Physical Society of Indonesia Articles Journal of the Physical Society of Indonesia 2.
, 25-33 .
https://doi.
org/10.
35895/jpsi.
Multiscale Atmospheric Drivers of the April 2023 Cilacap Flood: Insights from Himawari-9 RGB Imagery and Radiosonde Data Jamrud Aminuddin1*.
Sri Susanti1.
Adnan Dendy Mardika2, and Arief Darmawan3 1Physics Study Program.
Faculty of Mathematics and Natural Sciences.
Jenderal Soedirman University.
Central Java, 53122.
Indonesia 2Meteorology.
Climatology, and Geophysics Agency.
Central Java, 50145.
Indonesia 3Research Center for Climate and Atmosphere.
National Research and Innovation Agency.
Jakarta, 10340.
Indonesia (Received October 11, 2026.
revised January 17, 2026.
accepted January 30, 2026.
published online April 18, 2.
This study analyzes the atmospheric drivers of extreme rainfall during the 27 April 2023 flood event in Cilacap.
Indonesia, using satellite, reanalysis, and in situ data.
Convective evolution was examined using rainfall records.
Himawari-9 cloud-top temperatures, instability indices (CAPE.
CIN.
KI.
TTI), and surface wind fields.
Rainfall exceeded 135 mm from the afternoon to nighttime.
Satellite observations showed rapid convective growth, with cloud-top temperatures dropping below Oe75 AC, indicating mature Cumulonimbus clouds.
The 24-Hour Night Microphysics RGB imagery revealed convective initiation around 08:00 UTC and intensification during the evening.
Atmospheric instability increased prior to the event, with CAPE exceeding 1600 J kgAA between 09:00 and 12:00 UTC, while CIN decreased below 80 J kgAA.
Surface wind analysis identified a strong low-level convergence zone formed by moist onshore flow from the Indian Ocean interacting with inland winds.
This convergence provided dynamical lifting that released the stored instability, triggering deep convection that peaked around 15:00 UTC.
The results show that the extreme rainfall was driven by the coupling of high thermodynamic instability and local moisture convergence, demonstrating the value of combining NWP products and satellite RGB imagery for diagnosing coastal extreme rainfall.
Keywords: Atmosphere.
Clouds.
Extreme.
Himawari.
Instability.
Precipitation This is an open access article under the CC BY-NC license.
Copyright A 2026 by Author.
Published by Physical Society of Indonesia
INTRODUCTION
Indonesia's position as a maritime continent with high solar radiation drives significant evaporation, resulting in high atmospheric moisture content.
These conditions create an environment highly conducive to the formation of deep convective clouds, specifically Cumulonimbus (C.
On 27 April 2023, the Cilacap region experienced an extreme precipitation event exceeding 150 mm, which triggered catastrophic flooding and surpassed the intensity of the historical 1992 flood.
CilacapAos unique coastal geography, situated between the Indian Ocean and the southern Java mountain range, facilitates complex multiscale atmospheric interactions where regional phenomena, such as the Madden-Julian Oscillation (MJO) and the West Monsoon, enhance moisture transport toward the mainland.
Despite the high hydrometeorological risks in this area, a prominent research gap exists, as previous local studies have primarily utilized isolated approaches, relying either exclusively on Numerical Weather Prediction (NWP) methods or satellite-based RGB visualization.
There is a lack of integrated research that combines these methodologies to analyze the gradual evolution of convective systems in the southern coastal region of Java.
Consequently, a more comprehensive diagnosis is required to quantify the relationship between physical atmospheric parameters and the distinct developmental phases of clouds responsible for extreme rainfall.
This study addresses this gap by integrating high-resolution Himawari-9 satellite imagery .
minute temporal and 2-km spatial resolutio.
with atmospheric instability indices and radiosonde *Contact Author: jamrud.
aminuddin@unsoed.
ISSN: 2685- 3841 (Onlin.
2026 The Physical Society of Indonesia J.
Phys.
Soc.
Ind.
, 25-33 .
Articles Aminuddin et al.
observations from the BMKG Cilacap station (ID: 96.
The novelty of this research lies in the synergistic use of the 24-Hour Night Microphysics RGB technique and the quantification of instability parameters such as CAPE.
CIN, and KI to track the Cb cloud life cycle from initiation to maturity.
analyzing surface wind patterns and coastal convergence, this study aims to provide a deep understanding of the vertical lifting mechanisms that triggered the extreme flooding in Cilacap.
METHOD
Data This research employs a multi-source approach, integrating in-situ observational data, satellite remote sensing, and atmospheric reanalysis datasets to diagnose the triggering mechanisms of the extreme flood in Cilacap on April 27, 2023.
Precipitation data were obtained from the Global Satellite Mapping of Precipitation (GSMaP) Gauge-near-real-time (GNRT) type in .
csv format.
This dataset provides hourly rainfall estimates corrected with ground-based rain gauge data to ensure quantitative accuracy during the observation period from 12:00 to 16:30 UTC.
The microphysical characteristics and life cycle of convective clouds were analyzed using Himawari-9 Level 2 satellite imagery, specifically the Infrared (IR) Enhanced and 24-Hour Night Microphysics RGB products.
