s11069 025 07390 1

(1)

RESEARCH LETTER

Received: 3 January 2024 / Accepted: 23 April 2025

Erma Yulihastin

[email protected]

1 Research Center for Climate and Atmosphere, National Research and Innovation Agency, Jakarta, Indonesia

2 Agency for Meteorology, Climatology and Geophysics, Jakarta, Indonesia

3 Graduate Program of Earth Sciences, Bandung Institute of Technology, Bandung, Indonesia

Land effect on propagating convective systems during the heavy rainfall event of the 2002 Jakarta flood, Indonesia

Erma Yulihastin¹ · Danang Eko Nuryanto² · Robi Muharsyah² · Ibnu Fathrio¹ · Narizka Nanda Purwadani^1,3 · Albertus Sulaiman¹

Abstract

The movement direction of propagating convective systems originating from both inland and offshore over the north coast of West Java in Indonesia is determined primarily by the prevailing wind. However, the role of land composition over the western part of the Indonesian Maritime Continent (IMC) is also expected to affect the development of propagating convective systems by enhancing upward motion. This hypothesis is tested using a Weather Research and Forecasting model incorporating convection-permitting with 3 km spatial resolution to simulate the heavy rainfall event during the 2002 Jakarta flood. We addressed the influence of land on the local circulation, particularly in the area surrounding Jakarta, by replacing the inland over the western IMC (96°–119°E, 17°S–0°) with a water body with an altitude of 0 m. We then compared the results of model simulations with and without land. The results show that land-sea difference forcing has a significant role in enhancing upward motion and generates a deep convective cloud in response to the land-based convective system, which then continuously and rapidly propagates offshore due to the cold pool mechanism. This land effect mainly triggers gravity waves and also results in early morning convection over coastal regions.

Keywords Land · Propagating convective system · Early morning convection · Java

1 Introduction

Heavy rainfall over the north coast of West Java has triggered severe floods in Jakarta, Indonesia (Wu et al. 2007, 2013; Trilaksono et al. 2011, 2012; Nuryanto et al. 2019). These extreme events associated with landward -as well as offshore- propagating convective sys-

(2)

tems have previously been found to be controlled by prevailing background winds (Yuli- hastin et al. 2020). However, the effects of land on this propagating convective system are not fully understood. On a regional scale, a previous study suggested that the sea breeze convergence is strengthened by mountain–valley winds, which enhance precipitation over inland Java (Qian 2008).

Moreover, the localized spatial asymmetry of the land over this region produces a hetero- geneous rainfall pattern between southern and northern Java. It implies that rainfall variability occurs on an interannual timescale. The strengthened sea and valley breeze convergence produces large amounts of precipitation over mountainous regions (Qian et al. 2010). How- ever, no previous studies have investigated the effects of land over the western of the Indo- nesia Maritime Continent (96°–119°E, 17°S–0°) on coastal precipitation associated with extreme flooding events in Jakarta.

In the past four decades, the capital city of Jakarta has experienced floods more fre- quently during the active period of the Asian monsoon (January–February). However, a severe flood occurred in Jakarta from 29 January to 10 February 2002, which was reported to be the worst flood in the last three decades (1980–2010) in terms of its impact. Given that offshore and landward propagating convective systems may produce extreme precipitation, this event is an ideal case study to investigate the role of land and physical processes involved in such events using a numerical model.

Interestingly, from 29 to 30 January 2002, satellite data showed lower precipitation over Jakarta compared to ground-based station observations. During the period, a cyclonic vortex that developed over the Indian Ocean near southwest Java generated a predominantly meridional monsoonal flow, resulting in near-surface winds aligned with this orientation across West Java (Figs. S1–3). This prevailing northerly wind significantly impacts the topographically diverse and elevated inland of southern West Java.

To further investigate the role of land in the western Indonesia Maritime Continent (96°–

119°E, 17°S–0°) in modulating precipitation over its north coast (including Jakarta) of West Java, we conducted a numerical modeling study employing experimental scenarios with (CTL) and without (SCE) land using high-resolution (3 km spatial resolution) numerical weather predictions from the Weather Research and Forecasting (WRF). In the following sections, we explain the model setup and experimental design and discuss the modeling results.

2 Materials and methods 2.1 Model set-up

The simulations were conducted using WRF model version 4.1.2 (Skamarock et al. 2008) with one-way nested domains comprising a coarse domain (D01), an inner domain (D02), and an innermost domain (D03), with spatial resolutions of 27, 9, and 3 km, respectively (Fig. 1). The domains have 95 × 70, 109 × 76, and 97 × 91 grid points, respectively. The center of the coarse domain is 8.568˚S and 107.038˚E. Mercator map projections were used for the model domains.

The simulation configuration and physical parameters are summarized in Table 1. The initial and lateral boundary conditions were obtained from the National Center for Environ-

(3)

mental Predictions (NCEP) Final Analysis (FNL) Operational Global Analysis dataset (Kal- nay et al. 1996) and forecast grids, which have a horizontal resolution of 1.0˚ × 1.0˚ at 6 h intervals. The three-day forecasts were evaluated for each Planetary Boundary Layer (PBL) scheme, and the model was initialized at 00:00 UTC (07:00 LST) on 25 January 2002.

