean peninsula led to the following conclusions. This study emphasized the limitations of traditional drought indices such as SPI and SPEI in capturing the full extent of flash drought impacts, particularly in contrast to the EDDI and SPEDI indices, which demonstrated capability in delineating drought-affected areas and accurately representing the observed patterns of agricultural damage. The temporal analysis further revealed the strengths and weaknesses of each drought index in detecting and characterizing flash drought events, highlighting the importance of both precipitation deficits and temperature-induced evaporative stress. It was evident that EDDI and SPEDI indices were more effective in identifying a broader range of drought events, including those driven by high evaporative demand, and offered a more comprehensive view of flash drought occurrences. Notably, the assessment of SPEDI performance in comparison to SPEI illustrated the complex interactions between precipitation and evaporation in the context of drought analysis. The study revealed that SPEI may not fully account for the dynamics of evaporation, leading to a potential underestimation of drought severity under certain conditions. In contrast, SPEDI, by incorporating both precipitation and evaporation components, provided a detailed representation of drought conditions, effectively capturing the transition from normal to severe drought states. In conclusion, the findings of this study highlight the need to employ novel drought indices like SPEDI, which capture both precipitation deficits and evaporative demand. The results demonstrate that SPEDI offers a more balanced and responsive approach to detecting flash droughts compared to traditional indices. These insights gained here can contribute to the development of more effective flash drought monitoring and management strategies. Furthermore, SPEDI shows promise for integration into early warning systems, helping improve the timely detection and response to rapidly developing drought conditions. Declarations Author Contribution S. Kang: conceptualization, data curation, methodology, software, validation, formal analysis, writing - original draft preparation. H.J. Kim: methodology, validation, formal analysis, writing - original draft preparation. J.H. Lee: writing - review & editing, funding acquisition. H.-C. Yoon: conceptualization, writing - review & editing. H.-H. Kwon: conceptualization, methodology, software, validation, investigation, supervision, writing - review & editing, funding acquisition. ACKNOWLEDGEMENT This research was supported by a grant (2022-MOIS63-001(RS-2022-ND641011)) of the Cooperative Research Method and Safety Management Technology in National Disaster funded by the Ministry of the Interior and Safety (MOIS, Korea). References Abu Arra, A. and E. Şişman (2024). "Innovative drought classification matrix and acceptable time period for temporal drought evaluation." Water Resources Management, VoL. 38, No. 8: (2811-2833) Allen, R., et al. (2005). "RL Snyder, The ASCE Standardized Reference Evapotranspiration Equation." Environmental and Water Resources Institute of the American Society of Civil: Reston, FL, USA, Allen, R. G., et al. (1998). "Crop evapotranspiration-Guidelines for computing crop water requirements-FAO Irrigation and drainage paper 56." Fao, Rome, VoL. 300, No. 9: (D05109) Anderson, M. C., et al. (2013). "An intercomparison of drought indicators based on thermal remote sensing and NLDAS-2 simulations with US Drought Monitor classifications." Journal of Hydrometeorology, VoL. 14, No. 4: (1035-1056) Arra, A. A., et al. (2024). "Trends of meteorological and hydrological droughts and associated parameters using innovative approaches." Journal of Hydrology, VoL. 640: (131661) Christian, J. I., et al. (2019). "A methodology for flash drought identification: Application of flash drought frequency across the United States." Journal of Hydrometeorology, VoL. 20, No. 5: (833-846) Ford, T. W. and C. F. Labosier (2017). "Meteorological conditions associated with the onset of flash drought in the eastern United States." Agricultural and Forest Meteorology, VoL. 247: (414-423) Ford, T. W., et al. (2015). "On the utility of in situ soil moisture observations for flash drought early warning in Oklahoma, USA." Geophysical research letters, VoL. 42, No. 22: (9790-9798) Gavilán, P., et al. (2006). "Regional calibration of Hargreaves equation for estimating reference ET in a semiarid environment." Agricultural Water Management, VoL. 81, No. 3: (257-281) Hargreaves, G. H. and Z. A. Samani (1985). "Reference crop evapotranspiration from temperature." Applied engineering in agriculture, VoL. 1, No. 2: (96-99) Hobbins, M. T., et al. (2016). "The evaporative demand drought index. Part I: Linking drought evolution to variations in evaporative demand." Journal of Hydrometeorology, VoL. 17, No. 6: (1745-1761) Hoell, A., et al. (2020). "Lessons learned from the 2017 flash drought across the US Northern Great Plains and Canadian Prairies." Bulletin of the American Meteorological Society, VoL. 101, No. 12: (E2171-E2185) Hoffmann, D., et al. (2021). "Flash drought in CMIP5 models." Journal of Hydrometeorology, VoL. 22, No. 6: (1439-1454) Hu, J., et al. (2023). "The effects of high temperature, drought, and their combined stresses on the photosynthesis and senescence of summer maize." Agricultural Water Management, VoL. 289: (108525) Hunt, E. D., et al. (2009). "The development and evaluation of a soil moisture index." International Journal of Climatology: A Journal of the Royal Meteorological Society, VoL. 29, No. 5: (747-759) Hunt, E. D., et al. (2014). "Monitoring the effects of rapid onset of drought on non-irrigated maize with agronomic data and climate-based drought indices." Agricultural and Forest Meteorology, VoL. 191: (1-11) Kim, C. J., et al. (2014). "Analysis of climate change impacts on the spatial and frequency patterns of drought using a potential drought hazard mapping approach." International Journal of Climatology, VoL. 34, No. 1: (61-80) Kim, H.-J., et al. (2023). "A novel multi-scale parameter estimation approach to the Hargreaves-Samani equation for estimation of Penman-Monteith reference evapotranspiration." Agricultural Water Management, VoL. 275: (108038) KMA (2017). 2016 Abnormal climate report, Korea Meteorological Administration. Kwon, H. H., et al. (2016). "The unusual 2013–2015 drought in South Korea in the context of a multicentury precipitation record: Inferences from a nonstationary, multivariate, Bayesian copula model." Geophysical research letters, VoL. 43, No. 16: (8534-8544) Kwon, M., et al. (2019). "Spatio‐temporal drought patterns of multiple drought indices based on precipitation and soil moisture: A case study in South Korea." International Journal of Climatology, VoL. 39, No. 12: (4669-4687) Lee, H.-J., et al. (2021). "Detection of flash drought using evaporative stress index in South Korea." Journal of Korea Water Resources Association, VoL. 54, No. 8: (577-587) Lisonbee, J., et al. (2022). "Making sense of flash drought: Definitions, indicators, and where we go from here." Martınez-Cob, A. and M. Tejero-Juste (2004). "A wind-based qualitative calibration of the Hargreaves ET0 estimation equation in semiarid regions." Agricultural Water Management, VoL. 64, No. 3: (251-264) McEvoy, D. J., et al. (2012). "An evaluation of multiscalar drought indices in Nevada and Eastern California." Earth Interactions, VoL. 16, No. 18: (1-18) McKee, T. B., et al. (1993). The relationship of drought frequency and duration to time scales . Proceedings of the 8th Conference on Applied Climatology, California. Mo, K. C. and D. P. Lettenmaier (2016). "Precipitation deficit flash droughts over the United States." Journal of Hydrometeorology, VoL. 17, No. 4: (1169-1184) Mozny, M., et al. (2012). "Use of a soil moisture network for drought monitoring in the Czech Republic." Theoretical and Applied Climatology, VoL. 107: (99-111) Naveendrakumar, G., et al. (2019). "South Asian perspective on temperature and rainfall extremes: A review." Atmospheric Research, VoL. 225: (110-120) Noguera, I., et al. (2020). "Characteristics and trends of flash droughts in Spain, 1961–2018." Annals of the New York Academy of Sciences, VoL. 1472, No. 1: (155-172) Otkin, J. A., et al. (2013). "Examining rapid onset drought development using the thermal infrared–based evaporative stress index." Journal of Hydrometeorology, VoL. 14, No. 4: (1057-1074) Otkin, J. A., et al. (2018). "Flash droughts: A review and assessment of the challenges imposed by rapid-onset droughts in the United States." Bulletin of the American Meteorological Society, VoL. 99, No. 5: (911-919) Otkin, J. A., et al. (2022). "Getting ahead of flash drought: from early warning to early action." Bulletin of the American Meteorological Society, VoL. 103, No. 10: (E2188-E2202) Parker, T., et al. (2021). "Flash drought in Australia and its relationship to evaporative demand." Environmental Research Letters, VoL. 16, No. 6: (064033) Pendergrass, A. G., et al. (2020). "Flash droughts present a new challenge for subseasonal-to-seasonal prediction." Nature Climate Change, VoL. 10, No. 3: (191-199) Qin, N., et al. (2023). "Assessing the drought impact on sugarcane yield based on crop water requirements and standardized precipitation evapotranspiration index." Agricultural Water Management, VoL. 275: (108037) Raziei, T. and L. S. Pereira (2013). "Estimation of ETo with Hargreaves–Samani and FAO-PM temperature methods for a wide range of climates in Iran." Agricultural Water Management, VoL. 121: (1-18) Samani, Z. (2000). "Estimating solar radiation and evapotranspiration using minimum climatological data." Journal of irrigation and drainage engineering, VoL. 126, No. 4: (265-267) Svoboda, M., et al. (2002). "The drought monitor." Bulletin of the American Meteorological Society, VoL. 83, No. 8: (1181-1190) Valero, J. F. M., et al. (2013). "Regionalization of the Hargreaves coefficient to estimate long-term reference evapotranspiration series in SE Spain." Spanish Journal of Agricultural Research, VoL. 11, No. 4: (1137-1152) Vicente-Serrano, S. M., et al. (2010). "A multiscalar drought index sensitive to global warming: the standardized precipitation evapotranspiration index." Journal of climate, VoL. 23, No. 7: (1696-1718) Wang, L. and X. Yuan (2018). "Two types of flash drought and their connections with seasonal drought." Advances in Atmospheric Sciences, VoL. 35, No. 12: (1478-1490) Wilhite, D. A. and M. H. Glantz (1985). "Understanding: the drought phenomenon: the role of definitions." Water international, VoL. 10, No. 3: (111-120) Won, J., et al. (2020). "Copula-based Joint Drought Index using SPI and EDDI and its application to climate change." Science of the Total Environment, VoL. 744: (140701) Xiao, X., et al. (2024). "Leveraging multisource data for accurate agricultural drought monitoring: A hybrid deep learning model." Agricultural Water Management, VoL. 293: (108692) Yang, Y., et al. (2021). "Evaluation of 18 models for calculating potential evapotranspiration in different climatic zones of China." Agricultural Water Management, VoL. 244: (106545) Yao, N., et al. (2018). "Drought evolution, severity and trends in mainland China over 1961–2013." Science of the Total Environment, VoL. 616: (73-89) Yuan, X., et al. (2019). "Anthropogenic shift towards higher risk of flash drought over China." Nature communications, VoL. 10, No. 1: (4661) Zhang, Y., et al. (2022). "Changes and driving factors of compound agricultural droughts and hot events in eastern China." Agricultural Water Management, VoL. 263: (107485) Zhang, Y., et al. (2022). "A new multi-variable integrated framework for identifying flash drought in the Loess Plateau and Qinling Mountains regions of China." Agricultural Water Management, VoL. 265: (107544) Zhang, Y., et al. (2017). "Flash droughts in a typical humid and subtropical basin: A case study in the Gan River Basin, China." Journal of Hydrology, VoL. 551: (162-176) Additional Declarations No competing interests reported. Supplementary Files AppendixA.docx Cite Share Download PDF Status: Published Journal Publication published 09 Sep, 2025 Read the published version in Stochastic Environmental Research and Risk Assessment → Version 1 posted Editorial decision: Revision requested 27 Jul, 2025 Submission checks completed at journal 21 Apr, 2025 First submitted to journal 19 Apr, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4405968","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":445986141,"identity":"195edb10-184c-4050-9792-32a2087f5076","order_by":0,"name":"Subin Kang","email":"","orcid":"","institution":"Sejong University","correspondingAuthor":false,"prefix":"","firstName":"Subin","middleName":"","lastName":"Kang","suffix":""},{"id":445986143,"identity":"84b2f713-2bb5-4c5c-84bb-502d5563a40c","order_by":1,"name":"Ho-Jun Kim","email":"","orcid":"","institution":"Sejong University","correspondingAuthor":false,"prefix":"","firstName":"Ho-Jun","middleName":"","lastName":"Kim","suffix":""},{"id":445986145,"identity":"0e7581a0-a059-4c51-bab9-212e21ca2bfa","order_by":2,"name":"Joo-Heon Lee","email":"","orcid":"","institution":"Joongbu University","correspondingAuthor":false,"prefix":"","firstName":"Joo-Heon","middleName":"","lastName":"Lee","suffix":""},{"id":445986146,"identity":"21ae40cf-9156-4b45-8eff-0218a105ce40","order_by":3,"name":"Hyeon-Cheol Yoon","email":"","orcid":"","institution":"National Disaster Management Research Institute","correspondingAuthor":false,"prefix":"","firstName":"Hyeon-Cheol","middleName":"","lastName":"Yoon","suffix":""},{"id":445986147,"identity":"f50f70aa-512b-44a4-9a10-79ddcface669","order_by":4,"name":"Hyun-Han Kwon","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA1ElEQVRIiWNgGAWjYBACAx4QWcHAwNgAFZEgTssZkrUwtiGJENRiznPG8HHhvDo55vbewy8YauwYJGcfwK/FsrfH2HjmtsPGjD3n0iwYjiUzSPMlEHDYed5t0rzbDiQ2zsgxM2BgO8Agx0PAYRAtc+rqIVr+EaPlbC9QSwNzAuOMHOMHjG0HGKQJajlz/rMxz7HDho09Z8wYEvuSeSR7CGpJS3zMU1Mnb9jeY/zhwzc7OYkzBLTAgWEDA5tEAgMDIWchAXkGBuYPxCsfBaNgFIyCkQQAgp89YW4UElAAAAAASUVORK5CYII=","orcid":"","institution":"Sejong University","correspondingAuthor":true,"prefix":"","firstName":"Hyun-Han","middleName":"","lastName":"Kwon","suffix":""}],"badges":[],"createdAt":"2024-05-11 15:29:04","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4405968/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4405968/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s00477-025-03092-z","type":"published","date":"2025-09-09T15:57:01+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":81200890,"identity":"3a51dc32-f640-4712-98e9-73e5ecaec3b0","added_by":"auto","created_at":"2025-04-23 11:15:06","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":507286,"visible":true,"origin":"","legend":"\u003cp\u003eOverview of the research framework, including SPEDI development, flash drought criteria definition, comparative analysis with existing indices, identification of flash drought characteristics in South Korea, and result interpretation.