IntroductionAs a tropical archipelago, Indonesia is exceptionally susceptible to climate change impacts. Since mitigation requires accurate regional climate data, a reliable model assessment is essential to address the biases and uncertainties of Global Climate Model’s (GCMs). This study evaluates and ranks 35 NASA NEX-GDDP-CMIP6 models, including their multi-model mean ensemble (ENSMEAN), on their capacity to simulate monthly precipitation climatology over Indonesia.MethodsThe methodology employed MSWEP dataset as the observational reference, utilizing statistical metrics including Correlation Coefficient (CC), Normalized Standard Deviation (NSTD), Root Mean Square Deviation (RMSD) and Mean Bias (MB). A dual-scale evaluation framework was adopted, assessing the model’s spatio-temporal performance. Taylor Diagrams used to visualize model distribution, while Min-Max normalization and the Summation of Rank (SR) applied to ensure fair comparison and identify the best-performing models.DiscussionThe findings demonstrate that NEX-GDDP-CMIP6 models generally capture Indonesia’s seasonal precipitation patterns in close alignment with MSWEP observations. Notably, five models that consistently identified as high performers across spatio-temporal dimensions were ACCESS-CM2, CMCC-ESM2, TaiESM1, MRI-ESM2-0 and CESM2-WACCM. Specifically, ACCESS-CM2 showed the highest temporal accuracy, while TaiESM1 demonstrated the strongest spatial accuracy. ENSMEAN ranked seventh across spatio-temporal dimensions, proving its capability of reducing errors and enhancing simulation reliability. Despite the model’s overall accuracy, systematic biases persists, such as a “February dip” that underestimates peak wet-season precipitation and a tendency to overestimate precipitation during the dry season. These discrepancies suggests that simulating precipitation interactions among monsoon dynamics, topography and land-sea contrast remain challenging in Indonesia Maritime Continent. This study offers a benchmark for GCMs selection and underscores the need for improved regional models to support climate adaptation and hydrological policymaking.

Extreme rainfall events in the Maritime Continent are often driven by complex interactions between large-scale atmospheric circulation and local processes, making their prediction particularly challenging. This study examines the atmospheric drivers of an extreme rainfall event along the northern coast of Central Java on 13 March 2024, when daily rainfall exceeded 200 mm at several stations. Although extreme rainfall affected a broader region of northern Java during 11–17 March 2024, this study focuses on the northern coastal area of Central Java, which experienced the highest rainfall intensity and includes Semarang, a densely populated urban center with high flood risk. Multi-source observations, including rainfall and wind data from the Indonesian Agency for Meteorology, Climatology, and Geophysics and Automatic Weather Stations, C-band weather radar, Himawari-8 brightness temperature, Global Precipitation Measurement data, and fifth-generation ECMWF atmospheric reanalysis were used to analyze interactions between large-scale and local circulations.The event occurred during an active intraseasonal phase characterized by Madden–Julian Oscillation propagation from phase 4 to 5 with amplitudes exceeding 1.5. A persistent Westerly Wind Burst with zonal winds above 7 m s−1 for about 11 days and peaks exceeding 15 m s−1 on 12–13 March coincided with a strong Cross-Equatorial Northerly Surge, further reinforced by the presence of a tropical depression, producing meridional wind anomalies near −6 m s−1 around 7°S and enhancing cross-equatorial moisture transport. These large-scale wind systems transported substantial moisture toward the equatorial region, increasing atmospheric instability and promoting deep convection that leads to extreme rainfall. The Hybrid Single-Particle Lagrangian Integrated Trajectory back-trajectory analysis indicates that approximately 46% of the moisture originated from the lower troposphere over the South China Sea and Java Sea, with additional contributions from the Indian Ocean in the mid- and upper-troposphere (about 23% and 20%). These conditions supported the formation and regeneration of mesoscale convective systems.High-resolution Weather Research and Forecasting simulations demonstrate that local processes critically control rainfall intensity and spatial distribution. Reducing terrain height by 50% substantially weakens precipitation, while complete topographic removal nearly suppresses rainfall. These sensitivity experiments suggest that topography amplified and sustained convection through orographic lifting and flow modulation, rather than acting as the sole factor controlling the observed precipitation center. However, the simulated sea-breeze component tends to be overestimated, while the land-breeze component tends to be underestimated, likely contributing to the displacement of the simulated precipitation maximum from the observed coastal zone. Overall, large-scale circulation sets favorable conditions, while local processes govern convective triggering and rainfall concentration.