To further inspect the heavy rainfall event, we validated the simulation results with ground-based observational data from the Agency for Meteorology, Climatology, and Geo- physics (BMKG) station (Van den Besselaar et al. 2017) and TRMM Multi-Satellite Pre- cipitation Analysis (TMPA) Real Time 3B41RT (TMPA-RT). Rainfall gauge data calibrate the TMPA-RT data and have been used in several extreme rainfall studies over tropical Fig. 1 (a) Simulation domains of the WRF model. (b) The land-use real scheme was used for the control (CTL) and removal of the land with the water body or sea (SCE) simulations

(4)

regions and the Maritime Continent (Yulihastin et al. 2020; Huffman et al. 2007, 2012;

Harris et al. 2007; Liu et al. 2008; Yong et al. 2015). Furthermore, to identify the synoptic conditions related to background winds, we used high-resolution, near-surface ERA5 data from the European Center for Medium-Range Weather Forecasts (ECMWF), with a spatial resolution of 0.25˚ × 0.25˚ (Hersbach et al. 2020).

2.2 Experimental design

To investigate the effect of land-sea composition on a propagating convective system, we used an experimental design (Fig. 1) involving two scenarios: (1) the actual topographic conditions as the control simulation (CTL); and (2) a modified land that is flattened (altitude = 0 m) and replaced with the sea (i.e., a water body), which is the scenario simulation (SCE). The CTL and SCE scenario simulations were applied to all domains, focusing on domain 3 (D03). Both experiments include the outermost domain, which captures the effects of regional processes (e.g., wind circulation and sea surface temperature) on the innermost domain circulation.

In this SCE simulation, the initial condition of the water body temperature is fixed based on the default settings from the WRF model without integrating any temperature updates.

Given that this case study of extreme events is primarily influenced by prevailing winds and synoptic scale phenomena (such as cyclonic vortices and cold surges), we employed a static sea surface temperature throughout the simulation. This approach assumes that there are no significant changes or stratification of temperature within the entire water body during the simulation period.

Table 1 Configuration of the weather and research forecasting model for simulating precipitation over the maritime continent at horizontal resolutions of 27 km (D01), 9 km (D02), and 3 km (D03)

Configuration D01 D02 D03

Horizontal grids 95 × 70 175 × 121 199 × 166

Grid spacing (km) 27 9 3

Cumulus scheme Betts–Miller–Janjic scheme Betts–Miller–Janjic scheme -

Vertical grid 46 layers 46 layers 46 layers

Radiation RRTMG longwave scheme

RRTMG shortwave scheme RRTMG longwave scheme

RRTMG shortwave scheme

Microphysics WDM 5-class scheme WDM 5-class scheme WDM 5-class scheme Surface layer Revised MM5 Monin-Obuk-

hov scheme Revised MM5 Monin-Obuk-

hov scheme Revised MM5 Monin-

Obukhov scheme Land surface Unified Noah land-surface

model Unified Noah land-surface

model Unified Noah land-

surface model Planetary boundary

layer YSU scheme YSU scheme YSU scheme

Initial boundary

condition FNL 1.0 × 1.0 FNL 1.0 × 1.0 FNL 1.0 × 1.0

(5)

3 Result and discussion 3.1 Observed heavy rainfall

Rainfall recorded by five meteorology stations in Jakarta City and surrounding areas (i.e.

Bogor represented by Citeko Station) (Fig. 2a) revealed a significant increase from 26 to 30 January 2002 (Fig. 2b). Notably, the Citeko station was included in this inspection to demonstrate the heavy rainfall recorded in north and south areas over the mainland in Great Jakarta. Moreover, heavy rainfall (almost 150 mm) occurred on 29 January 2002 over the northern coast of Jakarta (Tanjung Priok Station). The rainfall began on the morning of 26 January 2002 over inland areas and tended to propagate rapidly offshore during the following morning (Fig. 2b). However, landward propagation of the convective system also occurred, with a slower phase of propagation from 26 to 28 January 2002. Furthermore, moderate-intensity precipitation over the coastal region became persistent (Fig. 2b).

Of note, the slower landward propagation was related to the presence of a cyclonic vortex that developed over the Indian Ocean close to the coast of southern West Java on 26–27 Jan- uary 2002. The cyclonic vortex caused the strengthening of a predominant northerly wind with a meridional monsoonal flow over West Java (Fig. 2c). This northerly flow was similar

Fig. 2 (a) A map of the BMKG station locations over the Great Jakarta area overlaid with the elevation of the region. (b) Rainfall amounts recorded by BMKG stations during 25–31 January 2002. (c) Composite of the streamlines of near-surface winds (925 hPa) from 25 to 31 January 2002. (d) Hovmöller diagram showing a time–latitude cross-section along 106.68˚E. The black dashed line is the coastline along the north coast of West Java

(6)

to the cold anomaly that was identified as the main cause of the torrential precipitation during the Jakarta flood from 31 January to 2 February 2007 (Wu et al. 2007; Trilaksono et al.