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-4405968/v1/c7e1dc1e5c73899e9c9fb2c6.png"},{"id":81200887,"identity":"f0b653ce-aa28-462b-9b2c-cb6fbdd5ad5d","added_by":"auto","created_at":"2025-04-23 11:15:06","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1501621,"visible":true,"origin":"","legend":"\u003cp\u003eAdministrative districts and geographic distribution of selected Automated Synoptic Observing Service (ASOS) stations across South Korea used in this study. Thiessen polygons were constructed to define the spatial domain of each station for spatial analysis.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-4405968/v1/a7a8fbe659ec0359f207f117.png"},{"id":81200888,"identity":"bb9d13a3-0d17-473a-8f0c-9135a309a108","added_by":"auto","created_at":"2025-04-23 11:15:06","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1220295,"visible":true,"origin":"","legend":"\u003cp\u003eSpatial distribution of agricultural damage areas(ha) to field crops as indicators of flash drought severity across districts for the years 2017, 2018, 2019, and 2022. The color intensity represents the extent of field crop damage, with darker shades indicating larger damaged areas. Gray shading indicates districts where no agricultural damage data were collected.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-4405968/v1/4a98f91e70542925a7469ad8.png"},{"id":81201434,"identity":"42db1a07-5e01-43c9-80f4-384e18ef74c5","added_by":"auto","created_at":"2025-04-23 11:23:06","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":2044808,"visible":true,"origin":"","legend":"\u003cp\u003eAggregated severity metrics of flash drought severity across districts for the years 2017, 2018, 2019, and 2022, aggregated over flash drought events from May to August, according to each drought index (SPI, SPEI, EDDI, and SPEDI). The color intensity represents the total severity scores, with darker shades indicating higher drought severity. Gray shading indicates districts where no agricultural damage data were collected.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-4405968/v1/2ffb6054f4ffddd5f4527ea8.png"},{"id":81200897,"identity":"80e57366-6f14-41d4-8626-0aa9a8953743","added_by":"auto","created_at":"2025-04-23 11:15:06","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":1486874,"visible":true,"origin":"","legend":"\u003cp\u003eAggregated duration metrics of flash drought duration across districts for the years 2017, 2018, 2019, and 2022, aggregated over flash drought events from May to August, according to each drought index (SPI, SPEI, EDDI, and SPEDI). The color intensity represents the cumulative duration of flash droughts in days, with darker shades indicating longer drought periods. Gray shading indicates districts where no agricultural damage data were collected.\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-4405968/v1/9618ae2130c2823cd922cce4.png"},{"id":81200910,"identity":"16f96c24-995e-4b1c-b8ae-c28e82c84314","added_by":"auto","created_at":"2025-04-23 11:15:07","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":5909566,"visible":true,"origin":"","legend":"\u003cp\u003eTemporal evolution of drought indices—SPI, SPEI, EDDI, and SPEDI—from 2017 to 2022, highlighting instances of flash droughts, identified based on agricultural damage data. Red-bordered empty boxes indicate periods with reported field crop damages, while gray-shaded areas represent periods with no available agricultural damage data. Colored shaded boxes indicate drought conditions as detected by each drought index. Lines represent SPI (dotted sky-blue line), SPEI (solid blue line), EDDI (dotted red line), and SPEDI (solid black line).\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-4405968/v1/573a7a139f3f63fec3a677c3.png"},{"id":81200904,"identity":"1e1392cb-2d3a-4a70-9bc3-ef517a22b9fa","added_by":"auto","created_at":"2025-04-23 11:15:06","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":170129,"visible":true,"origin":"","legend":"\u003cp\u003eCorrelation coefficient matrix between four drought indices: SPI, SPEI, EDDI, and SPEDI. The values represent Pearson correlation coefficients calculated from each index’s aggregated time series across all weather stations in South Korea.\u003c/p\u003e","description":"","filename":"7.png","url":"https://assets-eu.researchsquare.com/files/rs-4405968/v1/50231973ac07745032f610bb.png"},{"id":81200899,"identity":"404cd740-295a-4ed3-885e-fd2b974e0108","added_by":"auto","created_at":"2025-04-23 11:15:06","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":142503,"visible":true,"origin":"","legend":"\u003cp\u003eFrequency of each state in the Gangwon-do district spanning from 1980 to 2022. The classification of each state is based on atmospheric and land surface conditions, as defined in Table 4.\u003c/p\u003e","description":"","filename":"8.png","url":"https://assets-eu.researchsquare.com/files/rs-4405968/v1/1939715a74e4b6c9899cb2b9.png"},{"id":81200898,"identity":"3ee17d91-3d0f-4c07-ab4f-45415d7306f3","added_by":"auto","created_at":"2025-04-23 11:15:06","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":426491,"visible":true,"origin":"","legend":"\u003cp\u003eTemporal frequencydistribution of the four water balance states in the Gangwon-do district spanning from 1980 to 2022, analyzed across daily, monthly, and annual timescales. The classification of each state is based on atmospheric and land surface conditions, as defined in Table 4.\u003c/p\u003e","description":"","filename":"9.png","url":"https://assets-eu.researchsquare.com/files/rs-4405968/v1/fbf8bcc73b9e8f8621b7bcf9.png"},{"id":81200907,"identity":"37c50d2d-94fc-4bae-ba36-e3f8c33ed474","added_by":"auto","created_at":"2025-04-23 11:15:07","extension":"png","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":1773240,"visible":true,"origin":"","legend":"\u003cp\u003eDistinctions between SPI and EDDI states. Experimental comparison of precipitation and evaporation patterns in Gangwon-do according to simplified water balance states. Lines represent SPI (dotted sky-blue line), SPEI (solid blue line), EDDI (dotted red line), and SPEDI (solid black line).\u003c/p\u003e","description":"","filename":"10.png","url":"https://assets-eu.researchsquare.com/files/rs-4405968/v1/dc593478a8f2cd9ad5a081b4.png"},{"id":91359016,"identity":"abd8b1fd-e8e9-42e1-b00f-331388fe1e1d","added_by":"auto","created_at":"2025-09-15 16:04:11","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":16332846,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4405968/v1/c7e386b8-e0a0-4891-a29c-dd87c3a6141d.pdf"},{"id":81201431,"identity":"7732a208-5c70-4919-867d-3ed86c06a038","added_by":"auto","created_at":"2025-04-23 11:23:06","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":39730,"visible":true,"origin":"","legend":"","description":"","filename":"AppendixA.docx","url":"https://assets-eu.researchsquare.com/files/rs-4405968/v1/50448d2e82e43b14b3eaa0da.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Introducing a Novel Standardized Precipitation Evaporation Differential Index (SPEDI) for Improved Flash Drought Detection and Assessment: A Case Study in South Korea","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eDrought is a recurring natural hazard that affects a wide range of climate zones (Naveendrakumar et al., \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Traditionally, droughts have been classified into four types based on their distinctive characteristics and underlying causes (Wilhite and Glantz, \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e1985\u003c/span\u003e): meteorological, agricultural, hydrological, and socioeconomic droughts. More recently, the recognition of \u0026ldquo;flash droughts\u0026rdquo;\u0026mdash;events characterized by rapid onset and intensification\u0026mdash;has led to refinements in this classification system (Otkin et al., \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Pendergrass et al., \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). In contrast to the gradual development of typical drought events, flash drought can emerge in just two to six weeks, making timely detection and response particularly difficult. In particular, after a drought event in the US Central Plains in 2012, the sudden onset of the droughts during a short period was identified as a flash drought (Lisonbee et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). This recognition has led to a significant increase in studies that revisit several historical flash droughts.\u003c/p\u003e \u003cp\u003eFlash droughts can develop even under varying initial moisture conditions (Christian et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Despite the fundamental requirement of precipitation deficit for drought events, the emergence of flash drought is contingent upon additional climatic factors. The absence of rainfall, coupled with factors contributing to heightened evaporative demand (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{ET}_{0}\\)\u003c/span\u003e\u003c/span\u003e), such as elevated surface temperatures (i.e., severe heatwave), strong wind velocities, and clear skies persisting over several weeks, plays a crucial role in the rapid-onset and fast intensification of flash droughts (Otkin et al., \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Christian et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Svoboda et al., \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2002\u003c/span\u003e). Especially high temperatures can create conditions vulnerable to drought(Hu et al., \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Zhang et al., \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Won et al., \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), making it more likely for flash droughts to occur. Under elevated \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\text{E}\\text{T}}_{0}\\)\u003c/span\u003e\u003c/span\u003e conditions, a consequential depletion in soil moisture content occurs, leading to a simultaneous increase in evapotranspiration (ET) (Otkin et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Anderson et al., \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). If this combination persists for an extended duration (days to weeks), it has the potential to transform energy-limited conditions into water-limited conditions. This transition intensifies vegetation stress and eventually triggers a flash drought (Ford and Labosier, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Ford et al., \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Mozny et al., \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Hunt et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2009\u003c/span\u003e; Hunt et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2014\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eFlash droughts can be further categorized into two types based on specific physical mechanisms. The first category, known as \u0026ldquo;heatwave flash drought,\u0026rdquo; is characterized by exceptionally high temperatures driving an increase in ET and a simultaneous decrease in soil moisture. Heatwave flash droughts tend to be more pronounced in densely vegetated areas with comparatively high humidity. The second category, termed \u0026ldquo;precipitation-deficit flash drought,\u0026rdquo; is associated with a deficiency in precipitation, leading to reduced soil moisture and a subsequent decrease in ET (Wang and Yuan, \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Mo and Lettenmaier, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Zhang et al., \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Understanding both types requires tools that can account for the interaction between precipitation and evaporative demand.\u003c/p\u003e \u003cp\u003eIn this study, we focus on South Korea, a country particularly susceptible to flash droughts due to its climate and topography. Most of the annual rainfall is concentrated in the summer months (June to August), while the rest of the year, especially winter and early spring, tends to be extremely dry (Kim et al., \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). In addition, nearly 70% of the country\u0026rsquo;s terrain is mountainous, which complicates precipitation patterns and hydrological analysis. These geographical characteristics were evident in 2022, when the country experienced severe drought in the south and flooding in central regions. Reports of flash drought impacts have become more frequent (KMA, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Lee et al., \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), highlighting the need for improved assessment tools. However, most drought-related studies in South Korea focus on seasonal or long-term trends, with flash droughts remaining underexamined, accounting for only 4.11% of global drought studies according to the Web of Science (as of April 2025).