Analysis of Meteorological–Hydrological Drought Propagation for the Development of a Drought Early Warning System in the Special Region of Yogyakarta (DIY): A Case Study of the 2023 El Niño Event

International Journal of Current Science Research and Review ISSN: 2581-8341 Volume 08 Issue 12 December 2025 DOI: 10.47191/ijcsrr/V8-i12-52, Impact Factor: 8.048 IJCSRR @ 2025 www.ijcsrr.org 6414 *Corresponding Author: Yana Arifin Volume 08 Issue 12 December 2025 Available at: www.ijcsrr.org Page No. 6414-6426 Analysis of Meteorological–Hydrological Drought Propagation for the Development of a Drought Early Warning System in the Special Region of Yogyakarta (DIY): A Case Study of the 2023 El Niño Event Yana Arifin1, Yohana Noradika Maharani2, Tedy Agung Cahyadi3 1Master’s Program in Disaster Management, UPN “Veteran” Yogyakarta, Indonesia 2Master’s Program in Disaster Management, UPN “Veteran” Yogyakarta, Indonesia 3Department of Geological Engineering, UPN “Veteran” Yogyakarta, Indonesia ABSTRACT: The drought that affected the Special Region of Yogyakarta (DIY) as a consequence of the 2023 El Niño event was characterized by prolonged dry conditions resulting from exceptionally low rainfall, depletion of water resources, and significant impacts on agriculture, dams, reservoirs, as well as the socio-economic conditions of local communities. This study aims to examine, in a spatiotemporal framework, the influence of El Niño on drought conditions across the DIY Province and to analyze the stages and propagation of meteorological, agricultural, and hydrological drought. The analysis uses rainfall observations from 122 rain gauges and three BMKG stations. Drought assessment begins with rainfall deficit analysis, followed by meteorological drought identification using the Standardized Precipitation Index (SPI), agricultural drought assessment using the groundwater availability index (KAT), and hydrological drought assessment using the Hydrological Drought Index (HDI/IKH). The results show that meteorological drought was first detected in April 2023 in Bantul, Gunungkidul, and Kulon Progo Regencies, then expanded toward the central and northern parts of DIY in May 2023. Agricultural drought emerged in June 2023 in the same three regencies, indicated by KAT values falling below 40% of field capacity. The drought reached its peak in October 2023 as hydrological drought, when reservoir volume availability at Q80 dropped below 50%. Meteorological drought appeared 1–2 months after the onset of rainfall deficit (dry season), progressed into agricultural drought with a lag of 2–3 months, and culminated in hydrological drought with a lag of 5–6 months. The development of a regional Integrated Drought Early Warning System should incorporate monitoring and forecasting outputs, enabling drought alerts to be issued 3–6 months before critical conditions arise. KEYWORDS: 2023 El Niño, agricultural drought, hydrological drought, meteorological drought, early warning system, drought propagation. INTRODUCTION In 2023, the Special Region of Yogyakarta (DIY) experienced an extreme drought triggered by the El Niño phenomenon. The Regional Disaster Management Agency (BPBD) of DIY reported that the event affected 100 hamlets, damaged 378 hectares of agricultural land, and left 8,316 residents with limited access to clean water (BPBD, 2023). The BMKG Dry Day Count (HTH) map indicated that parts of Java, including DIY, experienced more than 90 consecutive dry days in October 2023—classified as an extreme HTH category (BMKG, 2023). These conditions spurred a multidimensional crisis, including water scarcity, desiccated agricultural land, declining crop yields due to crop failure, and increased potential for social conflict driven by competition over water resources (Aprianto, 2023; Nugroho & Wahyuni, 2023). Field observations in DIY showed cracked soils, dried-up wells, and widespread reliance on water-tank distribution. According to the World Meteorological Organization (WMO, 2012), drought is categorized into four interrelated types: (1) meteorological drought (rainfall deficit), (2) agricultural/agronomic drought (insufficient soil moisture and groundwater availability for crops), (3) hydrological drought (reduced river discharge, reservoir storage, and groundwater levels), and (4) socio-economic drought. These types typically occur sequentially, beginning with meteorological drought as the initial trigger, followed