The 2024 Asianu2013Australian monsoon (AAM) year, defined as April 2024 to March 2025, was notable. Based on available data, a prolonged rainy season was observed in most parts of the AAM region, except for the Meiyu Region, which corresponds to the area affected by the second stage of the East Asian summer monsoon. The rainy season also featured elevated near-surface air temperatures, a boreal summer rainfall surplus of approximately 20% across South, East, and Mainland Southeast Asia, an austral summer rainfall deficit of approximately 30% in Northern Australia, and a surplus of approximately 30% in the Maritime Continent. During boreal winter, a strong East Asian winter monsoon circulation led to below-average precipitation along East Asiau2019s climatological rain belt, including South China and Japan, accompanied by above-average near-surface air temperatures. Meanwhile, the 2024 AAM exhibited notable subseasonal variability, with abrupt alternations between dry/drought and wet/flood conditions, as well as between warm and cold episodes over many regions. It was also characterized by widespread extreme events, including but not limited to heavy rainfall, heatwaves, cold surges, and tropical cyclones. Such AAM-associated variability and extremes exerted considerable social impacts and caused substantial economic losses, highlighting the ongoing challenges in understanding and predicting the AAM at regional scales and multiple timescales.

Wind-driven rain (WDR) is a significant contributor to building facade degradation, including surface weathering, corrosion, and various other forms of deterioration. The lack of long-term hourly rainfall observations poses a major challenge for WDR assessment. ERA5 provides a potential alternative source of hourly precipitation data, although it contains systematic biases. This study proposes a bias-correction framework applied to ERA5 precipitation at the daily scale, which is subsequently redistributed to the hourly scale to enable WDR calculations across multiple facade orientations. Based on an analysis of 106 sites over the period from 2011 to 2020, the bias correction reduces the frequency of light drizzle events (0-0.1 mm) from 41.5 % to 25.0 %, increases the proportion of no-rain conditions from 21.8 % to 50.9 %, and redistributes hourly rainfall occurrences toward higher-intensity classes (e.g., 5-10 mm events increase from 0.49 % to 0.82 %, and >20 mm events from 0.005 % to 0.06 %). Using the corrected dataset, WDR exposure was assessed across eight facade orientations spanning diverse climate zones in Indonesia, revealing that the orientations associated with maximum exposure vary according to local climatic characteristics. Annual WDR exposure patterns further indicate that west-facing facades commonly experience moderate to very heavy exposure levels. As a result, the proposed framework provides a pragmatic, first-order approximation of long-term WDR exposure based on daily bias-corrected ERA5 hourly precipitation data, supporting semi-empirical WDR calculations and offering practical guidance for identifying facade orientations with higher exposure to inform adaptive design and mitigation strategies across different climate zones.

This study provides a thorough analysis of stratiform and convective rainfall through the use of two advanced Frequency-Modulated Continuous-Wave (FMCW) radar systems: the Weather Radar (WR) and the Micro Rain Radar (MRR). The WR effectively distinguishes between these two types of rainfall by examining important metrics such as reflectivity, Doppler velocity, and spectrum width, which reveal the distinct characteristics of each rainfall type. For the MRR data, the research utilizes Spectral Feature Classification and Clustering (SFCC) to identify significant spectral features, including peak Doppler shifts, spectral width, skewness, kurtosis, and power centroid. These features are essential indicators in the classification process. The study employs machine learning (ML) techniques, specifically k-means clustering and the Support Vector Classifier (SVC), to categorize the rainfall types. Stratiform rainfall is identified by low Doppler shifts, narrow spectral widths, and minimal variability, while high Doppler shifts, broad spectral widths, and greater variability characterize convective rainfall. The SVC model exceeded expectations in accuracy when applied to synthetic data, taking advantage of the clear separability of the features. Conversely, k-means clustering demonstrated lower accuracy due to its reliance on linear boundary assumptions and its sensitivity to feature scaling. Importantly, both methodologies identified centroid and width features as crucial for effective classification.