2011). The meridional wind modulated the heavy rainfall (Wu et al. 2007), temperature, and relative humidity, which were associated with a cold anomaly after the Borneo Vortex had dissipated (Trilaksono et al. 2011).

One interesting factor is that the offshore propagation of the convective system was in a direction against the prevailing wind (Fig. 2d). The interaction between the predominant northerly wind and the rugged land of southern West Java might have contributed to the offshore propagation and enhanced convection inland. We further examined this issue with a no land scenario (SCE) over all domains (see Fig. 1) and compared the rainfall results for this scenario with the actual condition (CTL) simulations during 26–29 January 2002.

3.2 Effect of land on propagating convective systems

The presence of a land-based and oceanic convective system produced maximum precipitation on 26 January 2002 during the afternoon and at night, respectively (Fig. 3). In this case, both the land-based convective system and offshore propagation (black arrows) were present on 26 and 27 January 2002 in the CTL (Fig. 3a); however, these features were absent in the SCE (Fig. 3b) simulations.

Interestingly, early morning precipitation over the coastal region also became persistent during 28–30 January in the CTL simulation (Fig. 3a). In contrast, the SCE simulation only

Fig. 3 Hovmöller diagram showing a time–latitude cross-section along 106.86˚E, which represented the north coast of Jakarta, of daily rainfall from 26 to 30 January 2002 for the (a) CTL, (b) SCE, and (c) CTL–SCE simulations. The horizontal red lines denote the north (upper, N) and south (below, S) coastal areas of West Java. The black arrows in Fig. 3c denote the direction of rainfall propagation

(7)

captured the slightly landward propagation of the convective system on 28 January 2002 (Fig. 3b). The comparison between the CTL and SCE simulations (Fig. 3c) highlighted significant differences, particularly in precipitation, revealing substantial amounts over the land and sea during those days, except for the afternoon precipitation observed on 28 Janu- ary 2002.

As we examine the maximum precipitation recorded in the CTL simulation from 26 January to 28 January, it becomes clear that the effects of the land composition are significant, especially given the qualitative agreement between the model outcomes and the observational data. The spatial precipitation patterns observed during this time further validated this consistency (Figs. 4a–c).

The positive potential temperatures observed in the CTL and SCE simulations for West Java (Figs. 4d–f) broadly correspond to the maximum precipitation over the north coastal region exhibited in the previous figures (Figs. 4a–c). However, distinction in humidity lev- els between the northern and southern regions suggest that the rugged topography of the mountainous area has resulted in drier and more humid conditions in the north (Fig. 4f).

This result aligns with a previous finding that the land developed upward and downward motions due to the large-scale circulation such as Madden–Julian oscillation), thus leading to a quasi-stationary convective system near major topographic features over the Maritime Continent (Wu and Hsu 2009).

Moreover, to gain a deeper understanding of the physical mechanisms responsible for the offshore propagation that occurred from 26 to 27 January 2002, we thoroughly analyzed the hourly evolution of several key meteorological variables. These include cloud water vapor, which plays a vital role in precipitation formation; equivalent potential temperature, a mea- sure that helps to assess the stability of the atmosphere; and the meridional vertical wind vector, which gives insight into the direction and intensity of wind flows.

Fig. 5 depicts the insights drawn from this analysis, which illustrates the temporal changes of these variables and their relationship to the observed phenomena. Figures 5a and d demonstrated that strong offshore propagation (the speeds of 5.6 and 8.4 ms^{− 1}) occurred from 13:00 to 00:00 LT over a distance of > 220 km. In contrast, the offshore propagation did not occur in the SCE simulation (Figs. 5b and e). During 07:00–18:00 LT, the convective system was concentrated over the mountainous region, whereas during 19:00–00:00 LT, the system moved offshore (Figs. 5c and f). This strong offshore propagation is consistent with a previous study that found offshore-propagating convection over inland areas reaches a maximum in the daytime (Yulihastin et al. 2021).

In this case study, the land might have triggered the propagating convective system that is evident on 26–27 January 2002 (see Fig. 5) compared to 28–29 January 2002 (Fig. 6).

Figure 6 highlights key characteristics of the convection initiation over coastal areas during the early morning on 29 January 2002, specifically between 01:00 and 06:00 LT, which need further investigation. In this following discussion, we will explore the intricate physical and dynamic processes that influence how terrain affects two specific phenomena: (1) the propagation of convective systems and (2) the development of early morning convection.

To further understand the physical processes that caused the offshore propagation from 26 to 27 January 2002, we analyzed the hourly evolution of cloud water vapor, equivalent potential temperature, and meridional vertical wind vector (Fig. 7). On 26 January, the land- based convective system formed from 13:00 LT over the southern part of the mountainous area and propagated northward to the coastal area in the evening (18:00 LT) (Fig. 7a). This

(8)

Fig. 4 Spatial maps of the CTL, SCE, and CTL–SCE simulation results from 26 to 28 January 2002. (a–c) Daily rainfall (shading) overlain by the near-surface wind average and anomaly at 925 hPa (vectors).