by impacts on agriculture, then hydrological stress, and ultimately socio-economic crises (Mishra & Singh, 2010; Wilhite, 2010). Drought in DIY is not an isolated event but reflects structural vulnerability exacerbated by climate change. The region’s heterogeneous topography—from Mount Merapi in the north to the karst hills of southern DIY—relies heavily on seasonal rainfall International Journal of Current Science Research and Review ISSN: 2581-8341 Volume 08 Issue 12 December 2025 DOI: 10.47191/ijcsrr/V8-i12-52, Impact Factor: 8.048 IJCSRR @ 2025 www.ijcsrr.org 6415 *Corresponding Author: Yana Arifin Volume 08 Issue 12 December 2025 Available at: www.ijcsrr.org Page No. 6414-6426 patterns that have become increasingly erratic (BPS, 2024). El Niño, characterized by anomalous warming of sea surface temperatures in the central and eastern Pacific Ocean, disrupts atmospheric circulation and leads to substantial reductions in rainfall across Indonesia (Qian et al., 2023; Aldrian & Susanto, 2023). In 2023, El Niño was intensified by a positive Indian Ocean Dipole (IOD), further amplifying dry conditions in southern Indonesia, including DIY (Kurniadi et al., 2023; BMKG, 2023). BPBD DIY recorded 2,256 drought-related incidents throughout 2023 (Kompas, 2023), with impacts on 100 hamlets, 378 hectares of damaged farmland, and an increase in affected residents from 4,150 to 8,316 people. Reports indicate that prolonged drought in several villages across DIY led to severe water shortages (BPBD, 2023). Consequently, the provincial government allocated IDR 80,000,000 to address the clean-water crisis. At present, sectoral drought monitoring systems are routinely implemented by various agencies: the Meteorology, Climatology, and Geophysics Agency (BMKG) monitors rainfall anomalies and meteorological drought; the Ministry of Agriculture focuses on agronomic planting calendars; the Ministry of Public Works and Housing (PUPR) monitors hydrological conditions; and the National Disaster Management Agency (BNPB) addresses water-related crises (BMKG, 2023). BMKG provides early warnings for extreme climate conditions, the Ministry of Agriculture operates a drought and flood monitoring system for agricultural impacts, and the Ministry of Public Works maintains reservoir and dam monitoring systems targeting hydrological drought. However, these systems operate independently and remain unintegrated within a unified early warning platform, causing early warning information to be delayed or insufficiently aligned with local needs (BNPB, 2015). Addressing this gap, the present study proposes the development of an Integrated Drought Early Warning System using an interdisciplinary approach that combines climatological, hydrological, and agricultural analyses. Methodologically, the study applies the Standardized Precipitation Index (SPI) (McKee et al., 1993), the Soil Water Availability Index (KAT/ATi), and the Hydrological Drought Index (HDI/IKH) based on reservoir discharge. Spatial integration is conducted using Geographic Information Systems (GIS), while drought propagation is examined through cross-correlation and lag-time analysis. This approach aligns with recommendations from UNDRR (2022) and the IPCC AR6 (2022), which highlight the need for multi-indicator early warning systems. METHOD This study employs a quantitative approach to analyze the onset and progression of meteorological, agricultural, and hydrological drought during the 2023 El Niño event in the Special Region of Yogyakarta (DIY). Secondary data were obtained from reliable sources, including monthly rainfall records for 2023 from 122 rain gauges and three BMKG stations, soil water availability data (KAT/ATi) as an indicator of agricultural drought, and discharge and storage data from the Sermo Reservoir for hydrological drought assessment. Meteorological drought was analyzed using the monthly Standardized Precipitation Index (SPI-1) (McKee et al., 1993), standardized by WMO (2016). Agricultural drought was assessed through a land-water balance approach using the Soil Water Availability Index (KAT), while hydrological drought was evaluated using the Hydrological Drought Index (HDI/IKH) with a Q80 threshold following Avicenna (2015). Spatial interpolation was conducted using the Inverse Distance Weighting (IDW) method, allowing visualization of drought propagation across the DIY region. Drought propagation time was examined by