Urban open spaces, which are crucial for outdoor activities, are increasingly vulnerable to deviations in thermal comfort due to extreme/untypical climates. The use of typical meteorological year (TMY) datasets for designing built environments has limitations in accounting for climate deviations caused by extreme/untypical climates. Therefore, there is a need to develop new datasets that more accurately represent atypical climates and can be used for adaptive strategies under these atypical conditions. This study aims to develop the untypical meteorological year (UTMY) dataset consisting of minimum untypical meteorological year (UTMY-N) and maximum untypical meteorological year (UTMY-M) datasets which is applied for thermal stress assessment, through a case study in the hot and humid climate of Indonesia. Finkelstein-Schafer statistics with weighting applied to more climate variables, including global horizontal irradiance (GHI), temperature, dew point temperature, and wind speed were used to select extreme/untypical years. The large weighting of the GHI and temperature variables influenced the selection of extreme/untypical years that represent extremes for these variables compared with other variables in the UTMY dataset. The average annual temperature in the UTMY-M dataset across 106 sites is 0.2 degrees C to 1.0 degrees C higher than that in the TMY dataset, while the UTMY-N dataset is 0.4 degrees C to 1.8 degrees C lower. Compared with the TMY dataset, monthly and annual thermal stress calculations using the universal thermal climate index (UTCI) derived from the UTMY dataset are more effective in assessing the farthest thermal stress deviations (the upper maximum and lower minimum limits) under untypical climate conditions in each climate zone.

Convectively coupled Kelvin waves (CCKWs) are eastward‐propagating weather systems that organise convection locally and are linked to precipitation extremes across the Maritime Continent (MC). They are often embedded in active Madden–Julian Oscillation (MJO) phases. The MJO also propagates eastwards, but influences convection in the MC over longer time‐scales and larger areas. This article examines a case study during July 2021 of multiple CCKWs and westward‐propagating inertio‐gravity waves (WIGs) embedded within an active MJO. The final CCKW traversed the western MC, causing precipitation extremes across equatorial Indonesia and East Malaysia that led to numerous reports of flooding and landslides, with western Borneo the worst‐affected region. The MJO event was terminated abruptly following the passage of this CCKW. Analysis of the total column water budget reveals that the precipitation rate exceeded the vertically integrated moisture‐flux convergence provided by the CCKW, drying out the atmosphere and suppressing further convection. The performance of the UK Met Office prediction model was evaluated for this case study; parameterised convection configurations generally performed as well as or better than explicit convection models. This is possibly because they represented better the location and timing of the convective systems that developed because of interactions between CCKWs and WIGs. This research highlights how CCKWs should be viewed, not simply as convective systems that affect weather locally but, as having the potential to deliver larger‐scale impacts over the entire equatorial MC, as part of a complex multiscale interaction. Such interactions can involve the MJO influencing CCKWs downscale by providing enhanced convection. Conversely, the suppressed phase of CCKWs can dampen the MJO convective signal and terminate MJO propagation. Whilst weather prediction models may forecast rainfall associated with individual equatorial modes accurately, capturing their combined effect remains a challenge.

This study explores the transformative potential of supervised machine learning algorithms in improving rainfall prediction models for Indonesia. Leveraging the NEX-GDDP-CMIP6 dataset's high-resolution, global, and bias-corrected data, we compare various machine learning regression algorithms. Focusing on the EC Earth3 model, our approach involves an in-depth analysis of five weather variables closely tied to daily rainfall. We employed a diverse set of algorithms, including linear regression, K-nearest neighbor regression (KNN), random forest regression, decision tree regression, AdaBoost, extra tree regression, extreme gradient boosting regression (XGBoost), support vector regression (SVR), gradient boosting decision tree regression (GBDT), and multi-layer perceptron. Performance evaluation highlights the superior predictive capabilities of Gradient Boosting Decision Tree and KNN, achieving an impressive RMSE score of 0.04 and an accuracy score of 0.99. In contrast, XGBoost exhibits lower performance metrics, with an RMSE score of 5.1 and an accuracy score of 0.49, indicating poor rainfall prediction. This study contributes in advancing rainfall prediction models, hence emphasizing the improvement of methodological choices in harnessing machine learning for climate research.