(d–f) Daily equivalent potential temperature anomaly with respect to the five-day average (shading and contour lines; blue and red are negative and positive anomalies, respectively)

(9)

Fig. 5 Temporal evolution of the wind (vectors; vertical component multiplied by a factor of 40), cloud mixing ratio (shading), and equivalent potential temperature anomaly (contours) along 106.68˚E, which are shown as vertical–latitude cross-sections. For simplicity, 343 K was subtracted from the equivalent potential temperature, and blue (red) lines indicate negative (positive) values. The contour intervals are 1.5 K and contours start from − 0.5 K (0.5 K). (a–c) CTL, SCE, and CTL–SCE simulated six-hourly results from 07:00–12:00 LT to 01:00–06:06 LT on 26 January 2002. (d–f) Same as (a–c), but for 27 January 2002. The vertical red lines denote the north (right, N) and south (left, S) coastal areas of West Java. The black arrows indicate the direction of convection propagation with a rough estimation of the speed of propagation

(10)

Fig. 6 Same as Fig. 5, but for (a–c) 28 January 2002 and (d–f) 29 January 2002

(11)

Fig. 7 Same as Fig. 5, but for the hourly evolution from 13:00–00:00 LT on 26 January 2002. The blue contour line over the near-surface (< 1 km) is the cold pool

(12)

land-based convective system was not present in the SCE simulation (Fig. 7b), as depicted by the distinction between CTL to SCE that the convection mainly concentrated over mountain area (Fig. 7c).

Notably, at 20:00 LT (Fig. 7d), the cold pool created by the decaying convective cloud over the mountainous region traveled downward and triggered the formation of a new convective cloud over the coastal region. In this case, the presence of the cold pool is evident from the lower equivalent potential temperature over the near-surface (< 1 km). This is consistent with a previous study that showed offshore-propagating systems are driven by a surface cold pool (Yulihastin et al. 2021, 2022) and the squall line mechanism over the eastern Maritime Continent (Hassim et al. 2016).

This distinctive offshore propagation occurred the following day (Fig. 8). A land-based convective system developed over the mountainous area from 13:00 LT and strengthened until 18:00 LT (Fig. 8a). However, the cold pool that formed from the decay of the convective cloud system appeared at 16:00 LT over the southern mountainous region. The cold pool created a new convective cloud system over the southern mountainous region and forced the existing convective cloud system to move northward. Interestingly, several convective cells connected the mountain and sea regions at 19:00 LT, and then the convective system over the mountainous region diminished at 20:00–00:00 LT (Fig. 8a).

A cold pool that developed from the decaying cloud over the coastal region created a new convective cell over the ocean at 22:00 LT, propagating landward due to the strong northerly background winds (Figs. 8a and c). In this case, we conclude that the interaction between the predominant prevailing wind and land, in association with the cold pool mechanism, controlled the propagation direction of the convective system. However, this was not present for the SCE simulation (Fig. 8b). It should be noted that the strong landward propagation of convective system (~ 15.4 ms^{− 1}) occurred in the midnight from Java Sea to north coastal region.

The land-based convection over south coast of Werst Java slowly moving rapid (~ 15.4 ms^{− 1}) landward propagation of the oceanic convective system triggered a deep convective system over the mountainous region during the early morning of the following day (Fig. 9).

For the SCE simulation, early morning convection also occurred and appears to be strongly connected to the convective system over the ocean.

In such a case, the early morning convection may also be affected by a gravity wave, which was further examined for the following day (Fig. 10). The SCE simulation results (see Figs. 7, 8, 9 and 10) suggest that the sea scenario does not significantly enhance northerly winds over the ocean or produce enhanced precipitation. This result is consistent with a previous study that reported greater surface roughness on a flat island has an influence on a high-wind regime (Wang and Sobel 2017). Thus, forcing caused by the thermal contrast between the land and sea is the main factor (Wang and Sobel 2017) controlling the diurnal cycle of precipitation over the Maritime Continent (Ruppert and Chen 2020).

3.3 Gravity wave effect on early morning convection

Another interesting feature the simulation results can explain is the early morning precipitation over the coastal region on 28–30 January 2002 (recall Fig. 6). A previous study noted that the early morning precipitation peak over the north coast of West Java is strongly related to extreme precipitation (Yulihastin et al. 2020). We found the early morning precipitation

(13)

Fig. 8 Same as Fig. 7, but for 27 January 2002

(14)

peak was prominent on 28–29 January 2002 (Figs. 9 and 10). In the case of 29 January 2002, initial convection over the coastal region occurred suddenly at night (22:00 LT) and rapidly propagated landward (02:00 LT) (Fig. 10a).

On the other hand, for SCE, the convection cells are connected between land and sea and tend to be steady (Fig. 10b). During early morning-to-morning time windows (Fig. 10d), the land-based convection indicates a stationary convection system. The intensification of land-based convection around the mountain did not exist for SCE (Fig. 10c).

The initiation of this convection over the coastal region is difficult to explain solely in terms of the prevailing northerly wind. We found a wave-like structure expressed as a negative equivalent potential temperature anomaly at ~ 500–700 hPa, corresponding to a recent study (Yulihastin et al. 2022). The layered structure of the anomaly was disrupted when initial convection developed at 22:00 LT. This result corresponds to a previous study (Hassim et al. 2016) that suggested the strong prevailing wind could create strong asymmetry between the windward and leeward areas, thus generating gravity waves via a thermal contrast, as well as an upward deviation of the horizontal flow in response to the enhanced diurnal heating mechanism.