comparing time series of SPI, KAT, and IKH. The overall workflow—from data collection, drought index computation, and spatial analysis to the development of the early warning system. FINDINGS AND DISCUSSION El Niño in 2023 The El Niño phenomenon is monitored through the ENSO index, which reflects positive (warm) or negative (cool) phases of sea surface temperature conditions in the equatorial Pacific. At the beginning of 2023, the ENSO index was in a Neutral phase (ENSO ±0.5), following NOAA standards. It gradually shifted toward an El Niño phase starting in May 2023, as illustrated in Figure 1. International Journal of Current Science Research and Review ISSN: 2581-8341 Volume 08 Issue 12 December 2025 DOI: 10.47191/ijcsrr/V8-i12-52, Impact Factor: 8.048 IJCSRR @ 2025 www.ijcsrr.org 6416 *Corresponding Author: Yana Arifin Volume 08 Issue 12 December 2025 Available at: www.ijcsrr.org Page No. 6414-6426 Figure 1. Monthly ENSO Index for 2023 El Niño continued to intensify, indicated by a steady increase in the monthly ENSO index from June to October 2023. The event reached its peak in November 2023 and gradually weakened from December onward, returning to Neutral conditions by April of the following year. At the same time, the Indian Ocean Dipole (IOD) was also active, beginning in July 2023 and peaking in November and December 2023. The evolution of the IOD index is shown in Figure 2 below. Figure 2. Monthly IOD Index for 2023 Impacts of the 2023 El Niño on the Climate of DIY The El Niño phenomenon generally affects Indonesia, including the Special Region of Yogyakarta (DIY), by reducing rainfall and increasing air temperature. El Niño can also delay the onset of the rainy season in DIY, resulting in lower rainfall totals and heightened drought risk. The spatial distribution of low rainfall throughout 2023 is shown in Figure 3. International Journal of Current Science Research and Review ISSN: 2581-8341 Volume 08 Issue 12 December 2025 DOI: 10.47191/ijcsrr/V8-i12-52, Impact Factor: 8.048 IJCSRR @ 2025 www.ijcsrr.org 6417 *Corresponding Author: Yana Arifin Volume 08 Issue 12 December 2025 Available at: www.ijcsrr.org Page No. 6414-6426 Figure 3. Monthly Rainfall Map of DIY in 2023 Low rainfall was first observed in Kulon Progo and Bantul starting in April 2023. By May, areas with low rainfall (brown shading) expanded and became more widespread. Spatial patterns show that rainfall deficits initially emerged in the eastern and southern parts of the region. In June and July, the reduction in rainfall intensified, spreading from southern DIY toward central, northern, and western areas. The most severe rainfall deficits occurred during August, September, and October, when monthly rainfall generally fell below 20 mm. Even in November and December 2023, rainfall across DIY remained low, ranging from only 20–50 mm per month. Meteorological Drought A significant reduction in monthly rainfall can trigger meteorological drought as measured using the Standardized Precipitation Index (SPI), consistent with findings by McKee et al. (1993), McKee (1997), Purnamasari (2017), Andika (2014), and Muliawan (2012). During the 2023 El Niño event, the prolonged dry season (Supari et al., 2018) also contributed to sectoral drought, beginning with meteorological drought (Surmaini, 2015; Van Loon, 2015). The analysis of the SPI throughout 2023 reveals spatial and temporal fluctuations in drought patterns across the study area. Figure 5 presents SPI-1 (monthly scale) used to identify monthly meteorological drought conditions. Monthly meteorological drought analysis for DIY is presented temporally in Figure 5. The results indicate that meteorological drought began to develop in April 2023, although initially only in Gunungkidul Regency. By June 2023, drought conditions intensified and expanded, reaching moderate levels (brown shading). The drought further escalated to dry (orange) and very dry (red) conditions during September–October, and persisted in the “very dry” category throughout November and December 2023 across nearly all regencies in DIY, as shown in Figure 4. Figure 4. Monthly Meteorological Drought (SPI) Map for DIY in 2023 International Journal of Current Science Research and Review ISSN: 2581-8341 Volume 08 Issue 12 December 2025 DOI: 10.47191/ijcsrr/V8-i12-52, Impact Factor: 8.048 IJCSRR @ 2025 www.ijcsrr.org 6418 *Corresponding Author: Yana Arifin Volume 08 Issue 12 December 2025 Available at: