Ruth Mahubessy, Indonesian Meteorological, Climatological, and Geophysical Agency (BMKG) and WMO Fellow at ECMWF
Indonesia, the world's largest archipelago with more than 17,000 islands, has a climate pattern dominated by monsoons. Rainfall patterns across the country vary, shaped by its geographical location, mountain ranges and monsoon circulation (Figure 1). Most regions follow the monsoon cycle, with the rainy season from December to February and the dry season from June to August. However, other regions, such as Ambon in Maluku, experience the opposite pattern. Some regions even have no difference between the rainy and dry seasons, so the rainfall pattern tends to be the same throughout the year. This variety poses a challenge for forecasters, but it’s what also makes it fascinating.
Forecasting in a land of contrasts
The diversity in rainfall patterns in Indonesia makes the weather in Indonesia rather unique. As a weather forecaster who has worked in three regions with different climate zones, I have experienced firsthand how complex forecasting can be in such a diverse environment. Using the classical forecasting methods that I learned at State College of Meteorology Climatology and Geophysics, my team and I have long relied on observational data and model outputs to anticipate heavy rainfall. But sometimes the atmosphere surprises us. There have been occasions where extreme rainfall (sometimes exceeding 300 mm/day) occurred despite the model guidance showing no significant convective signals. Several indicators, such as streamline analysis, atmospheric instability indexes, moisture profile, equatorial wave activity, precipitation forecast, all suggested unremarkable conditions. In such moments, we found ourselves asking: what exactly went wrong in the model?
Motivated by these challenges, I have recently started at ECMWF as a World Meteorological Organization (WMO) Fellow, from the Indonesian Meteorological, Climatological, and Geophysical Agency (BMKG), as part of an initiative to strengthen forecasting capacity in developing and least developed countries. My project over the next year focuses on analysing convective parameters that can act as precursors to extreme weather phenomena such as heavy rainfall and landspouts.
Figure 1: Indonesia has a climate regime influenced by topography and monsoon circulation. This map illustrates eight climate zones ranging from equatorial to sub-savanna: (1A) equatorial, (1B) subequatorial, (2A) highland tropical, (2B) very highland tropical, (3A) monsoonal, (3B) sub-monsoonal, (4A) savannah, and (4B) sub-savannah. This diversity results in a unique rainfall pattern across Indonesia. Source: Putra, I.D. G. A. et al. 2024
The mystery of the landspout
Another phenomenon that sometimes occurs in Indonesia is landspouts. Landspouts are a type of non-mesocyclonic tornado and referred to locally as Puting Beliung in Indonesia and Malaysia. According to the BMKG, 1,221 landspout cases have been reported over the past five years (1 January 2021–20 October 2025). These are small events normally less than 1 km across and short-lived, lasting just 5 to 15 minutes. However, they can be surprisingly destructive, damaging buildings and disrupting communities (Figure 2), and yet, they remain very difficult to predict, even Doppler radar often fails to detect them. To understand why this small-scale phenomenon can occur suddenly, we need to look more closely at convective activity in Indonesia.
Damage caused by a landspout in Rancaekek, West Java, on 21 February 2024. The event damaged at least 503 houses and 31 factory buildings, injured 33 people, and displaced 21 families. © gracetansc/iStock /Getty Images Plus
Continuous convection
Located on the equator, Indonesia experiences continuous convection processes. We have heat continuously throughout the year and the humidity is high, creating atmospheric conditions that tend to be unstable and very conducive to the formation of convective clouds. In tropical weather forecasting, rainfall prediction is the most important weather parameter for supporting early warning systems. This is especially true for impact-based forecasting services, which have become a new paradigm implemented by the WMO for all National Meteorological and Hydrological Services (NMHS). Rather than simply describing the weather, impact-based forecasts communicate what the weather is likely to do.
Figure 2: Over the past five years (2021–2025), BMKG data shows more than 13,968 incidents of heavy rain (more than 60 mm/day), 5,787 cases of strong winds, 1,221 cases of landspouts/waterspouts, and 270 hailstorms. These events resulted in 7,551 cases of flooding, 4,419 cases of fallen trees, 3,579 landslides, 6,817 damaged buildings, 1,177 fatalities/injuries, and 5,119 cases of transportation disruptions. Source: BMKG
To predict convective activity, forecasters generally use several atmospheric instability indices such as the K-Index (KI), Total Totals Index (TT or Totalx), Convective Available Potential Energy (CAPE), Convective Inhibition (CIN), and Showalter Index (SI). However, the thresholds for these indices were developed based on atmospheric conditions at mid-latitudes, which have different atmospheric conditions from those in the tropics. I believe that the thresholds need to be adjusted to better suit the characteristics of Indonesia's atmosphere.
By combining model data produced by ECMWF with BMKG observational data, this project aims to develop a diagnostic tool that can assess the probability of convective activity in Indonesia more accurately. The results of this project are expected to contribute to improving the accuracy of extreme weather forecasts to help forecasters, decision makers, and society in general, while enriching our understanding of tropical meteorology.
ECMWF–WMO Fellowship
The ECMWF–WMO fellowship is an initiative by the WMO to strengthen forecasting capacity in least developed and developing countries. Its main goal is to expand global collaboration and to serve as a bridge between forecasters/researchers from NMHS and experts to enhance scientific knowledge in weather forecasting. Being part of this programme is a valuable personal and professional experience. For me, the most rewarding part is the opportunity to learn from the experts at the Centre. Being here also opens doors to collaborate with other international institutions, such as the European Severe Storms Laboratory (ESSL), and to attend high-quality training sessions which is a great opportunity that many forecasters rarely have. It truly feels like a privilege.
Top banner image: © GeloKorol/iStock /Getty Images Plus