Our proposed mechanism for early morning convection and precipitation over the north coast of West Java involving meridionally propagating gravity waves is consistent with recent studies (Yulihastin et al. 2022), and is also supported by previous studies of offshore propagation of convective systems due to gravity waves (Hassim et al. 2016; Wang and Sobel 2017; Ruppert and Chen 2020; Ruppert and Zhang 2019; Wei et al. 2020; Mapes et al. 2003; Wu et al. 2009; Love et al. 2011). Topographically induced gravity waves have a Fig. 9 Same as Fig. 8, but for 01:00–06:00 LT on 28 January 2002

(15)

Fig. 10 Same as Fig. 7, but for 19:00–06:00 LT on 28–29 January 2002

(16)

larger amplitude in the windward direction and descend on leeward slopes (Wang and Sobel 2017). The land produces low-level circulation that is strengthened by the height of the land.

This possible mechanism occurred persistently during the three days of heavy rainfall over the coastal region during the severe 2002 Jakarta flood.

4 Discussion: multiscale interaction and their contribution to coastal rainfall

To further discuss the role of land in obstructing the incoming northerly wind, enhancing moisture convergence, and developing deep convection along with extended diurnal rainfall events from 26 to 29 January 2002, we propose the following detailed mechanism (recall Figs. 5, 6, 7, 8, 9 and 10):

(1) Land topography can decelerate northerly winds, enhance vertical lifting, and generate mountain-valley breeze. The diverse topography of the land, characterized by moun- tains and valleys, plays a crucial role in impeding the flow of northerly winds. This interaction results in a deceleration of these winds, which in turn enhances the vertical lifting of air masses. Additionally, this topographical variation can create localized mountain-valley breezes, where cool air flows down from higher elevations during the evening, further affecting the wind dynamics. However, it should be noted that the stronger the northerly wind, the faster landward rainfall propagation created from 26 to 28 January 2002, and it consistently intensified the early morning coastal rainfall as depicted in CTL simulation (Fig. 11).

(2) Differences in thermal inertia between land and sea can create pressure gradients that generate land-sea breeze, which either oppose or support the northerly wind. In this case study, northerly wind from local-to-synoptic scales was mainly associated with increased activities of cold surge (CS) and CENS, which started from 26 January and reached the maximum on 30 January 2002 (Fig. S4). Figure S4 shows that the CS, which coincided with CENS, influenced the strength of the sea breeze, which is represented by background wind (BW) over the north of West Java. As depicted in Fig. 12, the deep sea breezes are initiated from midnight to morning from the surface level to around 7 km in height. In contrast, the land breeze intensified from the afternoon to nighttime. However, the changing time between sea breeze and land breeze does not appear during 28 to 29 January 2002 from 03:00 LT. Throughout the day, the land breeze decelerated by the strength of CS and CENS (> 10 m s^-1) (Fig. S4).

(3) Developing convective regions can decelerate the northerly wind and enhance upstream moisture convergence. Thus, wind dynamics and moisture convergence influence the location, timing, and propagation direction of convective systems. Here, the multiscale interaction from local-to-synoptic circulation, as mentioned in previous studies (Satiadi et al. 2023; Trismidianto et al. 2024; Saufina et al. 2025) in creating the speed of landward propagation and developing maximum coastal rainfall, is exhibited in Fig. 13. The strengthening of northerly wind influences rapid landward propagation of convective systems, and the maximum rainfall over the northern coastal region results from the

(17)

convergence between sea breeze and mountain breeze. This positive feedback interaction does not exist for SCE simulation (Fig. 13).

This study also demonstrated that the diurnal cycle is vital for triggering convective storms and high local rain rates in the West Java region. This result aligns with recent studies sug- gesting that these diurnal rainfall disturbances can be explained as diurnally phase-locked gravity waves through nocturnal mesoscale convective systems events (Ruppert et al.

2020). To this end, we recommend further investigation into the influence of orography on triggering deep convection through orographic flows, as noted in prior studies. This aspect is essential for generating convective storms (Ruppert and Chen 2020; Ruppert et al. 2020) and enhancing mesoscale circulations. Thus, to demonstrate the interaction between cold pool and gravity wave in developing the deep convective cloud clusters, we use analytical modeling by using wave packets termed Meridional Tropical Waves or Meridional Tropi- Fig. 11 A time-latitude cross section over Jakarta (106.86^oE) during 07:00 LST 26 January 2002 to 07:00 LT 29 January 2002 of hourly rainfall (blue to red shaded contour), daily accumulation rainfall (blue filled bar), and topography (grey filled line) for (a) CTL and (b) SCE model simulated. A vertical red line indicates the northern coast of Jakarta. The red arrow indicates the direction of the landward propagation, and the value represents the speed of rainfall propagation calculated from maximum-to-maximum intensity.

The blue arrow represents offshore propagation

(18)

cal Disturbances (Yang 1991). The interaction of Meridional Tropical Wave Packets with orographic winds is expressed by Figures S5–6.