www.ijcsrr.org Page No. 6414-6426 Agricultural Drought The 2023 El Niño not only triggered meteorological drought but, when prolonged, also led to agricultural drought, affecting crops—particularly rice—consistent with Surmaini et al. (2015) and Irawan et al. (2025). Agricultural drought was first observed in August 2023, indicated by soil water availability (KAT) dropping below 40% of field capacity (Figure 4). The analysis of agricultural drought based on KAT indicates that in June 2023, soil water availability fell into the “Moderate” category in only two regencies— Kulon Progo and Gunungkidul—while the other regencies had already reached the “Low” category. By July 2023, nearly all regencies in DIY were classified as “Low” to “Very Low,” as shown in Figure 4. Agricultural drought typically emerges after the onset of meteorological drought (Boken, 2005; Irawan et al., 2023), supporting the use of meteorological indicators as early warnings for agricultural drought (Surmaini et al., 2015). Spatial and temporal analysis of the KAT index shows that water availability for vegetation in DIY had already declined to moderate-to-low levels by June 2023. This aligns with the regional climate, as August–October marks the peak dry season when rainfall is extremely low, with more than 90 consecutive dry days recorded. Rainfall remained very low during November and December, contributing to the expansion and intensification of agricultural drought. Initially observed in western and northern DIY, the drought gradually spread toward central, southern, and eastern areas, becoming most pronounced in the southern and eastern regions by December 2023. Hydrological Drought Hydrological drought is detected through declines in reservoir water levels, river discharge, or lake volume. Monitoring data from the Sermo Reservoir throughout 2023, at both 10-day and monthly scales, indicate a significant decline beginning in late September. This marks the onset of hydrological drought. By October 2023, reservoir volume had fallen below the Q80 threshold (reliable discharge), and continued decreasing to below Q50, indicating the start of hydrological drought warning conditions, as shown in Figure 5. Figure 5. Sermo Reservoir Volume (million m³) Hydrological drought in this study was assessed using the Hydrological Drought Index (HDI/IKH), as shown in Table 2. According to Avicenna (2015) and Affandy et al. (2024), the “Dry” category (IKH 0.000–0.0155) began to appear in June 2023. The “Very Dry” category (–0.0006 to 0.000) emerged in late September. In October 2023, IKH values reached –0.033 to –0.56, significantly below the threshold for “Very Dry” conditions. These negative values persisted until the end of 2023 (Table 1). International Journal of Current Science Research and Review ISSN: 2581-8341 Volume 08 Issue 12 December 2025 DOI: 10.47191/ijcsrr/V8-i12-52, Impact Factor: 8.048 IJCSRR @ 2025 www.ijcsrr.org 6419 *Corresponding Author: Yana Arifin Volume 08 Issue 12 December 2025 Available at: www.ijcsrr.org Page No. 6414-6426 Table 1. Reservoir Discharge and IKH Values Overall, the analysis shows that hydrological drought progressively intensified over time. Hydrological drought was first detected in June 2023 with an IKH of –0.09, then further declined to –0.03 and beyond, indicating severe reservoir deficits from late September through December 2023. Drought Propagation: Empirical Evidence of a Tiered Transmission The progression of drought is quantified by correlating monthly indicators of meteorological drought (SPI), agricultural drought (KAT), and hydrological drought (IKH). This chain of propagation unfolds through a sequence that starts with rainfall deficits, followed by meteorological drought, reduced soil moisture for crops, and ultimately a decline in river discharge or reservoir/lake volume. This pattern aligns with previous studies (Budianto et al., 2020; Sutanto et al., 2024). Although such conditions can escalate into socio-economic drought, that aspect is not covered in this research. Figure 6. Drought Propagation in the Special Region of Yogyakarta (DIY) International Journal of Current Science Research and Review ISSN: 2581-8341 Volume 08 Issue 12 December 2025 DOI: 10.47191/ijcsrr/V8-i12-52, Impact Factor: 8.048 IJCSRR @ 2025 www.ijcsrr.org 6420 *Corresponding Author: Yana Arifin Volume 08 Issue 12 December 2025 Available at: www.ijcsrr.org Page No. 6414-6426 Temporal propagation