\u003c/p\u003e \u003cp\u003eA key step toward better flash drought management is having a clear definition that distinguishes flash droughts from general droughts. The definition of flash drought includes rapid onset and intensification, and various studies have proposed different indicators for identification. For instance, Ford and Labosier (\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) utilized soil moisture observations at 0-40cm and quantified temporal differences in soil moisture to assess flash droughts in the United States. Conversely, Otkin et al. (\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2013\u003c/span\u003e) employed the evaporative stress index (ESI) as an alternative to soil moisture observations to assess flash drought. Pendergrass et al. (\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) proposed using the evaporative demand drought index (EDDI) based on standardized anomalies of \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\text{E}\\text{T}}_{0}\\)\u003c/span\u003e\u003c/span\u003e, highlighting shifts toward moisture-stressed environments. Noguera et al. (\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) evaluated flash drought in Spain using the Standardized Precipitation Evapotranspiration Index (SPEI; Vicente-Serrano et al., \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2010\u003c/span\u003e), considering the effects of both precipitation and ET. Parker et al. (\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) conducted a study in the Australian region, investigating flash drought detection criteria based on standardized precipitation index (SPI; McKee et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e1993\u003c/span\u003e) and EDDI, ESI, and standardized soil-moisture index (SSI).\u003c/p\u003e \u003cp\u003eIn line with earlier efforts, our study evaluates the suitability of commonly used drought indices\u0026mdash;SPI, SPEI, and EDDI\u0026mdash;for detecting flash droughts. SPI, based solely on precipitation, remains a widely used benchmark for meteorological droughts (McKee et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e1993\u003c/span\u003e; Kwon et al., \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Abu Arra and Şişman, \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Arra et al., \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Its simplicity, however, limits its ability to capture rapid drought development influenced by atmospheric demand. To address this, SPEI was developed by Vicente-Serrano et al. (\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2010\u003c/span\u003e) to incorporate both precipitation and potential evapotranspiration (PET), offering a more comprehensive perspective (Mozny et al., \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Qin et al., \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Yet, concerns remain over its reliability, especially when PET is overestimated. McEvoy et al. (\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2012\u003c/span\u003e) suggested using reference evapotranspiration (ET₀) as a more accurate input\u0026mdash;a method adopted in this study.\u003c/p\u003e \u003cp\u003eEDDI, unlike SPI and SPEI, is based entirely on ET₀ and represents atmospheric moisture demand through anomalies in water vapor flux. It was developed by Hobbins et al. (\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2016\u003c/span\u003e) and is particularly suited to identifying rapid drought developments. While these indices each capture important drought signals, SPI and EDDI rely on single variables and may fall short when applied independently. SPEI attempts to bridge this gap but has been shown to correlate more closely with SPI than with EDDI, suggesting it may not fully capture the interplay between precipitation and evaporative demand (Yao et al., \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Won et al., \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eTo address the limitations of existing drought indices, we propose a new composite index: the Standardized Precipitation Evaporation Differential Index (SPEDI). Designed to combine the benefits of SPI and EDDI, SPEDI highlights conditions in which low precipitation coincides with high evaporative demand\u0026mdash;characteristics typical of flash droughts. By independently calculating and then differencing SPI and EDDI, SPEDI retains the key features of both, offering a more balanced and responsive measure.\u003c/p\u003e \u003cp\u003eThis study investigates flash drought behavior in South Korea by analyzing four indices\u0026mdash;SPI, SPEI, EDDI, and SPEDI\u0026mdash;across multiple seasonal timescales. Our objectives are threefold: First, we seek to map the spatiotemporal patterns of flash droughts to better understand their frequency and distribution across South Korea. Second, we evaluate the effectiveness of the SPI, SPEI, EDDI, and SPEDI indices in capturing seasonal differences in flash drought behavior. Third, we explore the key meteorological drivers that influence the seasonal and regional variations of flash droughts, with an emphasis on improving early detection and management.\u003c/p\u003e \u003cp\u003eTo evaluate the effectiveness of each drought index, we use the observed agricultural damage area as a proxy for flash drought severity. This allows us to assess how well each index captures drought conditions that result in actual impacts. In doing so, we aim to demonstrate the practical advantages of SPEDI compared to conventional indices. The study is structured around four key steps: (1) development of the SPEDI index, (2) definition of flash drought criteria, (3) identification of flash drought events in South Korea, and (4) comparative analysis of index performance, as summarized in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e[Insert Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e]\u003c/p\u003e"},{"header":"2. Study Area and Meteorological Data","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Study area\u003c/h2\u003e \u003cp\u003eSouth Korea spans a peninsular area of approximately 100,364 km\u003csup\u003e2\u003c/sup\u003e, with its northern border adjacent to the continent and all remaining borders surrounded by sea. The topography is approximately 30% flat terrain and 70% mountainous regions. Despite the abundance of mountains, only 15% of them surpass an elevation of 1,000 meters above sea level, and more than 65% peak below 500 meters. Of note is the distinctive topographical layout of the Korean Peninsula, which features prominent mountainous regions primarily lining the eastern border of the peninsula, as exemplified by the Taebaek Mountains. The key characteristic of the peninsula\u0026rsquo;s river systems is the geographical orientation of their main watersheds, with the Taebaek Mountains limiting the area for the watershed toward the East Sea. Consequently, major rivers predominantly flow southwestward, ultimately connecting to the Yellow or South Seas. In contrast, those rivers that discharge into the East Sea follow shorter courses.\u003c/p\u003e \u003cp\u003eIn addition to topography, substantial seasonal variations in precipitation and temperature shape hydrological characteristics. Located within the geographic coordinates of 33\u0026ndash;38\u0026deg;N and 126\u0026ndash;132\u0026deg;E, the region experiences significant influence from monsoon winds, leading to cold and dry conditions in winter and a hot and humid climate in summer. The annual precipitation averages around 1,306mm, accompanied by a mean temperature range of 7\u0026ndash;15℃. Approximately 54% of the annual precipitation falls during the summer season, establishing it as the dominant period for rainfall (Kwon et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). The mean annual temperature remains consistently at 13℃, with seasonal temperature extremes ranging from 1℃ in winter to 24℃ during the summer. These climatic variations contribute to the dynamic hydrological patterns observed in South Korea.\u003c/p\u003e \u003cp\u003e[Insert Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e]\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Meteorological Dataset\u003c/h2\u003e \u003cp\u003eThis study utilized data from 59 Automated Synoptic Observing Service (ASOS) stations operated by the Korea Meteorological Administration. The ASOS network, initially established in the early 1900s, provides consistent and systematic records of meteorological variables, including precipitation, solar radiation, and air temperature, at both hourly and daily resolutions. For the primary drought analysis, we used daily meteorological data from 1980 to 2022 to ensure sufficient temporal coverage for retrospective evaluation. In addition, a subset of data from 2017 to 2022 was used to align with the available validation dataset on agricultural damage (2018\u0026ndash;2022).\u003c/p\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e presents the spatial distribution of the ASOS stations used in this study along with their associated Thiessen polygons. Based on the computed polygons, each station covers an average area with a diameter of approximately 40.05 km, ranging from a minimum of 8.76 km to a maximum of 61.18 km. These values reflect notable spatial variability in station density. While the overall coverage is generally acceptable for regional-scale analysis, some mountainous regions appear to be underrepresented, which may affect the accuracy in localized drought detection. Nevertheless, the primary aim of this study is to provide broad insights into flash drought detection across South Korea using multiple drought indices. Accordingly, we prioritized the use of observation data with high temporal and spatial resolution to support the core analysis.\u003c/p\u003e \u003cp\u003eThe dataset served as the foundation for generating the drought indices SPI, SPEI, EDDI, and SPEDI. Recognizing the absence of directly measured ET, this study intentionally derived \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\text{E}\\text{T}}_{0}\\)\u003c/span\u003e\u003c/span\u003e using a multiscale surrogate model based on the Hargreaves-Samani equation proposed by Kim et al. (\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). To facilitate a comprehensive evaluation of flash droughts by region, point data from ASOS datasets were interpolated into an administrative district areal dataset. This spatial approach enhances the regional assessment of drought conditions, contributing to a more comprehensive understanding of the phenomenon under investigation.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3 Field Crop Damage Dataset During the Summer Season\u003c/h2\u003e \u003cp\u003eFlash drought can result in a variety of damages, including agricultural losses, wildfires, and ecological impacts. The government evaluates the damage to each field and crop by surveying the affected areas, subsequently calculating the damage ratio. The agricultural crop damage ratio represents the proportion of the area that cannot be harvested (determined by the damaged area multiplied by the damage ratio) in comparison to the entire cultivated area of crops. In this study, our primary emphasis is on agricultural damage incurred in field crops rather than rice paddies, which we have identified as the key indicator for assessing flash droughts in South Korea. Field crops in South Korea primarily depend on river or groundwater irrigation, whereas rice paddies benefit from irrigation supplied through agricultural reservoirs. This distinction in irrigation sources makes rice paddies less susceptible to the direct impacts of flash droughts compared to field crops. Moreover, most damage to field crops is caused by flash droughts that occur during the summer. In these contexts, we gathered a dataset on agricultural damage to field crops from the Ministry of Agriculture, Food, and Rural Affairs, as summarized in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. The dataset includes officially reported damage during the summer period (June to August) for the years 2017, 2018, 2019, and 2022, in which field-level drought damage was documented. In some administrative districts, no drought damage was reported or recorded during these years. A summary of the dataset is provided in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eTotal areas, field crop area, and flash drought-induced field crop damage by district from June to August for the years 2017, 2018, 2019, and 2022 (ha) (N/A indicates that no investigation was conducted or data was not collected for the respective district.)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"7\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eDistrict Name\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eTotal\u003c/p\u003e \u003cp\u003eArea\u003c/p\u003e \u003cp\u003e(ha)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eField Crop Land Use (ha)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"4\" nameend=\"c7\" namest=\"c4\"\u003e \u003cp\u003eField Crop Damage (ha)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2017\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003e2018\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003e2019\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003e2022\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSeoul\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e60,500\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e407\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eN/A\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eN/A\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eN/A\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eN/A\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBusan\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e77,100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e2,519\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e286\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e795\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDaegu\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e149,900\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e3,873\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2,620\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1,605\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e560\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e55\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIncheon\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e106,700\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e5,154\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGwangju\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e50,100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e4,165\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eN/A\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eN/A\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eN/A\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eN/A\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDaejeon\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e54,000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e2,656\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e64\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e4,402\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e2,078\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e118\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUlsan\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e106,300\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e4,722\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSejong-si\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e46,500\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e3,401\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2,447\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e630\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e133\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGyeonggi-do\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1,020,000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e70,288\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1,086\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1,805\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e13\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGangwon-do\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1,683,100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e70,037\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e279\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e1,216\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eChungcheongbuk-do\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e740,700\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e60,128\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e850\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e384\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eChungcheongnam-do\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e824,800\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e73,657\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e8,649\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eJeollabuk-do\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e807,300\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e67,595\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2,383\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e2,875\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e238\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eJellanam-do\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1,236,200\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e110,747\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e291\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1,126\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGyeongsangbuk-do\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1,842,400\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e143,350\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGyeongsangnam-do\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1,054,300\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e69,864\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eN/A\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eN/A\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eN/A\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eN/A\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eJeju-do\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e185,000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e55,593\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e[Insert Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e]\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Methodology","content":"\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e3.1 Thiessen Polygon Method\u003c/h2\u003e \u003cp\u003eThe meteorological dataset used in this study consists of point-based observations collected from individual stations, whereas the validation data dataset is aggregated by administrative districts. To bridge this mismatch in spatial format, we applied the Thiessen polygon method to convert point observations into area-weighted averages.\u003c/p\u003e \u003cp\u003eThis method assigns weights to each observation station based on its proximity to neighboring stations. This is achieved by constructing polygons through the perpendicular bisector between stations, forming Thiessen polygons that delineate each station\u0026rsquo;s zone of influence (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Using these polygons, spatially weighted averages of meteorological variables were computed for each administrative district used in the validation dataset. The weighted value \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{P}_{n}\\)\u003c/span\u003e\u003c/span\u003e for each district is calculated as follows:\u003cdiv id=\"Equ1\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ1\" name=\"EquationSource\"\u003e\n$$\\:{P}_{n}=\\frac{\\sum\\:{P}_{i}{A}_{i}}{\\sum\\:{A}_{i}}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e1\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{P}_{n}\\)\u003c/span\u003e\u003c/span\u003e is the meteorological value at station \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:i\\)\u003c/span\u003e\u003c/span\u003e, and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{A}_{i}\\)\u003c/span\u003e\u003c/span\u003e is the area of the Thiessen polygon within the target administrative district.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e\u003cb\u003e3.2 Computation of Reference Evapotranspiration\u003c/b\u003e \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\varvec{E}\\varvec{T}}_{\\mathbf{o}}\\)\u003c/span\u003e\u003c/span\u003e\u003c/h2\u003e \u003cp\u003eIn regions with limited hydrometeorological data, it is crucial to adopt surrogate equations for calculating evapotranspiration (ET) or reference evapotranspiration \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\text{E}\\text{T}}_{0}\\)\u003c/span\u003e\u003c/span\u003e. Numerous empirical equations have been derived from extensive research to reflect the underlying physical relationships between ET and meteorological input variables. Various approaches for estimating \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{ET}_{0}\\)\u003c/span\u003e\u003c/span\u003e have been widely applied in hydrology, including the Hargreaves method (Hargreaves and Samani, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e1985\u003c/span\u003e), FAO56 Penman-Monteith method (Allen et al., \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e1998\u003c/span\u003e), and ASCE PM method (Allen et al., \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2005\u003c/span\u003e). This study employs the temperature-based Hargreaves equation and physically-based FAO56 Penman-Monteith equation (FAO56 PM) to estimate \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\text{E}\\text{T}}_{0}\\)\u003c/span\u003e\u003c/span\u003e:\u003cdiv id=\"Equ2\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ2\" name=\"EquationSource\"\u003e\n$$\\:{ET}_{0\\left(PM\\right)}=\\frac{0.408\\varDelta\\:\\left({R}_{n}-G\\right)+\\gamma\\:\\frac{900}{\\gamma\\:+273}{u}_{2}({e}_{s}-{e}_{a})}{\\varDelta\\:+\\gamma\\:(1+0.34{u}_{2})}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e2\u003c/div\u003e\u003c/div\u003e\u003cdiv id=\"Equ3\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ3\" name=\"EquationSource\"\u003e\n$$\\:{ET}_{0\\left(HS\\right)}=0.0023\\bullet\\:RA\\times\\:(T℃+17.8)\\times\\:{({T}_{max}-{T}_{min})}^{0.50}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e3\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eEquation (\u003cspan refid=\"Equ1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) represents the FAO56 PM equation, where \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{R}_{n}\\)\u003c/span\u003e\u003c/span\u003e is the net radiation (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:MJ/{m}^{2}/day\\)\u003c/span\u003e\u003c/span\u003e), \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:G\\)\u003c/span\u003e\u003c/span\u003e is the soil heat flux density (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:MJ/{m}^{2}/day\\)\u003c/span\u003e\u003c/span\u003e), \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\text{T}\\)\u003c/span\u003e\u003c/span\u003e is the average temperature at a 2m height (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:℃\\)\u003c/span\u003e\u003c/span\u003e), \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{u}_{2}\\)\u003c/span\u003e\u003c/span\u003e is the wind speed at a 2m height (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:m/s\\)\u003c/span\u003e\u003c/span\u003e), \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{e}_{s}\\)\u003c/span\u003e\u003c/span\u003e is the saturation vapor pressure (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:kPa\\)\u003c/span\u003e\u003c/span\u003e), \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{e}_{a}\\)\u003c/span\u003e\u003c/span\u003e is the actual vapor pressure (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:kPa\\)\u003c/span\u003e\u003c/span\u003e), \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{e}_{s}-{e}_{a}\\)\u003c/span\u003e\u003c/span\u003e is the saturation vapor pressure deficit (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:kPa)\\)\u003c/span\u003e\u003c/span\u003e, and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\varDelta\\:\\)\u003c/span\u003e\u003c/span\u003e is the slope vapor pressure curve (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:kPa/℃\\)\u003c/span\u003e\u003c/span\u003e) (Allen et al., \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e1998\u003c/span\u003e). Eq.\u0026nbsp;(\u003cspan refid=\"Equ2\" class=\"InternalRef\"\u003e2\u003c/span\u003e) represents the HS equation, where \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:RA\\)\u003c/span\u003e\u003c/span\u003e is the extraterrestrial radiation in \u003cem\u003emm/day\u003c/em\u003e; \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\text{T}\\)\u003c/span\u003e\u003c/span\u003e is the average temperature; and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\text{T}}_{\\text{m}\\text{a}\\text{x}}\\)\u003c/span\u003e\u003c/span\u003e and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\text{T}}_{\\text{m}\\text{i}\\text{n}}\\)\u003c/span\u003e\u003c/span\u003e are the maximum and minimum temperature in \u0026deg;C, respectively (Hargreaves and Samani, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e1985\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe HS equation, initially formulated with fixed regression parameters, has the potential to produce biased estimates when applied in different climatic regimes (Valero et al., \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). In response to this concern, Samani (\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2000\u003c/span\u003e) recommended the calibration of coefficients to better represent local climate conditions. Several publications have suggested improved parameters for the HS equation in the different climate zones (Gavil\u0026aacute;n et al., \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Martınez-Cob and Tejero-Juste, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Raziei and Pereira, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Yang et al., \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). The general form of the HS equation can be represented by Eq.\u0026nbsp;(\u003cspan refid=\"Equ3\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003cdiv id=\"Equ4\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ4\" name=\"EquationSource\"\u003e\n$$\\:{ET}_{0}=\\alpha\\:\\bullet\\:RA\\times\\:(T+\\beta\\:)\\times\\:{({T}_{max}-{T}_{min})}^{\\gamma\\:}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e4\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eIn a recent study conducted by Kim et al. (\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), a Bayesian approach was developed within a multi-scale modeling framework to estimate parameters for the HS equation, maintaining accurate estimations of evapotranspiration across various temporal scales. More specifically, the proposed surrogate model offers a new perspective for extending the existing HS equation to encompass daily, monthly, and annual scales. The three parameters (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\alpha\\:,\\:\\beta\\:,\\:\\gamma\\:\\)\u003c/span\u003e\u003c/span\u003e) of the multi-scale surrogate model were recalibrated and validated using \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\text{E}\\text{T}}_{0}\\)\u003c/span\u003e\u003c/span\u003e derived from the FAO56 PM method. In a previous study, the performance of the multi-scale surrogate model was assessed by categorizing it based on different temporal scales (i.e., daily, monthly, annual). The findings demonstrated that the multi-scale model outperformed a single-scale model in minimizing systematic bias across various temporal scales. In our current study, we build upon the findings from that previous research (Kim et al., \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), and the calibrated parameters were used for estimating \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{ET}_{0}\\)\u003c/span\u003e\u003c/span\u003e, which can be used to estimate EDDI. The parameters used for this study are detailed in Appendix A.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e3.3 Computation of Drought Indices\u003c/h2\u003e \u003cp\u003eIn this study, we utilize three widely accepted drought indices\u0026mdash;SPI, SPEI, and EDDI\u0026mdash;with a one-month accumulation period and daily temporal resolution to effectively capture flash drought conditions, which are characterized by rapid onset and intensification within a few weeks.