5 Conclusions

We investigated the role of land on propagating convective systems based on a case study of the heavy rainfall event during 26–29 January 2002 that produced a severe flood in Jakarta.

We used high-resolution (3 km) numerical weather predictions from the WRF model. The simulated rainfall is qualitatively consistent with satellite and station data. In this case, synoptic weather conditions were influenced by a cyclonic vortex off the south coast of West Java, which enhanced the near-surface prevailing northerly wind.

We found that the land over the western Indonesia Maritime Continent significantly enhances upward convective motion over the mountainous region. These results are consistent with a previous study (Ruppert and Chen 2020) that showed diurnal and topographic forcing led to ~ 20% rainfall enhancement, as reflected by offshore-propagating convection Fig. 12 A height-time cross section over north coast of Jakarta (5.9 S; 106.86^oE) during 07:00 LST 26 January 2002 to 07:00 LST 29 January 2002 for meridional wind (shaded) for (a) CTL, (b) SCE, and (c) the difference between both of CTL and SCE (CTL–SCE) model simulated. Positive (negative) values indicated southerly (northerly) wind for (a–b) and increase (decrease) of northerly wind for (c)

(19)

modes. The land drives offshore propagation due to the cold pool mechanism, which is in a direction opposite to the prevailing northerly wind. The cold pool also triggers thunder- storms and allows convective cells to coalesce as well as separate into new cells. With the strengthening of the prevailing northerly winds, the new convective cells tend to propagate landward, leading to maximum precipitation over the coastal and mountainous regions. This cold pool has previously been identified from night-time land surface cooling and a low- level flow anomaly that occurs near the leading edge of the cold anomaly (Ruppert and Zhang 2019).

The land over West Java also controls wave-like activity associated with a meridionally propagating gravity wave. This is in agreement with a previous study (Wei et al. 2020) that showed near-surface cooling induced by enhanced rainfall-generated gravity waves, which then caused an offshore-propagating diurnal cycle of precipitation. Our results demonstrate that disruption of the gravity wave coincides with initial convection over the coastal region in the early morning, which is consistent with a previous study (Yulihastin et al. 2022). This also corresponds to the early morning precipitation peak that is related to extreme precipitation and strong landward and offshore propagation of a convective system. The landward propagation is consistent with a previous study that suggested this results in the timing of peak precipitation being more random (Yulihastin et al. 2020; Ruppert and Zhang 2019).

However, the early morning precipitation peak identified in the present case study was also a feature of the 2007 Jakarta flood, which had a semi-diurnal pattern superimposed on the diurnal cycle, with the largest precipitation peak during the early morning (Trilaksono et al. 2011). Our study suggests that synoptic disturbance associated with cyclonic vortex could modify rainfall diurnal patterns in the mesoscale, which contributed to the enhancement of rainfall intensity related to the severe flood in Jakarta in 2002. This study concluded a chance for the same process to occur again in the future, needing the interaction between Fig. 13 A day-to-day variation from 25 to 30 January 2002 of cold surge, CENS, and background wind (sea breeze) denoted by line, average or rainfall over north coastal region (6–6.5 ^oS, 106.86^oE) indicated by bar graph for CTL (red) and SCE (blue) scenarios

(20)

the land and the sea over the western Indonesia Maritime Continent associated with a similar combination of synoptic and mesoscale dynamics.

Supplementary Information The online version contains supplementary material available at h t t p s : / / d o i . o r g / 1 0 . 1 0 0 7 / s 1 1 0 6 9 - 0 2 5 - 0 7 3 9 0 - 1 .

Acknowledgements The authors also greatly acknowledge BMKG for supporting rainfall station data. This research supports the development of Numerical-based Atmosphere-ocean prediction and Knowledge Using deep Learning Artificial intelligence (NAKULA) for better rainfall prediction in the Indonesia Maritime Continent.

Author contributions Erma Yulihastin was the primary contributor to this study who designing the concept and writing the initial and revised manuscript. Danang Eko Nugroho and Ibnu Fathrio conducted the numerical model simulations. Robi Muharsyah and Narizka Nanda Purwadani contributed to the model valida- tion using station and reanalysis data. Albertus Sulaiman calculated the analytical model. All authors were involved in discussions regarding the paper before and after the review process and read and approved the final manuscript.

Funding This research was supported by the National Research and Innovation Agency (BRIN) under the Research and Innovation for Indonesia Progression in second year (2023–2025) with grant number 57/II.7/

HK/2024 and the Joint Collaboration Research Organization of Electronica and Informatics of Program House Decision Support System Based on Remote Sensing Analysis 2024 grant number 1/III.6/HK/2024.

Data availability The NCEP-FNL, BMKG, TMPA-RT, and ERA5 ECMWF datasets generated during and analyzed in the current study are publicly available in the permanent archive as follows: DOI: h t t p s : / / d o i . o r g / 1 0 . 5 0 6 5 / D 6 M 0 4 3 C 6 (Kalnay et al. 1996), DOI:https://doi.org/10.1175/JCLI-D-16-0575.1 ( B a s s e l a a r et al., 2017), DOI:https://doi.org/10.5067/TRMM/TMPA/3H-E-IR/7 (Huffman et al. 2007, 2012), and DOI:https://

doi.org/10.24381/cds.adbb2d47 (Hersbach et al. 2020), respectively.