is assessed by comparing monthly drought indices in tabular form, while spatial propagation is identified through comparisons of monthly drought maps throughout 2023, as shown in Figure 8. The findings reveal that drought phenomena interact across stages—beginning with atmospheric anomalies (Zhou et al., 2025), followed by rainfall anomalies, leading to meteorological drought (Surmaini et al., 2015; Avicenna et al., 2015), and further advancing into hydrological drought as the cycle continues (Budianto et al., 2020; Irawan et al., 2023). The sequence of drought development, as summarized in Table 2, begins with the onset of the dry season in the Special Region of Yogyakarta (DIY). Table 2. Lag Time of Drought Propagation Low rainfall was first observed in April 2023. This early deficit set off the initial stage of drought progression. The reduced rainfall in April was followed by the emergence of meteorological drought (SPI) in May 2023, classified as Normal to Dry in two regencies. By June 2023—around two months after the onset of the dry season—meteorological drought intensified, with the Very Dry category spreading across all regencies in the region. The prolonged rainfall deficit gradually reduced soil moisture, reflected in declining soil water availability (KAT) recorded in July 2023. This marks the onset of agricultural drought, which appeared three months after the beginning of the dry season and about two months after meteorological drought began to intensify. Hydrological drought was assessed using the water volume of Sermo Reservoir in Kulon Progo as a representative indicator of regional hydrological conditions. The reservoir volume, initially around 19,173,383 m³ between January and April 2023, dropped to approximately 11,298,257 m³ by October 2023. This level is far below the Q80 threshold of 17,718,774 m³, consistent with the calculated Hydrological Drought Index (IKH) for October 2023. These findings demonstrate a clear onward propagation of drought— temporally showing a 2–3 month delay between stages and extending to 5–6 months from the first rainfall deficit in April. This pattern aligns with earlier research (Sutanto et al., 2024; Budianto et al., 2020), indicating that each phase of drought requires a lag time before influencing the next. When the deficits persist for an extended period, they culminate in a prolonged drought disaster. A comparative analysis of SPI, KAT, and IKH illustrates a systematic drought progression during the 2023 El Niño event in the Special Region of Yogyakarta. As shown in Figure 7, the sequence unfolded as follows: April 2023: Start of the dry season marked by low rainfall June 2023: SPI-1 ≤ –1.5, indicating meteorological drought August 2023: KAT ≤ 40%, signaling the onset of agricultural drought October 2023: IKH ≤ –0.006 and reservoir volume < 50% of normal, marking hydrological drought The empirical lag time between each drought stage is 2–3 months from meteorological to agricultural drought, and another 2– 3 months before hydrological drought emerges—totaling roughly 5–6 months from the onset of the dry season. These findings parallel the conclusions of Sutanto et al. (2024) and Zhou et al. (2025). Integrated Early Warning System Based on Drought Propagation Building on the drought analysis conducted in the Special Region of Yogyakarta, particularly during the El Niño year, drought disasters can essentially be anticipated 3 to 6 months before they peak. By combining historical drought records with climate forecasts for the next several months, an integrated Drought Early Warning System (DEWS) can be developed. The foundation of such a system International Journal of Current Science Research and Review ISSN: 2581-8341 Volume 08 Issue 12 December 2025 DOI: 10.47191/ijcsrr/V8-i12-52, Impact Factor: 8.048 IJCSRR @ 2025 www.ijcsrr.org 6421 *Corresponding Author: Yana Arifin Volume 08 Issue 12 December 2025 Available at: www.ijcsrr.org Page No. 6414-6426 lies not only in real-time monitoring but also in establishing reliable meteorological drought prediction capabilities. Near real-time data, along with mediumto long-range forecasts (3–7 months ahead), is crucial for generating timely warnings. This study formulates an integrated drought early warning system designed to produce proactive, tiered, and time-based responses, illustrated in Figure 7. Figure 7. Integrated Regional/National Drought Early Warning System The proposed