\u003c/p\u003e \u003cp\u003e \u003cb\u003eComputation of SPI\u003c/b\u003e \u003c/p\u003e \u003cp\u003eThe SPI computation process involves fitting the accumulated precipitation data to a probability distribution, typically the gamma distribution, computing cumulative probability for the corresponding precipitation deficits, and then mapping these probabilities into standardized scores using the inverse cumulative standard normal distribution (McKee et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e1993\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eFirst, precipitation anomalies are quantified by fitting accumulated precipitation for a given time period (e.g., 1, 3, or 12 months) to a gamma distribution. The gamma probability density function is defined as:\u003cdiv id=\"Equ5\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ5\" name=\"EquationSource\"\u003e\n$$\\:f\\left(x\\right)=\\:\\frac{1}{{\\Gamma\\:}\\left(\\alpha\\:\\right){\\beta\\:}^{\\alpha\\:}}{x}^{\\alpha\\:-1}\\text{e}\\text{x}\\text{p}\\left(-\\frac{x}{\\beta\\:}\\right)$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e5\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:x\\)\u003c/span\u003e\u003c/span\u003e is the accumulated precipitation amount, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\alpha\\:\\)\u003c/span\u003e\u003c/span\u003e is the shape parameter, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\beta\\:\\)\u003c/span\u003e\u003c/span\u003e is the scale parameter, and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\Gamma\\:}\\left(\\alpha\\:\\right)\\)\u003c/span\u003e\u003c/span\u003e is the gamma function. These parameters \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\alpha\\:\\)\u003c/span\u003e\u003c/span\u003e and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\beta\\:\\)\u003c/span\u003e\u003c/span\u003e are estimated using the maximum likelihood method. Because precipitation time series may include zero values, the cumulative distribution function (CDF) \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:G\\left(x\\right)\\)\u003c/span\u003e\u003c/span\u003e is adjusted to account for the probability of zero precipitation. The adjusted cumulative probability, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:H\\left(x\\right)\\)\u003c/span\u003e\u003c/span\u003e, is computed as:\u003cdiv id=\"Equ6\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ6\" name=\"EquationSource\"\u003e\n$$\\:H\\left(x\\right)=q+\\left(1-q\\right)G\\left(x\\right)$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e6\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:q\\)\u003c/span\u003e\u003c/span\u003e is the probability of precipitation events in the time series., Finally, PI is obtained by transforming \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:H\\left(x\\right)\\)\u003c/span\u003e\u003c/span\u003e to a standard normal distribution with a mean value of 0 and a standard deviation of 1. Positive SPI values indicate the wet-than-average conditions, while negative values indicate drier-than-average conditions.\u003c/p\u003e \u003cp\u003e \u003cb\u003eComputation of SPEI\u003c/b\u003e \u003c/p\u003e \u003cp\u003eThe SPEI extends the SPI framework by incorporating precipitation and reference evapotranspiration (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{ET}_{0}\\)\u003c/span\u003e\u003c/span\u003e) to account for atmospheric moisture demand (Vicente-Serrano et al., \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). It uses the climatic water balance, defined as \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:D=P-{ET}_{0}\\)\u003c/span\u003e\u003c/span\u003e, ​where \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:P\\)\u003c/span\u003e\u003c/span\u003e is precipitation and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{ET}_{0}\\)\u003c/span\u003e\u003c/span\u003e is reference evapotranspiration, as its input variable. This balance is accumulated over a specified time scale to capture short- to medium-term drought conditions.\u003c/p\u003e \u003cp\u003eUnlike SPI, which typically uses the gamma distribution, SPEI is fitted to a log-logistic distribution, which can accommodate both positive and negative values. The log-logistic CDF is given by:\u003cdiv id=\"Equ7\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ7\" name=\"EquationSource\"\u003e\n$$\\:f\\left(D\\right)=\\frac{\\beta\\:}{\\alpha\\:}\\times\\:\\frac{{\\left(D/\\alpha\\:\\right)}^{\\beta\\:-1}}{\\alpha\\:{\\left[1+{\\left(D/\\alpha\\:\\right)}^{\\beta\\:}\\right]}^{2}}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e7\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\alpha\\:\\)\u003c/span\u003e\u003c/span\u003e, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\beta\\:\\)\u003c/span\u003e\u003c/span\u003e, and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\gamma\\:\\)\u003c/span\u003e\u003c/span\u003e represent the scale, shape, and location parameters, respectively. The resulting probability is then transformed into a standard normal variate, consistent with the SPI procedure.\u003c/p\u003e \u003cp\u003e \u003cb\u003eComputation of EDDI\u003c/b\u003e \u003c/p\u003e \u003cp\u003eEDDI, originally proposed by Hobbins et al. (\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2016\u003c/span\u003e), uses accumulated \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{ET}_{0}\\)\u003c/span\u003e\u003c/span\u003e as the input to represent atmospheric moisture demand. EDDI reflects how strongly the atmosphere is \u0026ldquo;pulling\u0026rdquo; moisture from the surface, and is particularly sensitive to conditions such as heatwaves, persistent clear skies, and strong winds.\u003c/p\u003e \u003cp\u003eWhile the original method relies on a non-parametric empirical approach using the Tukey plotting position formula, we adopted a parametric approach in this study for consistency with other indices. Specifically, we applied the gamma distribution to fit the accumulated \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{ET}_{0}\\)\u003c/span\u003e\u003c/span\u003e data over a one-month time scale. This choice was based on Kolmogorov-Smirnov (K-S) goodness-of-fit tests conducted across multiple locations in South Korea, which confirmed the gamma distribution as a suitable model for \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{ET}_{0}\\)\u003c/span\u003e\u003c/span\u003e ​.\u003c/p\u003e \u003cp\u003eThe computation procedure follows the general SPI framework, substituting precipitation with accumulated \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{ET}_{0}\\)\u003c/span\u003e\u003c/span\u003e. The resulting values are then transformed to standard normal variates to produce standardized EDDI scores. Unlike SPI and SPEI, where negative values indicate drier conditions, EDDI conventionally defines positive values as indicative of dryness. To maintain consistency across all indices used in this study, we inverted the sign of the EDDI values so that negative values also represent dry conditions, aligning the interpretation of all indices.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e3.4 Development of Composite Drought Index SPEDI\u003c/h2\u003e \u003cp\u003eAccurate assessment of flash drought requires accounting for both precipitation deficit and increased atmospheric evaporative demand. While SPI and EDDI each represent one of these components, their single-variable nature limits their ability to fully characterize flash drought conditions. SPEI was developed to integrate both precipitation and evapotranspiration; however, previous studies have shown that it tends to correlate more closely with SPI than with EDDI, suggesting it may not adequately capture the independent effects of evaporative demand (Yao et al., \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Won et al., \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eGiven these limitations, we propose a novel composite drought index, SPEDI (Standardized Precipitation Evaporation Differential Index), designed to integrate the distinct signals from SPI and EDDI. The index is intended to provide a more balanced representation of drought severity by combining independently derived anomalies from precipitation and atmospheric dryness.\u003c/p\u003e \u003cp\u003eWe define an intermediate variable Total Water Deficit (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:TWD\\)\u003c/span\u003e\u003c/span\u003e) as the sum of SPI and sign-reversed EDDI:\u003cdiv id=\"Equ8\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ8\" name=\"EquationSource\"\u003e\n$$\\:TWD=SPI+EDDI$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e8\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eNote that in this study, we reversed the sign of EDDI values so that negative values indicate dry conditions, aligning with the interpretation of SPI and SPEI. If the original (non-reversed) EDDI values are used, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:TWD\\)\u003c/span\u003e\u003c/span\u003e should instead be computed as:\u003cdiv id=\"Equ9\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ9\" name=\"EquationSource\"\u003e\n$$\\:TWD=SPI-EDDI$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e9\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eHigher value of \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:TWD\\)\u003c/span\u003e\u003c/span\u003e represent conditions where both precipitation is below normal and atmospheric moisture demand \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{ET}_{0}\\)\u003c/span\u003e\u003c/span\u003e is elevated, hallmarks of flash drought. To standardize \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:TWD\\)\u003c/span\u003e\u003c/span\u003e and allow direct comparison with other drought indices, we apply Z-score normalization:\u003cdiv id=\"Equ10\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ10\" name=\"EquationSource\"\u003e\n$$\\:SPEDI=\\frac{TWD-\\mu\\:}{\\sigma\\:}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e10\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\mu\\:\\)\u003c/span\u003e\u003c/span\u003e and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\sigma\\:\\)\u003c/span\u003e\u003c/span\u003e are the mean and standard deviation of TWD over the reference period.\u003c/p\u003e \u003cp\u003eBy explicitly combining two independently derived indicators, SPEDI retains the full signal of both variables and offers a more integrated and sensitive measure of flash drought conditions.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e3.5 Flash Drought Identification (criteria)\u003c/h2\u003e \u003cp\u003eThe criteria for identifying dry conditions are outlined in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e, following the percentile-based classification system proposed by Svoboda et al. (\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2002\u003c/span\u003e). In this framework, drought severity is classified based on the percentile ranking of drought index values: severe drought corresponds to values below the 10th percentile, moderate drought to the 10th -20th percentile range, and mild drought to the 20th -40th percentile range. Values above the 40th percentile are considered normal. This percentile-based approach enables consistent detection of drought intensity across different regions and time periods.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eCriteria for identifying dry conditions using drought indices (SPI, SPEI, EDDI, and SPEDI) based on the percentile ranking system by Svoboda et al. (\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2002\u003c/span\u003e)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePercentile of Drought Indices\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDrought conditions\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLess than 10%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSevere drought\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e10\u0026thinsp;~\u0026thinsp;20%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eModerate drought\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e20\u0026thinsp;~\u0026thinsp;40%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMild drought\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGreater than 40%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNormal\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eFlash droughts, in contrast to seasonal or long-term droughts, are defined by two distinguishing features: (2) rapid onset, typically developing within 2 to 6 weeks. (2) rapid intensification, often resulting in substantial impacts before conventional monitoring systems can detect them. These characteristics distinguish flash drought from traditional drought events that develop gradually over months or seasons.\u003c/p\u003e \u003cp\u003eBuilding on these characteristics, numerous studies have proposed criteria for flash drought identification (Ford and Labosier, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Pendergrass et al., \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Parker et al., \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). In this study, we developed a region-specific definition tailored to South Koreans. A flash drought event is defined as a rapid decline in drought index values, where the percentile shifts from above the 50\u003csup\u003eth\u003c/sup\u003e percentile to below the 20\u003csup\u003eth\u003c/sup\u003e percentile within a 14-day span. This indicates a rapid shift from normal to moderate drought within two weeks. This threshold captures a transition from normal to at least moderate drought within two weeks. This criterion provides a quantifiable and operational definition of flash drought onset, facilitating both early detection and more responsive drought management in rapidly evolving scenarios.\u003c/p\u003e \u003cp\u003e[Insert Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e]\u003c/p\u003e \u003c/div\u003e"},{"header":"4. Results","content":"\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e4.1 Spatial Distribution of Flash Drought\u003c/h2\u003e \u003cp\u003eIn this study, it is assumed that the extent of agricultural damage areas to field crops can serve as an indicator of flash drought severity, enabling their use in assessing flash drought indices. The distribution of agricultural damage areas is presented visually for four specific years (2017, 2018, 2019, and 2022) in each district, as illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. Metrics related to severity and duration were aggregated for the flash drought events that occurred between May and August, as presented in Figs.