Convectively coupled equatorial Kelvin waves (CCKWs) are eastward propagating weather systems that locally organise convection and have been linked to precipitation extremes across the Maritime Continent (MC). They are often embedded in convectively active phases of the Madden-Julian Oscillation (MJO) which too propagates eastwards but influences convection in the MC over longer timescales and larger areas. Previous high impact weather case studies have linked CCKWs to local precipitation extremes. In this study, we examine a case study during July 2021 of multiple CCKWs embedded within an active MJO. The final CCKW traversed the western MC causing precipitation extremes across equatorial Indonesia that lead to numerous reports of flooding and landslides, with the West Kalimantan region the worst affected. The MJO event itself was abruptly terminated following the passage of this CCKW. Through analysis of the moisture budget we find that the rainfall exceeded the convergence of moisture to produce the pronounced drying. Prior to the local MJO termination, we find there was enhanced westward propagating diurnal activity across the equatorial MC coinciding with a steady increase of total column water. We also examine observations of the extreme rainfall event in the West Kalimantan province. Comparing different deterministic model configurations, we find that the convection permitting models generally perform better when there are not multiple CCKWs present within the initial conditions. This research highlights how CCKWs should not simply be viewed as convective systems that locally affect weather but have the potential to have devastating impacts over the entire equatorial MC especially when involved in multiscale interactions both with the diurnal cycle and with the MJO.

Puting beliung (PB), or small-scale tornado, is a significant and under-researched extreme weather phenomenon in Indonesia, often causing severe damage to infrastructure and posing risks to public safety despite their brief localized nature. Therefore, this research aimed to examine spatial and temporal patterns and trends of PB events across Indonesia from 2011 to 2024, applying statistical analysis, geospatial mapping, and the Mann-Kendall trend test to a database of 2,434 PB events. The results showed that PB events primarily cluster in western and central regions, specifically on Java Island, and the highest frequencies were observed in East Java, West Java, and Central Java. These events typically occur in low-lying zones (0–500 meters above sea level), affecting agricultural and residential land in flat terrain. Temporally, most PB arises in the afternoon (1:00–3:00 pm local time), with peak frequencies in January, March, and November, coinciding with Indonesian monsoonal and transitional seasons. A trend analysis shows a statistically significant nationwide yearly increase of approximately 12 PB events, with 8 provinces exhibiting notable upward patterns. When compared to other PB-prone nations, Indonesia records a higher annual PB frequency than Japan, Australia, and Bangladesh, but remains well below the United States. The novelty of this research lies in its long-term, nationwide dataset and thorough spatiotemporal assessment, providing the first comprehensive examination of PB trends at national and provincial scales in Indonesia. These results provide crucial insights for disaster risk mapping, mitigation strategies, and early warning systems. Play This Podcast Article

Sub-seasonal to seasonal (S2S) prediction has emerged as an important tool in anticipating climate variations over shorter timescales, from a few weeks to several months ahead. This research undertakes multiple evaluations of verification results derived from various deterministic and probabilistic forecasting approaches at the S2S scale, employing diverse techniques accessible within the Python tool named Xcast. Developed as a proficient utility, Xcast is a tool capable of utilizing statistical and machine learning methodologies to rapidly and effectively process various gridded climate data. The study conducts a comparative analysis of several methods, including multiple linear regression (MLR), extreme learning machine (ELM), and probabilistic output extreme learning machine (POELM). The assessment employs blended rain data from rain posts and Global Satellite Mapping of Precipitation (GSMaP), alongside S2S the European Center for Medium-Range Weather Forecasts (ECWMF) forecast data—both data are on a 10-day time scale with the period from 1996–2021 tailored for Indonesia region. The research domain employs a condition whereby the tercile probability is determined by data points that accumulate rainfall of over 50 mm per 10 days, with a 30