Declarations

Competing interests The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. The authors declare the following financial interests/personal relationships which may be considered as potential competing interests.

References

Harris A, Rahman S, Hossain F, Yarborough L, Bagtzoglou AC, Easson G (2007) Satellite-based flood modeling using TRMM-based rainfall products. Sensors 7:3416–3427. https://doi.org/10.3390/s7123416 Hassim MEE, Lane TP, Grabowski WW (2016) The diurnal cycle of rainfall over new Guinea in convection-

permitting WRF simulations. Atmos Chem Phys 16(1):161–175. h t t p s : / / d o i . o r g / 1 0 . 5 1 9 4 / a c p - 1 6 - 1 6 1 - 2 0 1 6

Hersbach H, Bell B, Berrisford P, Hirahara S, Horányi A, Muñoz-Sabater J, Thépaut J (2020) The ERA5 global reanalysis. Q J R Meteorol Soc. https://doi.org/10.1002/qj.3803

Huffman GJ, Bolvin DT, Nelkin EJ, Wolff DB, Adler RF, Gu G, Stocker EF (2007) The TRMM multisatellite precipitation analysis (TMPA): Quasi-Global, multiyear, Combined-Sensor precipitation estimates at fine scales. J Hydrometeorol 8(1):38–55. https://doi.org/10.1175/jhm560.1

Huffman GJ, Adler RF, Bolvin DT, Nelkin EJ (2012) The TRMM Multi-Satellite precipitation analysis (TMPA) (Chap. 1). In: Hossain F, Gebremichael M (eds) Satellite rainfall applications for surface hydrology. Springer Netherlands, Dordrecht, pp 3–22. https://doi.org/10.1007/978-90-481-2915-7_1 Kalnay E, Kanamitsu M, Kistler R, Collins W, Deaven D, Gandin L, Joseph D (1996) The NCEP/NCAR

40-Year reanalysis project. Bull Am Meteorol Soc 77(3):437–471. h t t p s : / / d o i . o r g / 1 0 . 1 1 7 5 / 1 5 2 0 - 0 4 7 7 ( 1 9 9 6 ) 0 7 7 - 0 4 3 7 : t n y r p - 2 . 0 . c o ; 2

Liu C, Moncrieff MW, Tuttle JD (2008) A note on propagating rainfall episodes over the Bay of Bengal. Q J R Meteorol Soc 134(632):787–792. https://doi.org/10.1002/qj.246

(21)

Love BS, Matthews AJ, Lister GMS (2011) The diurnal cycle of precipitation over the maritime continent in a high-resolution atmospheric model. Q J R Meteorol Soc 137(657):934–947. h t t p s : / / d o i . o r g / 1 0 . 1 0 0 2 / q j . 8 0 9

Mapes BE, Warner TT, Xu M (2003) Diurnal patterns of rainfall in Northwestern South america. Part III:

diurnal gravity waves and nocturnal convection offshore. Mon Wea Rev 131:830–844. h t t p s : / / d o i . o r g / 1 0 . 1 1 7 5 / 1 5 2 0 - 0 4 9 3 ( 2 0 0 3 ) 1 3 1 - 0 8 3 0 : D P O R I N - 2 . 0 . C O ; 2

Nuryanto DE, Pawitan H, Hidayat R et al (2019) Characteristics of two mesoscale convective systems (MCSs) over the greater jakarta: case of heavy rainfall period 15–18 January 2013. Geosci Lett 6:1.

https://doi.org/10.1186/s40562-019-0131-5

Qian J-H (2008) Why precipitation is mostly concentrated over islands in the maritime continent. J Atmos Sci 65(4):1428–1441. https://doi.org/10.1175/2007jas2422.1

Qian J, Robertson AW, Moron V (2010) Interactions among ENSO, the monsoon, and diurnal cycle in rainfall variability over Java. Indonesia J Atmos Sci 67:3509–3524. https://doi.org/10.1175/2010JAS3348.1 Ruppert JJH, Chen H (2020) Island rainfall enhancement in the maritime continent. Geophys Res Lett

47(5):1–9. https://doi.org/10.1029/2019GL086545

Ruppert JH, Zhang F (2019) Diurnal forcing and phase locking of gravity waves in the maritime continent. J Atmos Sci. https://doi.org/10.1175/jas-d-19-0061.1

Ruppert JJH, Chen X, Zhang F (2020) Convectively forced diurnal gravity waves in the maritime continent.