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e and \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e. Here, the flash drought events are defined as described in Section \u003cspan refid=\"Sec8\" class=\"InternalRef\"\u003e3.2\u003c/span\u003e.\u003c/p\u003e \u003cp\u003eThe flash drought assessment data used in this study rely exclusively on agricultural damage to field crops, which only accounts for agricultural areas associated with field crops. Therefore, this study does not account for other types of damage that may result from flash drought events, primarily due to limitations associated with the available dataset.\u003c/p\u003e \u003cp\u003eIn 2017, SPI and SPEI exhibited limitations in accurately representing the extent of damage caused by flash drought. In contrast, both EDDI and SPEDI effectively identified and delineated the drought-affected areas. Indeed, SPEDI offered a notably more comprehensive and extensive representation compared to other drought indices.\u003c/p\u003e \u003cp\u003eIn 2018, a concentration of drought was evident on the western side of the continent, with all drought indices indicating flash droughts throughout the country. The Chungcheongnam-do district, in particular, emerged as a focal point due to its substantial agricultural damage during the validation period. Significantly, SPEDI exhibited a similar spatial trend, highlighting its effectiveness in accurately representing the observed patterns of agricultural damage.\u003c/p\u003e \u003cp\u003eIn 2019, a year marked by limited agricultural damage, both EDDI and SPEDI revealed a distribution of flash droughts across the entire country. In contrast, both SPI and SPEI suggested that these events were concentrated in the northern regions, failing to capture the drought event in the Daegu district. This emphasizes the capability of EDDI and SPEDI to offer a more comprehensive perspective of flash drought occurrences, particularly in regions that may not be identified by other indices.\u003c/p\u003e \u003cp\u003eIn 2022, the Gangwon-do district experienced significant agricultural damage due to drought. The SPI and EDDI indices effectively identified these drought events. Conversely, the SPEI and SPEDI indices failed to detect drought events in this district during the same period. This discrepancy emphasizes the varied sensitivities of these indices to the regional patterns of drought, indicating the importance of drought indices for specific drought characteristics and locations.\u003c/p\u003e \u003cp\u003e[Insert Figs.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e\u0026ndash;5]\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003e4.2 Temporal Variation of Drought Indices\u003c/h2\u003e \u003cp\u003eThis chapter\u0026rsquo;s primary objective is to assess the effectiveness of drought indices in detecting and characterizing flash drought events. The criteria for identifying these events, as previously outlined, were applied to recognize these events. Subsequently, validation was carried out using the dataset listed in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. Figure\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e6\u003c/span\u003e presents the temporal evolution of drought indices\u0026mdash;SPI, SPEI, EDDI, and SPEDI\u0026mdash;from 2017 to 2022 and highlights the instances of flash drought, as determined through agricultural damage data.\u003c/p\u003e \u003cp\u003e[Insert Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e6\u003c/span\u003e]\u003c/p\u003e \u003cp\u003eDuring periods marked by significant flash drought occurrences, such as in 2018, the four drought indices exhibited consistent results, except in the Jeju-do district, as illustrated in Fig.\u0026nbsp;5(g). The drought event in Gyeonggi-do in 2017 was also detected effectively by all four indices. SPI, SPEI, and SPEDI proactively detected the drought, while EDDI precisely aligned with the validation period, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e6\u003c/span\u003e(a).\u003c/p\u003e \u003cp\u003eSome drought events, particularly those categorized as \u0026ldquo;precipitation-deficit flash drought\u0026rdquo;, were not detected by EDDI. For example, the 2018 event in Jeju-do district (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e6\u003c/span\u003e(g)) and the 2019 event in Chungcheongbuk-do district (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e6\u003c/span\u003e(d)) were captured by SPI, SPEI, and SPEDI but missed by EDDI. This highlights the importance of precipitation-based indices in identifying droughts driven primarily by rainfall deficiency.\u003c/p\u003e \u003cp\u003eOn the other hand, there were cases where SPEI failed to detect drought, while other indices responded effectively. A notable example is the 2022 drought in Gangwon-do (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e6\u003c/span\u003e(b)), which was identified by SPI, EDDI, and SPEDI, but not by SPEI. This inconsistency further supports findings that SPEI may underrepresent evaporative demand under certain conditions.\u003c/p\u003e \u003cp\u003eEvents classified as \u0026ldquo;heatwave flash drought\u0026rdquo; were also observed. For the 2022 drought in the Gyeonggi-do district, only EDDI and SPEDI detected the event in advance, as illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e6\u003c/span\u003e(a). In contrast, SPI and SPEI indices did not indicate a drought, suggesting that precipitation levels alone were not abnormally low. In this case, elevated temperatures and resulting high evaporative demand elevated temperatures and resulting high evaporative demand played a key role in the development of drought. This reinforces the importance of incorporating both precipitation deficits and atmospheric evaporative demand when identifying flash droughts.\u003c/p\u003e \u003cp\u003eLast, in some instances, drought events were identified by a single drought index. Specifically, the 2017 drought in the Jeollanam-do district was detected by the SPEI, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e6\u003c/span\u003e(e). Similarly, the drought in the Gangwon-do district in the same year was only identified during the period highlighted by the EDDI, as displayed in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e6\u003c/span\u003e(b). Again, this suggests that each index has strengths and limitations in capturing specific drought patterns and events.\u003c/p\u003e \u003cp\u003eOur comparative analysis of drought indices in evaluating flash drought events revealed distinct patterns. EDDI consistently detected a higher number of flash drought events compared to the other drought indices. The SPI and SPEI exhibited similar tendencies, often underestimating the duration of drought events in the historical period. SPEDI effectively addressed the issues of overestimation and underestimation that were evident in other indices. Throughout the study period, across the 14 flash drought events identified in South Korea, the SPEDI index demonstrated better performance in accurately detecting the periods of agricultural damage (see Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eComparative performance of drought indices (EDDI, SPI, SPEI, SPEDI) in detecting flash drought events based on agricultural damage periods in South Korea\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSPI\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSPEI\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eEDDI\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eSPEDI\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eTotal\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEarly Detection\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e14\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMatching\u003c/p\u003e \u003cp\u003ePeriod\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e14\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e[Insert Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e]\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e4.3 Assessment of SPEDI Performance Compared to SPEI\u003c/h2\u003e \u003cp\u003eIn our study, we categorized the complex interactions between precipitation and evaporation into a simplified model, outlining four distinct states that characterize the water balance dynamics between the atmosphere and the land surface, as summarized in Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eSimplified model of water balance dynamics: categorization of atmospheric and land surface interactions into four distinct states\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eState\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAtmosphere\u003c/p\u003e \u003cp\u003eCondition\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eLand\u003c/p\u003e \u003cp\u003eCondition\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eEvaporative\u003c/p\u003e \u003cp\u003eDemand\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eActual\u003c/p\u003e \u003cp\u003eEvaporative\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eDrought\u003c/p\u003e \u003cp\u003eRisk\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eWet\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eWet\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eLow\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eLow\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eVery low\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eWet\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eDry\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eLow\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eLow\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eLow\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDry\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eWet\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eHigh\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eHigh\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eModerate\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDry\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eDry\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eHigh\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eLow\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eHigh\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eState 1 is defined by very humid atmospheric conditions, where there is no additional evaporation as the atmosphere is saturated, even if the land surface is moist. State 2 also represents a humid atmosphere but with dry land conditions. Here, both the demand for evaporation and the actual evaporation rate are low because the dry land surface tends to absorb moisture from the air rather than release it. In contrast, State 3 marks the onset of a precipitation deficit characterized by a dry atmosphere and a still-wet land surface. Under these conditions, the atmosphere exerts a higher evaporative demand, and the land surface, having adequate moisture availability, responds with increased evaporation. Although this state may be considered the onset of a drought, it is described more accurately as a transitional (or normal condition), pending further development. Finally, State 4 emerges if the precipitation deficit continues, marking a progression from the conditions described in State 3. State 4 can be classified as a severe drought, marked by a dry atmosphere in need of moisture recovery but with a dry land surface unable to provide it. This condition sustains a high level of evaporative demand that exceeds the land\u0026rsquo;s capacity to supply moisture, leading to a failure in atmospheric recovery and an ongoing pattern of moisture demand. Consequently, the atmosphere continues to demand greater evaporation, resulting in a progressively drier land surface until soil moisture is depleted. Unlike State 3, where actual evaporation may be significant, in State 4, evaporative demand continues to increase while actual evaporation decreases due to the deficiency of available moisture. This state of severe drought can rapidly intensify under abnormal weather conditions, further exacerbating the moisture imbalance between the atmosphere and the land surface.\u003c/p\u003e \u003cp\u003e[Insert Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e]\u003c/p\u003e \u003cp\u003eTheoretically, States 2 and 3 can be regarded as neutral in the context of drought analysis as precursors to the onset of drought, which is clearly marked by State 4. Drought indices that account for both precipitation and evaporation, particularly for evaporative demand, such as SPEI, are anticipated to classify these states effectively. However, some studies have pointed out that SPEI may be insufficient in a full consideration of the complexities of evaporation dynamics within this framework (Yao et al., \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Won et al., \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), leading to higher correlation with SPI compared to EDDI (Xiao et al., \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). To evaluate the relationships among drought indices, we aggregated index values from all weather stations across South Korea into single time series vectors for each index. Pearson correlation coefficients were then computed between these vectors to assess the linear association among SPI, SPEI, EDDI, and SPEDI on a national scale, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e7\u003c/span\u003e. Notably, SPI and SPEI demonstrate a strong correlation, whereas EDDI exhibits a lower correlation with SPI and SPEI. Interestingly, SPEDI exhibits relatively high correlations with all drought indices. These suggest that SPEI may not fully account for evaporative trends, leading to its lower correlation with EDDI. Additionally, the high correlations near one between SPEI and SPI indicate that SPEI may not effectively complement the limitation of SPI.