The formation of these twin vortices was triggered by a double Intertropical Convergence Zone (ITCZ) in the Pacific Ocean which broke down, leading to accumulation of convective clouds from the Pacific Ocean over the Banda-Maluku Sea experiencing extreme heat. Storm vortex prediction simulations based on wind parameters carried out in the Seroja tropical cyclone research have confirmed the ability of high resolution (3 km) numerical weather forecasting models to simulate pairs of vortices.
The research results also provide important insights regarding the potential for the formation of tropical cyclones similar to Seroja, which could occur more frequently in the Banda Sea, through a similar mechanism. This is confirmed by the March 28, 2023 incident, when twin vortex storms formed and the southern vortex continued to grow into tropical cyclone Ilna on April 9, 2023, with the core of the vortex forming in the Arafura Sea.
Strategies to master weather forecasting technology
The greatest scientific challenge in meteorology and climatology is producing an accurate rainfall prediction model for BMI. Therefore, all forms of studies from upstream to downstream related to meteorology and climatology have the same ultimate goal, namely, to produce more accurate weather and climate forecasts.
Therefore, various studies on extreme weather mechanisms based on numerical weather prediction models that have been carried out essentially have the ultimate goal of producing scientific contributions in formulating strategies to increase the accuracy of rainfall predictions, not only from a scientific aspect but also in efficiently carrying out operational weather forecasting models so that they can be realistically implemented in Indonesia. The recommendations are as follows:
First, increase the spatial resolution, namely 0.5 km for western Java, 1 km for Indonesia, and 5 km for Southeast Asia, with a temporal resolution of every hour for the following five days. Increasing resolution aims to increase the model’s ability to predict extreme phenomena, from micro to global scales (tornadoes to tropical cyclones).
Second, use the correct configuration scheme for the BMI region. A configuration test can be carried out on existing schemes or finding new schemes.
Third, develop a joint atmosphere-ocean model to improve rainfall predictions, considering that existing weather forecasting models have yet take into account aspects of ocean dynamics, even though the BMI region is mostly made up of oceans. In addition, it has been found that sea surface temperatures experience sub-daily changes, thus influencing diurnal rainfall at BMI.
Fourth, assimilate the data with atmospheric humidity data from the Global Navigation Satellite System (GNSS) satellite data, to improve the humidity input from the global model that has been used by weather forecasting models.
Read: Covid-19 Cases In Indonesia Surge By 30-40 Per Day
Fifth, apply Artificial Intelligence-based deep learning, which can be done before and after the numerical weather prediction model is being run. This aims to select the best configuration scheme, while the AI applied afterward aims to improve the quality of the rain prediction output.
As part of the endeavors en route to “Golden Indonesia Vision 2045” and to achieve the target of keeping the earth’s temperature from exceeding 1.5 degrees Celsius by 2050, Indonesia must immediately master weather and climate forecasting technology. The time is ripe for weather prediction information in Indonesia to be produced indigenously by the nation’s sons and daughters. This aside, research and development in the field of prediction technology requires the widest possible collaboration with other developed countries. Indonesia also needs adequate High Performance Computing facilities and observation technology network support, to be able to produce weather forecast data with high resolution and accuracy.