system is structured to take advantage of a 6-month predictive lead time to build three interconnected warning levels, each with its own triggers, target stakeholders, and mitigation actions tailored to the regional context of Yogyakarta. The findings establish a scientific basis showing that an Integrated Early Warning System can fully utilize lag time to deliver proactive, coordinated, and sequential response strategies aligned with the drought stages. Constructing the Integrated Drought Early Warning System (DEWS) Developing a National–Regional DEWS requires strong intersectoral collaboration at the national scale. The formation of DEWS involves coordinated work across several national and regional priorities, including sectoral engagement, stakeholder participation, public–private partnerships, and collaboration between central and local governments. Observation and prediction stages play an essential role in planning and responding to drought events. With nationwide observation networks, sector-specific drought indicators can be monitored and predicted. Combining real-time observations with forecasts makes it possible to generate timely drought warnings. Dissemination of the warnings is carried out through official channels to relevant sectors such as Agriculture, Public Works and Water Resources, and Disaster Management agencies (BNPB/BPBD). Each sector can then implement its own early mitigation actions, from institutional coordination to community-based response and public information through websites or social media. Because there is a clear lag time between meteorological, agricultural, and hydrological drought, this progression forms a dynamic structure that can serve as a scientific backbone for a time-based DEWS. The system consists of three warning levels, each tied to specific triggers and mitigation actions. Rather than simply mapping risk, it offers a proactive, time-sensitive risk management framework that provides up to 6 months of strategic lead time before hydrological crisis emerges. The system demonstrates that drought impacts can be anticipated early, allowing mitigation measures to be carried out more effectively while the region is still in the “alert” stage. The mechanism also incorporates continuous monitoring and evaluation, ensuring the system remains adaptive, dynamic, and responsive to real-time data rather than static alerts. The algorithm developed in this research defines three interconnected warning levels: Warning Level 1 — Meteorological Drought Trigger: SPI-1 ≤ –1.5 sustained for 1–2 months Timing: Detected 1–2 months after the dry season begins International Journal of Current Science Research and Review ISSN: 2581-8341 Volume 08 Issue 12 December 2025 DOI: 10.47191/ijcsrr/V8-i12-52, Impact Factor: 8.048 IJCSRR @ 2025 www.ijcsrr.org 6422 *Corresponding Author: Yana Arifin Volume 08 Issue 12 December 2025 Available at: www.ijcsrr.org Page No. 6414-6426 Purpose: Early alert phase Recommended actions: Outreach to farmers regarding early drought risk, Preparation of irrigation water reserves, Adjustment of planting schedules. These actions are preventive, aimed at reducing potential impacts. Warning Level 2 — Agricultural Drought Trigger: Soil Water Availability (KAT) < 40% of field capacity Timing: Detected around 2–3 months after the dry season starts Purpose: Critical phase for agriculture Recommended actions: Promotion of drought-tolerant crop varieties, Use of high-yield seeds, Implementation of water-saving practices (mulching, drip irrigation). These actions are corrective, meant to minimize yield loss and maintain land productivity. Warning Level 3 — Hydrological Drought Trigger: IKH ≤ –0.0006 and river/ reservoir discharge < 50% of Q80 Example: October 2023 in Yogyakarta Purpose: Crisis-phase action Recommended actions: Emergency water distribution to affected villages, Restrictions on non-essential water use. These actions are responsive, aimed at protecting lives and ensuring basic needs. The system is designed to streamline coordination among agencies such as BMKG, the Agriculture Office, PUPR, BNPB, and BPBD. Through a unified national platform, each institution can collaborate efficiently under an Integrated Drought Early Warning System, ensuring that early warnings translate into timely, actionable mitigation at every level—from national policymakers to local communities. Figure 8. Flowchart of the Drought Early Warning System