These satellite data feature a high temporal resolution of 10 minutes and a spatial resolution of 2 km, enabling precise monitoring of cloud phase transitions.
Furthermore, vertical atmospheric profiles were evaluated using radiosonde data from the official BMKG Cilacap station (Station ID: 96.
located at 7.
73A S and 109.
01A E.
The data were retrieved via the University of Wyoming portal for the observation cycles at 00:00 UTC and 12:00 UTC.
To support the analysis of atmospheric fluid dynamics, the European Centre for Medium-Range Weather Forecasts (ECMWF) ERA5 reanalysis dataset was utilized to extract wind vectors .
c, y.
and instability indices across various pressure levels.
Data Processing Data processing was conducted systematically through the integration of several meteorological computing software packages, including Microsoft Excel.
QGIS.
SATAID, and GrADS.
Satellite data visualization using SATAID focused on identifying Cloud Top Temperature (CTT) values and cloud morphological classification.
In this study, a CTT threshold below -60AC was established as an indicator of deep convection.
Cloud type identification during nighttime conditions was performed by applying the 24-Hour Night Microphysics RGB composite technique, which combines brightness temperature differences between infrared channels to distinguish physical cloud characteristics.
The color scheme interpretation used for this classification is presented in detail in Table 1.
Table 1 Interpretation RGB 24-Hour Night Microphysics Color Interpretation Thick clouds with high peaks (C.
Thick clouds at medium altitudes Thick clouds at high levels Dust or sand Low clouds Thermodynamic analysis was performed by calculating a series of atmospheric stability indices to quantify the potential for thunderstorm development.
Key parameters calculated included Convective Available Potential Energy (CAPE).
Convective Inhibition (CIN).
K-Index (KI).
Lifted Index (LI), and the Total Totals Index (TTI) based on standard meteorological formulations.
To ensure data synchronization with global meteorological standards, all time variables are presented in Coordinated Universal Time (UTC).
Finally, air mass lifting mechanisms were analyzed through surface wind convergence patterns at the 925 hPa level to map Low-Level Moisture Transport (LLMT) from the Indian Ocean, which served as the primary driver for the extreme rainfall in the region.
ISSN: 2685- 3841 (Onlin.
2026 The Physical Society of Indonesia J.
Phys.
Soc.
Ind.
, 25-33 .
Articles Aminuddin et al.
RESULTS AND DISCUSSION
Spatio-Temporal Rainfall Distribution The first analysis was conducted on the spatial rainfall distribution to identify the areas most severely affected by the flood.
Based on Figure 1, it is observed that the daily rainfall accumulation on April 27, 2023, was concentrated along the coastal regions of Cilacap.
The highest rainfall values were detected in North Cilacap, reaching 177.
8 mm, while South Cilacap ranged between 160.
8 and 166.
This spatial pattern confirms that the convective cells triggering the flood had an extensive reach covering the entire urban area down to the coastline.
Figure 1 Spatial distribution of daily rainfall over Cilacap Furthermore, to understand the timing of the extreme rainfall, a time-series analysis was performed, as presented in Figure 2.
The graph indicates that the rainfall was not distributed evenly throughout the day but was concentrated during the night.
A very sharp and simultaneous intensity spike is visible across the three primary observation points.
Point 3 recorded the highest peak intensity, exceeding 135 mm at 15:00 UTC.
The temporal synchronization of these peaks across all points suggests that the convective cloud system was a mature mesoscale system that developed uniformly over the Cilacap region, directly triggering an instantaneous rise in surface water discharge.
Figure 2 Time series of rainfall ISSN: 2685- 3841 (Onlin.
2026 The Physical Society of Indonesia J.
Phys.
Soc.
Ind.
, 25-33 .
Articles Aminuddin et al.
Cloud Microphysics and Growth Characteristics The analysis of the convective systemAos evolution was conducted by monitoring changes in the Cloud Top Temperature (CTT) across three observation points Figure 3.
As shown in Figure 3.
and Figure 3.
, a significant cooling trend is observed starting at 08:00 UTC, with temperatures progressively dropping below the -60AC threshold.
Both locations maintained a mature deep convection phase from 10:00 to 16:00 UTC, indicating strong and persistent convective activity throughout the However.
Observation Point 3 Figure 3.
displays a slightly different temporal pattern.
Although convective initiation also began around 08:00 UTC, the mature phase was shorter, lasting only from 10:00 to 13:00 UTC.
Despite these local variations, all three points reached their minimum temperature peaks of approximately -70AC at 15:00 UTC, confirming that a widespread mesoscale convective system covered the entire Cilacap region.
Figure 3 Time series of cloud surface temperature at .
observation point 1, .
observation point 2, and .
observation point 3 The physical characteristics of these clouds were further confirmed through the 24-Hour Night Microphysics RGB satellite imagery in Figure 4.
At 15:00 UTC, the Cilacap region was covered by a dense, dark brown-to-purple hue.
Microphysically, this color scheme interprets the presence of very thick Cumulonimbus (C.
clouds with high-altitude peaks containing large ice crystals.