\u003c/p\u003e \u003cp\u003e[Insert Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e7\u003c/span\u003e]\u003c/p\u003e \u003cp\u003eConsidering the matter of insufficient consideration of evaporation in SPEI, we conducted a comparative analysis of the two drought indices, SPEI and SPEDI. While the two indices attempt to render the same information on the land-atmosphere water balance, they employ different computation methods. In the computation of SPEI, the difference between precipitation and potential evapotranspiration (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{ET}_{o}\\)\u003c/span\u003e\u003c/span\u003e) is calculated and then fitted to a probability distribution. Conversely, SPEDI computes the precipitation to potential evapotranspiration difference after individually fitting each component into its respective probability distributions.\u003c/p\u003e \u003cp\u003eIn fitting each factor to the probability distributions, we hypothesized that there would be a difference in the contributions of these factors to the resulting drought index. To validate this hypothesis, we conducted an experimental comparison using annual time series data in Gangwon-do. This region was selected due to its unique precipitation and evaporation patterns that give rise to a clear distinction between the states identified by SPI and EDDI. More importantly, Gangwon-do is a key region for field crop production in South Korea, particularly well-suited for evaluating flash droughts. The frequency and temporal distribution of each state in the Gangwon-do region from 1980 to 2022 are shown in Figs.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e8\u003c/span\u003e and \u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e9\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e[Insert Figs.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e8\u003c/span\u003e\u0026ndash;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e9\u003c/span\u003e]\u003c/p\u003e \u003cp\u003eTo assess the performance of SPEI and SPEDI across the categorized states, Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e10\u003c/span\u003e displays the time series of drought indices for selected states. State 1 was excluded from the Figure due to the absence of drought. In the majority of seasons categorized as States 2 and 3, SPEDI consistently indicated neutral conditions, while the SPEI tracked more closely with the SPI. For instance, in March and April of 2018, which were classified as State 3, SPEI did not account adequately for the increased evaporation effects characteristic of this period. On the other hand, SPEDI accurately reflected these neutral conditions. Furthermore, during periods identified as State 4, SPEI exhibited a reduction in the severity of drought conditions. In contrast, SPEDI more effectively captured instances of extreme drought, offering a more comprehensive representation that considered both precipitation and evaporation conditions. This comparative analysis of SPEI and SPEDI allows us to better understand the distinct contributions of each component to these drought indices. The findings emphasize that SPEI does not fully describe the influence of evaporation. In contrast, SPEDI offers a more comprehensive representation, effectively accounting for the interaction between precipitation and evaporation in the assessment of drought conditions.\u003c/p\u003e \u003cp\u003e[Insert Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e10\u003c/span\u003e]\u003c/p\u003e \u003c/div\u003e"},{"header":"5. Discussion","content":"\u003cp\u003eThis study demonstrates that traditional precipitation-based drought indices, such as SPI and SPEI, are limited in their ability to capture the rapid onset of flash droughts (Zhang et al., \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Parker et al., \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). While EDDI accounts for atmospheric moisture demand via reference evapotranspiration (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{ET}_{0}\\)\u003c/span\u003e\u003c/span\u003e​), it has been shown to generate false positives by ignoring the precipitation component (Hobbins et al., \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Parker et al., \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). These single-variable indices often fail to represent the complex interaction between elevated atmospheric demand and concurrent precipitation deficits that drives flash drought development (Hoell et al., \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eAlthough SPEI attempts to integrate both precipitation and evaporative demand​ (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{ET}_{0}\\)\u003c/span\u003e\u003c/span\u003e), previous studies have found that it tends to underestimate drought severity under conditions of high evaporative demand (Yuan et al., \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2019\u003c/span\u003e, Hoffmann et al., \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). To address this limitation, we proposed a new composite index, SPEDI, that preserves the distinct statistical properties of precipitation and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{ET}_{0}\\)\u003c/span\u003e\u003c/span\u003e, ​by standardizing them separately before integration. This formulation enables SPEDI to better capture the transition from normal to extreme drought conditions, particularly under rapidly intensifying scenarios. Our findings show that SPEDI more effectively delineates flash drought-affected areas, as evidenced by its strong agreement with reported agricultural damage data.\u003c/p\u003e \u003cp\u003eDespite these promising results, several limitations of this study should be acknowledged. First, the validation of flash drought events relied exclusively on agricultural damage records, which were limited to the summer season and to specific years. This restricts the ability to analyze seasonal variation and long-term trends. Second, the study focused solely on agricultural impacts and did not consider flash drought effects on ecosystems, water resources, or urban areas. As a result, our findings and the performance of SPEDI are most applicable to agricultural drought, and caution is advised when generalizing to other sectors without further validation.\u003c/p\u003e \u003cp\u003eAnother limitation concerns the spatial resolution of the input meteorological data and drought indices. While the current resolution is suitable for regional-scale assessment, it may not adequately capture localized drought events or microclimatic effects, especially in mountainous or coastal regions. Using finer-resolution datasets could help detect localized flash droughts more accurately and support more targeted management.\u003c/p\u003e \u003cp\u003eLooking ahead, several directions for future research can strengthen and expand upon this study. First, improving the validation dataset is essential. Future work should incorporate drought impact data from other sectors, such as ecosystems, water resources, and urban areas, and extend the analysis to different seasons. This would allow a more complete evaluation of index performance across a broader range of conditions. It would also be valuable to compare results across various regions and climate zones to assess how the index performs in different settings. Additionally, using higher-resolution meteorological data could improve the precision of flash drought detection, especially at the local scale. This would support the identification of smaller or localized events and help track drought development in areas with complex terrain, including mountainous and coastal regions.\u003c/p\u003e \u003cp\u003eFinally, given the increasing recognition of flash drought as a significant hazard, efforts should continue to bridge research with real-time monitoring. Integrating indices like SPEDI into operational early warning systems, validated against observed impacts, has the potential to significantly enhance drought preparedness and response (Otkin et al., \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Otkin et al., \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2022\u003c/span\u003e)\u003c/p\u003e"},{"header":"6. Conclusion","content":"\u003cp\u003eIn this study, we aimed to enhance our understanding of flash drought patterns by conducting a comprehensive analysis of four drought indices: the Standardized Precipitation Index (SPI), the Standardized Precipitation Evapotranspiration Index (SPEI), the Evaporative Demand Drought Index (EDDI), and a Standardized Precipitation Evaporation Differential Index (SPEDI). Specifically, this study investigated the occurrence of flash droughts in South Korea, with an emphasis on spatiotemporal patterns, assessing the effectiveness of the mentioned drought indices, and identifying the meteorological factors that influence their development. Our systematic analysis within the South Korean peninsula led to the following conclusions.\u003c/p\u003e \u003cp\u003e \u003col\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eThis study emphasized the limitations of traditional drought indices such as SPI and SPEI in capturing the full extent of flash drought impacts, particularly in contrast to the EDDI and SPEDI indices, which demonstrated capability in delineating drought-affected areas and accurately representing the observed patterns of agricultural damage.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eThe temporal analysis further revealed the strengths and weaknesses of each drought index in detecting and characterizing flash drought events, highlighting the importance of both precipitation deficits and temperature-induced evaporative stress. It was evident that EDDI and SPEDI indices were more effective in identifying a broader range of drought events, including those driven by high evaporative demand, and offered a more comprehensive view of flash drought occurrences.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eNotably, the assessment of SPEDI performance in comparison to SPEI illustrated the complex interactions between precipitation and evaporation in the context of drought analysis. The study revealed that SPEI may not fully account for the dynamics of evaporation, leading to a potential underestimation of drought severity under certain conditions. In contrast, SPEDI, by incorporating both precipitation and evaporation components, provided a detailed representation of drought conditions, effectively capturing the transition from normal to severe drought states.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003c/ol\u003e \u003c/p\u003e \u003cp\u003eIn conclusion, the findings of this study highlight the need to employ novel drought indices like SPEDI, which capture both precipitation deficits and evaporative demand. The results demonstrate that SPEDI offers a more balanced and responsive approach to detecting flash droughts compared to traditional indices. These insights gained here can contribute to the development of more effective flash drought monitoring and management strategies. Furthermore, SPEDI shows promise for integration into early warning systems, helping improve the timely detection and response to rapidly developing drought conditions.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eS. Kang: conceptualization, data curation, methodology, software, validation, formal analysis, writing - original draft preparation. H.J. Kim: methodology, validation, formal analysis, writing - original draft preparation. J.H. Lee: writing - review \u0026amp; editing, funding acquisition. H.-C. Yoon: conceptualization, writing - review \u0026amp; editing. H.-H. Kwon: conceptualization, methodology, software, validation, investigation, supervision, writing - review \u0026amp; editing, funding acquisition.\u003c/p\u003e\u003ch2\u003eACKNOWLEDGEMENT\u003c/h2\u003e \u003cp\u003eThis research was supported by a grant (2022-MOIS63-001(RS-2022-ND641011)) of the Cooperative Research Method and Safety Management Technology in National Disaster funded by the Ministry of the Interior and Safety (MOIS, Korea).\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eAbu Arra, A. and E. Şişman (2024). \"Innovative drought classification matrix and acceptable time period for temporal drought evaluation.\" \u003cem\u003eWater Resources Management,\u003c/em\u003e VoL. 38, No. 8: (2811-2833)\u003c/li\u003e\n \u003cli\u003eAllen, R., et al. (2005). \"RL Snyder, The ASCE Standardized Reference Evapotranspiration Equation.\" \u003cem\u003eEnvironmental and Water Resources Institute of the American Society of Civil: Reston, FL, USA,\u003c/em\u003e \u003c/li\u003e\n \u003cli\u003eAllen, R. G., et al. (1998). \"Crop evapotranspiration-Guidelines for computing crop water requirements-FAO Irrigation and drainage paper 56.\" \u003cem\u003eFao, Rome,\u003c/em\u003e VoL. 300, No. 9: (D05109)\u003c/li\u003e\n \u003cli\u003eAnderson, M. C., et al. (2013). \"An intercomparison of drought indicators based on thermal remote sensing and NLDAS-2 simulations with US Drought Monitor classifications.\" \u003cem\u003eJournal of Hydrometeorology,\u003c/em\u003e VoL. 14, No. 4: (1035-1056)\u003c/li\u003e\n \u003cli\u003eArra, A. A., et al. (2024). \"Trends of meteorological and hydrological droughts and associated parameters using innovative approaches.\" \u003cem\u003eJournal of Hydrology,\u003c/em\u003e VoL. 640: (131661)\u003c/li\u003e\n \u003cli\u003eChristian, J. I., et al. (2019). \"A methodology for flash drought identification: Application of flash drought frequency across the United States.\" \u003cem\u003eJournal of Hydrometeorology,\u003c/em\u003e VoL. 20, No. 5: (833-846)\u003c/li\u003e\n \u003cli\u003eFord, T. W. and C. F. Labosier (2017). \"Meteorological conditions associated with the onset of flash drought in the eastern United States.\" \u003cem\u003eAgricultural and Forest Meteorology,\u003c/em\u003e VoL. 247: (414-423)\u003c/li\u003e\n \u003cli\u003eFord, T. W., et al. (2015). \"On the utility of in situ soil moisture observations for flash drought early warning in Oklahoma, USA.\" \u003cem\u003eGeophysical research letters,\u003c/em\u003e VoL. 42, No. 22: (9790-9798)\u003c/li\u003e\n \u003cli\u003eGavilán, P., et al. (2006). \"Regional calibration of Hargreaves equation for estimating reference ET in a semiarid environment.\" \u003cem\u003eAgricultural Water Management,\u003c/em\u003e VoL. 81, No. 3: (257-281)\u003c/li\u003e\n \u003cli\u003eHargreaves, G. H. and Z. A. Samani (1985). \"Reference crop evapotranspiration from temperature.