The regional characteristics of the boreal summer intraseasonal oscillations (BSISO) over southeast Asia are presented. The northeastward transition of the BSISO is characterised by 4 phases, such that convection is enhanced over the Philippines and Indochina in phase 1 and suppressed over Peninsular Malaysia, Borneo and Java. The opposite is true in phase 3. The role of BSISO in modulating extreme precipitation is highlighted, showcasing how its phases impact both the frequency and intensity of extreme precipitation events across the region. Using a method to detect and characterise precipitation features in terms of precipitating areas and their associated object properties, this study shows distinct shifts in precipitation regimes during different phases of the BSISO. Phase 1 exhibits increased large-scale convective activity, particularly affecting regions like the South China Sea and northern Philippines, linked to increased tropical storm frequency but reduced localised extreme precipitation events. In contrast, Phase 3, with active convection over Peninsular Malaysia, Borneo and Sumatra, shows intensified extreme precipitation from smaller to medium-sized areas. BSISO phases also modify the distribution of small and large precipitation objects over land, ocean, and coastal regions. This classification of precipitation regimes provides detailed insights into how the BSISO's large-scale envelope modifies regional precipitation extremes through various precipitation properties. This information could benefit probabilistic predictions of regional extreme precipitation events at subseasonal time scales.

The study of global and vertical solar irradiance components under different sky conditions is very important for building design applications. This research has been carried out by analyzing the components of solar radiation based on global horizontal irradiance (GHI) data, and global vertical irradiance for east, west, north, and south oriented. Observational data was acquired for one year starting from July 2022 to June 2023 in Jembrana-Bali of Indonesia. The solar radiation component has been analyzed based on sub-diurnal (daytime) and annual patterns. The clearness index (Kt) was calculated based on the ratio of extra-terrestrial solar radiation to global solar radiation measured at the surface. Sun-path analysis was carried out by considering the relationship of the zenith angle, azimuth angle, and the intensity of each solar radiation component to changes in time. The results show that in all-sky conditions, the average monthly maximum global horizontal irradiance occurs in January at 432.4 W/m², while the minimum monthly average global horizontal irradiance occurs in July at 327.9 W/m². The clearness index statistic shows that the hourly average ranges from 0.12 to 0.68 with an average of 0.47. Based on sun-path analysis, east, north, and west orientations receive more solar radiation in the Jembrana-Bali area. The findings from the study provide important information for architects, engineers, and policymakers that can be used for sustainable building design and building energy planning.

This study aims to demonstrate the comprehensive development of typical meteorological years (TMYs) under relatively limited observational data. The distribution of missing hourly observational data of the 2011-2020 period at all sites was examined. This paper proposes a quality control method for filling the gaps in the missing hourly observational data using bias-corrected ERA5 reanalysis data in the process of developing TMYs. Initially, the temperature bias distribution from-4.5 degrees C to 2.7 degrees C was reduced to a range of-0.014 degrees C to 0.005 degrees C. The relative humidity bias distribution was-6 % to 10 %, and was reduced to-0.32 % to 0.07 %. The bias distribution of wind speeds ranging from-4 m/s to 2 m/s was reduced to-0.02 m/s to 0.35 m/s. The Sandia method with a modified weighting of Finkelstein-Shaffer (FS) statistics was applied to eight climate elements, namely, global horizontal irradiance, direct normal irradiance, diffuse horizontal irradiance, temperature, precipitation, wind speed, relative humidity, and dew point temperature to generate TMYs at 106 sites across eight climate zones in Indonesia. The verification results showed that the average correlation and RMSE between TMYs and their long-term averages were 0.96 and 75 w/m2 for global horizontal radiation, respectively, while those for temperature were 0.86 and 1.3 degrees C, respectively.