This specific color pattern distinguishes the Cb system from surrounding mid-level clouds, which appear yellowish.
The spatial distribution of the cloud top temperature over the observation area is presented as a contour visualization in Figure 5.
This visualization shows that the convective cell core with the coldest temperatures .
elow Oe60Ay.
was centered directly over the Cilacap coast.
The use of contours in this figure facilitates the identification of the mesoscale convective system's structure, which covers a vast area, aligning with the previously analyzed rainfall distribution.
ISSN: 2685- 3841 (Onlin.
2026 The Physical Society of Indonesia J.
Phys.
Soc.
Ind.
, 25-33 .
Articles Aminuddin et al.
Figure 4 Satellite imagery with 24-hour Microphysics configuration on April 27, 2023.
Figure 5 Cloud Contours on April 27, 2023 ISSN: 2685- 3841 (Onlin.
2026 The Physical Society of Indonesia J.
Phys.
Soc.
Ind.
, 25-33 .
Articles Aminuddin et al.
Atmospheric Instability Indices and Energy Release The thermodynamic evolution of the atmosphere is analyzed through four key stability indices presented in Figure 6.
The temporal variation of potential energy is illustrated in Figure 6.
, which shows a significant destabilization phase where Convective Available Potential Energy (CAPE) surged to a peak range of 1500Ae1600 J/kg between 09:00 and 12:00 UTC.
This accumulation of potential energy coincided with conditions shown in Figure 6.
, where Convective Inhibition (CIN) values remained sufficiently low .
elow 80 J/k.
, creating a favorable environment for air parcels to rise once triggered.
Crucially, the sharp decline in CAPE observed after 12:00 UTC.
Figure 6.
marks the phase of convective overturning.
This indicates that the stored atmospheric potential energy was effectively consumed to drive the intense updrafts of the mature storm system, temporally correlating with the peak rainfall event.
Complementing this energy analysis.
Figure 6.
and Figure 6.
confirm deep tropospheric saturation and instability, with K-Index (KI) values exceeding 37 and Total Totals Index (TTI) surpassing 44.
5 throughout the event.
CIN (J/k.
Time (UTC) Time (UTC) .
TTI (J/k.
Time (UTC) Time (UTC) .
Figure 6 Atmospheric Instability Indices Charts: .
CAPE, .
CIN, .
K-Index, and .
TTI on April 27, 2023.
The detailed vertical atmospheric profile is observed through the Skew-T Log-P diagram in Figure 7.
Observations at 00:00 UTC show a large positive area.
however, by the 12:00 UTC observation, this area had significantly shrunk, and the CAPE value dropped sharply to 126.
69 J/kg.
This confirms the occurrence of the storm consumption phenomenon, where the potential energy accumulated during the day was fully released and consumed by the thunderstorms to produce the extreme rainfall that persisted throughout the night.
Coastal Convergence and Lifting Mechanisms As the primary triggering mechanism for the aforementioned instability, the surface wind patterns at the 925 hPa level were analyzed in Figure 8.
A significant convergence zone, a meeting of air masses, was identified along the Cilacap coastline.
Sea breezes from the Indian Ocean, carrying high moisture, met with local land breezes, creating a vertical lifting force that pushed the air mass upward ISSN: 2685- 3841 (Onlin.
2026 The Physical Society of Indonesia J.
Phys.
Soc.
Ind.
, 25-33 .
Articles Aminuddin et al.
to condense into clouds.
The continuous moisture supply, further enhanced by MJO phase 4-5 activity, made this region highly productive in generating extreme rainfall.
Figure 7 Profile of the sounding measurement results on the SkewT-logP diagram on April 27, 2023, for 00 UTC and 12 UTC.
Figure 8 Wind Speed Distribution CONCLUSION Based on the diagnosis of atmospheric dynamics regarding the extreme flood event in Cilacap on April 27, 2023, it can be concluded that the disaster was triggered by the development of a highly intense mesoscale convective system.
Atmospheric conditions prior to the event exhibited extreme instability, identified through quantitative thermodynamic parameters exceeding critical thresholds, specifically CAPE values above 1500 J/kg peaking between 09:00 and 12:00 UTC.
CIN below 80 J/kg, and a K-Index exceeding 37.
The development of Cumulonimbus (C.
clouds was detected starting at 08:00 UTC, reaching full maturity at 15:00 UTC with cloud top temperatures (CTT) dropping to -70AC, which temporally aligned with the peak extreme rainfall intensity.
This extreme precipitation was mechanically driven by vertical lifting resulting from coastal convergence at the low-level .
, where the interaction between moisture-rich Indian Ocean sea breezes and local land breezes created a continuous energy supply.
The energy conversion process during the storm was clearly recorded in the Skew-T diagram through the storm consumption phenomenon, marked by a sharp decline in CAPE observed at 12:00 UTC.
This proves a massive release of atmospheric potential energy converted into kinetic energy, resulting in persistent heavy rainfall throughout the coastal region of Cilacap.
ISSN: 2685- 3841 (Onlin.
2026 The Physical Society of Indonesia J.
Phys.
Soc.
Ind.
, 25-33 .
Articles Aminuddin et al.
REFERENCE