\" \u003cem\u003eApplied engineering in agriculture,\u003c/em\u003e VoL. 1, No. 2: (96-99)\u003c/li\u003e\n \u003cli\u003eHobbins, M. T., et al. (2016). \"The evaporative demand drought index. Part I: Linking drought evolution to variations in evaporative demand.\" \u003cem\u003eJournal of Hydrometeorology,\u003c/em\u003e VoL. 17, No. 6: (1745-1761)\u003c/li\u003e\n \u003cli\u003eHoell, A., et al. (2020). \"Lessons learned from the 2017 flash drought across the US Northern Great Plains and Canadian Prairies.\" \u003cem\u003eBulletin of the American Meteorological Society,\u003c/em\u003e VoL. 101, No. 12: (E2171-E2185)\u003c/li\u003e\n \u003cli\u003eHoffmann, D., et al. (2021). \"Flash drought in CMIP5 models.\" \u003cem\u003eJournal of Hydrometeorology,\u003c/em\u003e VoL. 22, No. 6: (1439-1454)\u003c/li\u003e\n \u003cli\u003eHu, J., et al. (2023). \"The effects of high temperature, drought, and their combined stresses on the photosynthesis and senescence of summer maize.\" \u003cem\u003eAgricultural Water Management,\u003c/em\u003e VoL. 289: (108525)\u003c/li\u003e\n \u003cli\u003eHunt, E. D., et al. (2009). \"The development and evaluation of a soil moisture index.\" \u003cem\u003eInternational Journal of Climatology: A Journal of the Royal Meteorological Society,\u003c/em\u003e VoL. 29, No. 5: (747-759)\u003c/li\u003e\n \u003cli\u003eHunt, E. D., et al. (2014). \"Monitoring the effects of rapid onset of drought on non-irrigated maize with agronomic data and climate-based drought indices.\" \u003cem\u003eAgricultural and Forest Meteorology,\u003c/em\u003e VoL. 191: (1-11)\u003c/li\u003e\n \u003cli\u003eKim, C. J., et al. (2014). \"Analysis of climate change impacts on the spatial and frequency patterns of drought using a potential drought hazard mapping approach.\" \u003cem\u003eInternational Journal of Climatology,\u003c/em\u003e VoL. 34, No. 1: (61-80)\u003c/li\u003e\n \u003cli\u003eKim, H.-J., et al. (2023). \"A novel multi-scale parameter estimation approach to the Hargreaves-Samani equation for estimation of Penman-Monteith reference evapotranspiration.\" \u003cem\u003eAgricultural Water Management,\u003c/em\u003e VoL. 275: (108038)\u003c/li\u003e\n \u003cli\u003eKMA (2017). 2016 Abnormal climate report, Korea Meteorological Administration.\u003c/li\u003e\n \u003cli\u003eKwon, H. H., et al. (2016). \"The unusual 2013–2015 drought in South Korea in the context of a multicentury precipitation record: Inferences from a nonstationary, multivariate, Bayesian copula model.\" \u003cem\u003eGeophysical research letters,\u003c/em\u003e VoL. 43, No. 16: (8534-8544)\u003c/li\u003e\n \u003cli\u003eKwon, M., et al. (2019). \"Spatio‐temporal drought patterns of multiple drought indices based on precipitation and soil moisture: A case study in South Korea.\" \u003cem\u003eInternational Journal of Climatology,\u003c/em\u003e VoL. 39, No. 12: (4669-4687)\u003c/li\u003e\n \u003cli\u003eLee, H.-J., et al. (2021). \"Detection of flash drought using evaporative stress index in South Korea.\" \u003cem\u003eJournal of Korea Water Resources Association,\u003c/em\u003e VoL. 54, No. 8: (577-587)\u003c/li\u003e\n \u003cli\u003eLisonbee, J., et al. (2022). \"Making sense of flash drought: Definitions, indicators, and where we go from here.\" \u003c/li\u003e\n \u003cli\u003eMartınez-Cob, A. and M. Tejero-Juste (2004). \"A wind-based qualitative calibration of the Hargreaves ET0 estimation equation in semiarid regions.\" \u003cem\u003eAgricultural Water Management,\u003c/em\u003e VoL. 64, No. 3: (251-264)\u003c/li\u003e\n \u003cli\u003eMcEvoy, D. J., et al. (2012). \"An evaluation of multiscalar drought indices in Nevada and Eastern California.\" \u003cem\u003eEarth Interactions,\u003c/em\u003e VoL. 16, No. 18: (1-18)\u003c/li\u003e\n \u003cli\u003eMcKee, T. B., et al. (1993). \u003cem\u003eThe relationship of drought frequency and duration to time scales\u003c/em\u003e. Proceedings of the 8th Conference on Applied Climatology, California.\u003c/li\u003e\n \u003cli\u003eMo, K. C. and D. P. Lettenmaier (2016). \"Precipitation deficit flash droughts over the United States.\" \u003cem\u003eJournal of Hydrometeorology,\u003c/em\u003e VoL. 17, No. 4: (1169-1184)\u003c/li\u003e\n \u003cli\u003eMozny, M., et al. (2012). \"Use of a soil moisture network for drought monitoring in the Czech Republic.\" \u003cem\u003eTheoretical and Applied Climatology,\u003c/em\u003e VoL. 107: (99-111)\u003c/li\u003e\n \u003cli\u003eNaveendrakumar, G., et al. (2019). \"South Asian perspective on temperature and rainfall extremes: A review.\" \u003cem\u003eAtmospheric Research,\u003c/em\u003e VoL. 225: (110-120)\u003c/li\u003e\n \u003cli\u003eNoguera, I., et al. (2020). \"Characteristics and trends of flash droughts in Spain, 1961–2018.\" \u003cem\u003eAnnals of the New York Academy of Sciences,\u003c/em\u003e VoL. 1472, No. 1: (155-172)\u003c/li\u003e\n \u003cli\u003eOtkin, J. A., et al. (2013). \"Examining rapid onset drought development using the thermal infrared–based evaporative stress index.\" \u003cem\u003eJournal of Hydrometeorology,\u003c/em\u003e VoL. 14, No. 4: (1057-1074)\u003c/li\u003e\n \u003cli\u003eOtkin, J. A., et al. (2018). \"Flash droughts: A review and assessment of the challenges imposed by rapid-onset droughts in the United States.\" \u003cem\u003eBulletin of the American Meteorological Society,\u003c/em\u003e VoL. 99, No. 5: (911-919)\u003c/li\u003e\n \u003cli\u003eOtkin, J. A., et al. (2022). \"Getting ahead of flash drought: from early warning to early action.\" \u003cem\u003eBulletin of the American Meteorological Society,\u003c/em\u003e VoL. 103, No. 10: (E2188-E2202)\u003c/li\u003e\n \u003cli\u003eParker, T., et al. (2021). \"Flash drought in Australia and its relationship to evaporative demand.\" \u003cem\u003eEnvironmental Research Letters,\u003c/em\u003e VoL. 16, No. 6: (064033)\u003c/li\u003e\n \u003cli\u003ePendergrass, A. G., et al. (2020). \"Flash droughts present a new challenge for subseasonal-to-seasonal prediction.\" \u003cem\u003eNature Climate Change,\u003c/em\u003e VoL. 10, No. 3: (191-199)\u003c/li\u003e\n \u003cli\u003eQin, N., et al. (2023). \"Assessing the drought impact on sugarcane yield based on crop water requirements and standardized precipitation evapotranspiration index.\" \u003cem\u003eAgricultural Water Management,\u003c/em\u003e VoL. 275: (108037)\u003c/li\u003e\n \u003cli\u003eRaziei, T. and L. S. Pereira (2013). \"Estimation of ETo with Hargreaves–Samani and FAO-PM temperature methods for a wide range of climates in Iran.\" \u003cem\u003eAgricultural Water Management,\u003c/em\u003e VoL. 121: (1-18)\u003c/li\u003e\n \u003cli\u003eSamani, Z. (2000). \"Estimating solar radiation and evapotranspiration using minimum climatological data.\" \u003cem\u003eJournal of irrigation and drainage engineering,\u003c/em\u003e VoL. 126, No. 4: (265-267)\u003c/li\u003e\n \u003cli\u003eSvoboda, M., et al. (2002). \"The drought monitor.\" \u003cem\u003eBulletin of the American Meteorological Society,\u003c/em\u003e VoL. 83, No. 8: (1181-1190)\u003c/li\u003e\n \u003cli\u003eValero, J. F. M., et al. (2013). \"Regionalization of the Hargreaves coefficient to estimate long-term reference evapotranspiration series in SE Spain.\" \u003cem\u003eSpanish Journal of Agricultural Research,\u003c/em\u003e VoL. 11, No. 4: (1137-1152)\u003c/li\u003e\n \u003cli\u003eVicente-Serrano, S. M., et al. (2010). \"A multiscalar drought index sensitive to global warming: the standardized precipitation evapotranspiration index.\" \u003cem\u003eJournal of climate,\u003c/em\u003e VoL. 23, No. 7: (1696-1718)\u003c/li\u003e\n \u003cli\u003eWang, L. and X. Yuan (2018). \"Two types of flash drought and their connections with seasonal drought.\" \u003cem\u003eAdvances in Atmospheric Sciences,\u003c/em\u003e VoL. 35, No. 12: (1478-1490)\u003c/li\u003e\n \u003cli\u003eWilhite, D. A. and M. H. Glantz (1985). \"Understanding: the drought phenomenon: the role of definitions.\" \u003cem\u003eWater international,\u003c/em\u003e VoL. 10, No. 3: (111-120)\u003c/li\u003e\n \u003cli\u003eWon, J., et al. (2020). \"Copula-based Joint Drought Index using SPI and EDDI and its application to climate change.\" \u003cem\u003eScience of the Total Environment,\u003c/em\u003e VoL. 744: (140701)\u003c/li\u003e\n \u003cli\u003eXiao, X., et al. (2024). \"Leveraging multisource data for accurate agricultural drought monitoring: A hybrid deep learning model.\" \u003cem\u003eAgricultural Water Management,\u003c/em\u003e VoL. 293: (108692)\u003c/li\u003e\n \u003cli\u003eYang, Y., et al. (2021). \"Evaluation of 18 models for calculating potential evapotranspiration in different climatic zones of China.\" \u003cem\u003eAgricultural Water Management,\u003c/em\u003e VoL. 244: (106545)\u003c/li\u003e\n \u003cli\u003eYao, N., et al. (2018). \"Drought evolution, severity and trends in mainland China over 1961–2013.\" \u003cem\u003eScience of the Total Environment,\u003c/em\u003e VoL. 616: (73-89)\u003c/li\u003e\n \u003cli\u003eYuan, X., et al. (2019). \"Anthropogenic shift towards higher risk of flash drought over China.\" \u003cem\u003eNature communications,\u003c/em\u003e VoL. 10, No. 1: (4661)\u003c/li\u003e\n \u003cli\u003eZhang, Y., et al. (2022). \"Changes and driving factors of compound agricultural droughts and hot events in eastern China.\" \u003cem\u003eAgricultural Water Management,\u003c/em\u003e VoL. 263: (107485)\u003c/li\u003e\n \u003cli\u003eZhang, Y., et al. (2022). \"A new multi-variable integrated framework for identifying flash drought in the Loess Plateau and Qinling Mountains regions of China.\" \u003cem\u003eAgricultural Water Management,\u003c/em\u003e VoL. 265: (107544)\u003c/li\u003e\n \u003cli\u003eZhang, Y., et al. (2017). \"Flash droughts in a typical humid and subtropical basin: A case study in the Gan River Basin, China.\" \u003cem\u003eJournal of Hydrology,\u003c/em\u003e VoL. 551: (162-176)\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"stochastic-environmental-research-and-risk-assessment","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"serr","sideBox":"Learn more about [Stochastic Environmental Research and Risk Assessment](https://www.springer.com/journal/477)","snPcode":"477","submissionUrl":"https://submission.nature.com/new-submission/477/3","title":"Stochastic Environmental Research and Risk Assessment","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Flash Drought, SPEDI, Drought Detection, EDDI, SPEI, and SPI","lastPublishedDoi":"10.21203/rs.3.rs-4405968/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4405968/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThis study introduces the Standardized Precipitation Evaporation Differential Index (SPEDI), a new composite drought index designed to better capture flash drought conditions by accounting for both precipitation deficits and evaporative demand. The performance of SPEDI is compared with three established indices\u0026mdash;the Standardized Precipitation Index (SPI), the Standardized Precipitation Evapotranspiration Index (SPEI), and the Evaporative Demand Drought Index (EDDI)\u0026mdash;in detecting and characterizing flash droughts across South Korea. To evaluate their effectiveness, we analyzed spatial and temporal patterns of flash droughts and compared outputs from each index with records of agricultural damage during four representative years (2017, 2018, 2019, and 2022). Results show that while SPI and SPEI often fail to capture the rapid onset and intensity of flash droughts, EDDI and SPEDI more reliably reflect observed impacts. In particular, SPEDI demonstrated the highest agreement with actual crop damage, offering early warnings in 12 out of 14 events and accurate timing in 10 out of 14 cases. A simplified model examining the relationship between precipitation and evaporative demand further supports SPEDI\u0026rsquo;s improved ability to represent water balance conditions, especially under high evaporative stress, where SPEI tends to fall short. These results suggest that SPEDI offers a more accurate and practical approach for identifying flash droughts, with clear potential for use in early warning systems and drought risk management. This study presents SPEDI as a valuable tool for supporting drought response planning and improving understanding of flash drought behavior in a changing climate.\u003c/p\u003e","manuscriptTitle":"Introducing a Novel Standardized Precipitation Evaporation Differential Index (SPEDI) for Improved Flash Drought Detection and Assessment: A Case Study in South Korea","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-04-23 11:15:01","doi":"10.21203/rs.3.rs-4405968/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-07-27T10:18:41+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-04-22T01:01:03+00:00","index":"","fulltext":""},{"type":"submitted","content":"Stochastic Environmental Research and Risk Assessment","date":"2025-04-19T16:48:56+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"stochastic-environmental-research-and-risk-assessment","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"serr","sideBox":"Learn more about [Stochastic Environmental Research and Risk Assessment](https://www.springer.com/journal/477)","snPcode":"477","submissionUrl":"https://submission.nature.com/new-submission/477/3","title":"Stochastic Environmental Research and Risk Assessment","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"7449bab3-4336-417d-b2ab-a602bbc96aa0","owner":[],"postedDate":"April 23rd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-09-15T15:59:55+00:00","versionOfRecord":{"articleIdentity":"rs-4405968","link":"https://doi.org/10.1007/s00477-025-03092-z","journal":{"identity":"stochastic-environmental-research-and-risk-assessment","isVorOnly":false,"title":"Stochastic Environmental Research and Risk Assessment"},"publishedOn":"2025-09-09 15:57:01","publishedOnDateReadable":"September 9th, 2025"},"versionCreatedAt":"2025-04-23 11:15:01","video":"","vorDoi":"10.1007/s00477-025-03092-z","vorDoiUrl":"https://doi.org/10.1007/s00477-025-03092-z","workflowStages":[]},"version":"v1","identity":"rs-4405968","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4405968","identity":"rs-4405968","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}