During Nyepi, all activities are eliminated, including public services, such as closing access to land, sea, and air transportation routes that transit to Bali for one day. In 2022, the BMKG Research and Development Center observed air quality in Nyepi. The measurement aims to determine the relative reduction level of greenhouse gas (GHG) and particulate emissions on Nyepi Day compared to before and after. Greenhouse gases and particulates were measured for seven days, February 28 - March 6, 2022. The measurement locations were over 3 points: Denpasar Regional Office III, Karangasem Geophysics Post, and Jembrana Climatology Station. The results of this field measurement show that daily human activities significantly influence the concentration of pollutant gases and particulates in the air. During Nyepi Day, air concentration improves when all human actions are reduced. This is marked by a decrease in the air’s concentration of pollutant gases (CO, NO2) and dust particles. The reduction in the concentration of pollutant gases and dust particles did vary at each observation location; this was due to the sensitivity of the equipment, the character of each observation location, and the placement of measuring devices. Urban areas filled with community activities have experienced the greatest improvement in air quality compared to suburban areas. Order of measuring instruments close to the location of pollutant sources, even in suburban areas, will decrease the concentration of pollutant gases, similar to that in urban areas. The difference between PM25 data during Nyepi 2022 and the average PM25 for 2020 - 2022 shows a decrease in PM25 concentrations of around 47%, with the highest drop at night.

High-impact weather events in the form of hail have affected Rejang Lebong Regency in Bengkulu Province on March 27, 2023. Two areas are affected by hail with a diameter of 1 cm, such as Simpang Nangka Village and Cawang Lama Village, around 17.00 LT–18.00 LT (10.00–1.00 UTC) marked by heavy rain with lightning. Based on the hail incident report, an analysis of the extreme weather phenomenon of hail in Rejang Lebong district was carried out, which focused on the utilization of Bengkulu weather radar data. The results of the study showed the growth of strong convective clouds, such as Cumulonimbus (Cb) clouds detected by radar on March 27, 2023, at 08.00 UTC (15.00 LT) and continued to grow to maturity at 09.30 UTC (16.30 LT) in the Rejang Lebong district area. At the time of the event, a reflectivity value of 65 dBZ was observed. Analysis of the height of the Cb radar cloud around the event location found that the height reached more than 15 km. In addition, a special analysis was conducted in the area of the hail event by utilizing the radar maximum reflectivity value and Quantitative Precipitation Estimation (QPE) from the results of the Z-R relationship equation. In addition, tropical wave analysis shows that the Madden–Julian Oscillation (MJO) is active and is a trigger factor for the growth of massive convective clouds. On the other hand, Equatorial Rossby (ER) waves and Kelvin waves are in an inactive phase in western Indonesia when a hail event occurs.

Indonesia, with its tropical and monsoonal climate, is exposed to heavy precipitation and enormous rainfall accumulation which results in weather-driven hazards, including extreme rainfall events and floods. There are several conventional sources of data to estimate potential of anomalously high precipitation in Indonesia, including rain gauge data, satellite data and meteorological reanalysis. Even though they allow assessment of precipitation variability, their usefulness is limited by biases and data gaps. Furthermore, assessment of a variability in precipitation patterns is not the same as identification of their adverse societal effects, such as floods. Due to the proliferation of social media, these conventional data sets can be supplemented with crowd-sourced information that can potentially provide longer-term, accurate records and cover a larger area. In this study, we demonstrated that Twitter is a useful source for flood detection and created a flood database. Twitter-based flood database is derived for subregions of major islands within Indonesia: Java, Sumatra, Borneo and Sulawesi, and validated against data from governmental reports and local paper articles. Results show that Twitter-based retrieval performs well in comparison with other sources, but only in regions characterized by sufficiently large pool of active users. Flood events and extreme rainfall events (defined using in-situ and satellite data) were compared in terms of their spatial and temporal distribution, as well as their meteorological drivers. In general, on each of the island, there is a seasonal cycle: a wet season during boreal winter, when the Southeast Asian monsoon provides an environment supportive of rain events, and a dry season during boreal summer. On intraseasonal scale, Madden-Julian Oscillation (MJO) creates the conditions favorable for weather extremes. MJO activity causes an increase in the local rainfall rate, with a significant increase in a chance of observing extreme precipitation during favorable MJO phase.