J Atmos Sci 77(3):1119–1136. https://doi.org/10.1175/JAS-D-19-0236.1

Satiadi D, Purwaningsih A, Harjupa W, Trismidianto et al (2023). Impacts of CENS and MJO phenomena on diurnal cycle of local convection, moisture convergence, and rainfall over land, coast, and sea areas in the western part of Java Island. Meteorol Atmos Phys 135(5):42. h t t p s : / / d o i . o r g / 1 0 . 1 0 0 7 / s 0 0 7 0 3 - 0 2 3 - 0 0 9 7 9 - w

Saufina E, Satiadi D, Harjupa W, Purwaningsih A et al (2025) The role of Madden-Julian oscillation, Westerly wind bursts, and Kelvin waves in triggering extreme rainfall through mesoscale convective systems: A case study of West sumatra, March 7–8, 2024. Atmos Res 318:107993. h t t p s : / / d o i . o r g / 1 0 . 1 0 1 6 / j . a t m o s r e s . 2 0 2 5 . 1 0 7 9 9 3

Skamarock WC, Klemp JB, Dudhia J, Gill DO, Barker D, Duda MG et al (2008) A Description of the Advanced Research WRF Version 3 (No. NCAR/TN-475 + STR). University Corporation for Atmo- spheric Research. https://doi.org/10.5065/D68S4MVH

Trilaksono NJ, Otsuka S, Yoden S, Saito K, Hayashi S (2011) Dependence of Model-Simulated heavy rainfall on the horizontal resolution during the Jakarta flood event in January-February 2007. SOLA 7(1):193–

196. https://doi.org/10.2151/sola.2011-049

Trilaksono NJ, Otsuka S, Yoden S (2012) A Time-Lagged ensemble simulation on the modulation of precipitation over West Java in January–February 2007. Mon Wea Rev 140:601–616. h t t p s : / / d o i . o r g / 1 0 . 1 1 7 5 / M W R - D - 1 1 - 0 0 0 9 4 . 1

Trismidianto, Satiadi D, Harjupa W, Fathrio I et al (2024) Southerly surge impact on rainfall patterns in southern Indonesia during Winter Monsoon and Madden–Julian Oscillation (MJO). Atmosphere 15(7):840.

https://doi.org/10.3390/atmos15070840

Van den Besselaar EJM, van der Schrier G, Cornes RC, Iqbal AS, Klein Tank AMG (2017) SA-OBS: A daily gridded surface temperature and precipitation dataset for Southeast Asia. J Clim 30(14):5151–5165.

https://doi.org/10.1175/jcli-d-16-0575.1

Wang S, Sobel AH (2017) Factors controlling rain on small tropical islands: diurnal cycle, Large-Scale wind speed, and land. J Atmos Sci 74(11):3515–3532. https://doi.org/10.1175/jas-d-16-0344.1

Wei Y, Pu Z, Zhang C (2020) Diurnal cycle of precipitation over the maritime continent under modulation of MJO: perspectives from Cloud-Permitting scale simulations. J Geophys Research: Atmos, 125(13) Wu C-H, Hsu H-H (2009) Topographic influence on the MJO in the maritime continent. J Clim 22(20):5433–

5448. https://doi.org/10.1175/2009jcli2825.1

Wu P, Hara M, Fudeyasu H, Yamanaka M, Matsumoto D, Syamsudin J, Sulistyowati F, Djajadihardja R, Y, S (2007) The impact of Trans-equatorial monsoon flow on the formation of repeated torrential rains over Java Island. SOLA 3:093–096. https://doi.org/10.2151/sola.2007-024

Wu P, Hara M, Hamada J, Yamanaka MD, Kimura F (2009) Why a large amount of rain falls over the sea in the vicinity of Western Sumatra Island during nighttime. J Appl Meteorol Climatology 48(7):1345–

1361. https://doi.org/10.1175/2009jamc2052.1

Wu P, Arbain AA, Mori S, Hamada J, Hattori M, Syamsudin F, Yamanaka MD (2013) The Effects of an Active Phase of the Madden-Julian Oscillation on the Extreme Precipitation Event over Western Java Island in January 2013. SOLA, 9(0), 79–83.https://doi.org/10.2151/sola.2013-018

Yang H (1991) Wave packets and their bifurcations in geophysical fluid dynamics, vol 85. Springer, Berlin.

https://doi.org/10.1007/978-1-4757-4381-4

(22)

Yong B, Liu D, Gourley JJ, Tian Y, Huffman GJ, Ren L, Hong Y (2015) Global view of Real-Time Trmm multisatellite precipitation analysis: implications for its successor global precipitation measurement mission. Bull Am Meteorol Soc 96(2):283–296. https://doi.org/10.1175/bams-d-14-00017.1

Yulihastin E, Hadi WT, Ningsih NS, Syahputra MR (2020) Early morning peaks in the diurnal cycle of precipitation over the Northern Coast of West Java and possible influencing factors. Ann Geophys 38(1):231–242. https://doi.org/10.5194/angeo-38-231-2020

Yulihastin E, Fathrio I,. Trismidianto, Nauval F, Saufina E, Harjupa W, Satiadi D, Nuryanto DE (2021) Convective cold pool associated with offshore propagation of convection system over the East Coast of Southern sumatra, Indonesia. Adv Meteorol 2021:1–13. https://doi.org/10.1155/2021/2047609 Yulihastin E, Hadi TW, Abdillah MR, Fauziah IR, Ningsih NS (2022) Propagation of convective systems

associated with early morning precipitation and different Northerly background winds over Western Java. J Meteorol Soc Jpn 100(1):99–113. https://doi.org/10.2151/jmsj.2022-005

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.