Gridded precipitation datasets are widely available from satellite observations and reanalysis model outputs. However, its performance in specific regions in the world may vary and depends on several factors, such as grid data spatial resolution, rainfall estimation algorithms, geographical location, elevation and regional climate conditions. This study aims to report on 13 gridded precipitation datasets' performance over Indonesia through direct comparisons with rain gauge measurements at various time scales over a 12-year period (2001-2012). The results show that, at daily timescales, the MERRA2 and CPC outperformed other datasets but tended to underestimate the rain gauge data in Indonesia, followed by GPCC. However, MERRA2 has smaller variation and bias than CPC. On monthly and annually timescales, CPC was found to be the best-performing dataset, followed by MERRA2, GPM-IMERG, GPCC and TRMM (TMPA), while JRA55 registered the worst performance at all timescales, followed by ERA-Interim. The performance of all datasets was better during JJA and SON than during DJF and MAM. The best performances were found in the southern (S) region of Indonesia, while the worst were in the northeast (NE) region for all months and datasets. The best performances during DJF (Asian Winter Monsoon) and JJA/SON (Australian Winter Monsoon) were found in the northwest (NW) and southern (S) regions, respectively. Most datasets overestimate the rain gauge data over Indonesia, except for GSMaP, MERRA2, CPC and CMORPH.

Convectively coupled equatorial Kelvin waves (CCKWs), along with other inertia-gravity waves, form an important mode of equatorial tropical rainfall variability. A CCKW with its genesis in the Indian Ocean will travel eastward with a self-similar structure, at speeds of around 12-15ms-1. Padang is an equatorial city that lies on the west coast of the island of Sumatra, Indonesia and is the first city in the path of these CCKWs. Of the 417 extreme precipitation events identified in the IMERG and TRMM datasets we find that there are 62 events that coincide with the passage of a CCKW. Further analysis reveals that there is a 55% increase in chance of observing extreme precipitation at Padang given the presence of a CCKW. Owing to its location on the coast, the diurnal cycle of precipitation at Padang is a mixture of that typical of land and sea, with peak rainfall occurring in the evening which moves offshore overnight. We find that the presence of a CCKW alters the diurnal cycle, firstly through enhancing the onshore rainfall that persists overnight and secondly through pushing the peak rainfall time to earlier in the day. We then examine a case study of a CCKW that passed over Padang on 2017-08-21 where a clear band of eastward propagating westerlies in ERA5 reanalysis 10m zonal wind coincides with a band of IMERG precipitation. These westerlies also tilt westward with height. To understand the mechanisms by which the CCKW triggers extreme precipitation, we extract key thermodynamical variables from the reanalysis, such as divergence and specific humidity, and examine 4.4km and 8.8km resolution Met UM South East Asia forecasts of the event. We discuss the performance of the forecasts in capturing the CCKW.

This study aims to compare the relationship between climate variables and rice productivity under different irrigation systems (irrigated and rainfed) in the clustering area on Java Island, Indonesia. This study used the clustering areas resulting from the previous study. The climate variables used are bias-corrected MERRA2 data from the period 1987–2017, cropped for Java Island. The rice productivity and reference evapotranspiration data used in this study are the results of the simulation of Aquacrop modeling. The result from the cluster method used tends to divide Java Island into 2 clusters with different altitudes (lowland and highland) areas. The results show that the correlation values between the precipitation variable and rice productivity from Aquacrop simulation (both irrigated or rainfed) in cluster 1 (dominated lowlands) are higher than in cluster 2 (dominated highlands), contrary to that the correlation values between the reference evapotranspiration variable with rice productivity from Aquacrop simulation (both irrigated or rainfed) are higher in cluster 2 (dominated highlands) areas, compared to cluster 1 areas (dominated lowlands). R-square values from response surface methodology (RSM) on the rainfed system in both clusters are higher than those on the irrigated system. This indicates that rainfed agriculture is highly dependent on climate variables, especially precipitation and reference evapotranspiration variables compared with the regular irrigated agricultural system. The RSM result also shows that climate variables significantly contribute to the variation of rice productivity generated by Aquacrop modeling in irrigated and rainfed systems and in all clusters.