Jakarta, INTI - The National Research and Innovation Agency (BRIN) and Universitas Diponegoro (UNDIP) developed a deep learning-based extreme rainfall prediction model to support flood risk reduction strategies, particularly in Semarang and its surrounding areas.
This collaboration is particularly between the Research Center for Climate and Atmosphere (PRIMA) of BRIN and the Faculty of Science and Mathematics (FSM) of UNDIP. BRIN installed an Automatic Weather Station (AWS) in an area around the Geophysics and Geothermal Building of FSM UNDIP.
Eddy Hermawan, Head of the Hydrometeorology, Hydroclimatology, and Extreme Events Research Group of PRIMA BRIN, stated that this collaboration is part of the research project titled “Development of a Deep Learning-Based Extreme Rainfall Prediction Model to Support Flood Disaster Risk Reduction Strategies.” Eddy added that the research activity was led by Dr. Teguh Harjana, M.Sc.
He explained that the research involved interdisciplinary researchers from PRIMA BRIN, Research Center for Artificial Intelligence and Cybersecurity (PRKAKS) of BRIN, the Research Center for Computing (PRK) of BRIN, the Mathematics Department of Universitas Padjadjaran, and the Physics Department of FSM UNDIP.
According to Eddy, the installed AWS functions as an automatic meteorological observation system capable of continuously recording various weather and climate parameters, such as rainfall, air temperature, humidity, air pressure, solar radiation, wind speed, and wind direction.
The device is expected to strengthen the national climate observation network while providing high-quality local weather data that can be used for research, education, and community service activities.
"AWS observation data will be a crucial source of information for developing weather and climate prediction models, validating satellite and weather radar data, and analyzing extreme weather events in Central Java and the surrounding areas," Eddy explained.
Observation Data to be Used for Further Research
In addition to supporting BRIN research, the observation data can also be utilized by lecturers, students, and researchers in FSM UNDIP for various academic activities, from final project research and scientific publications to the development of climate data-based technology.
The scope of the collaboration also includes joint research, data and expertise exchange, student involvement in research activities, collaborative scientific publications, and human resource development in the fields of climatology, meteorology, and artificial intelligence (AI) to support disaster management and climate change adaptation.
"This collaboration is expected to foster strong synergy between research institutions and universities in producing innovations relevant to national development needs," said Eddy.
He added that the collaboration is expected to produce various strategic outcomes, including reputable scientific publications, the development of a prototype AI-based weather analysis and prediction system, improving the quality of national climate observation data, and science-based policy recommendations that support climate change adaptation and mitigation efforts in Indonesia.
Conclusion
BRIN and UNDIP developed a deep learning-based extreme rainfall prediction model to support flood risk reduction, particularly in Semarang and its surrounding areas. As part of this collaboration, BRIN installed an Automatic Weather Station (AWS) at FSM UNDIP to provide high-quality local weather data for research, education, and the development of artificial intelligence-based weather and climate prediction models. This research involves several BRIN research centers, UNDIP, and Universitas Padjadjaran, focusing on the development of prediction technology, validation of satellite and weather radar data, and analysis of extreme weather.
Read more: AI for Life Marks a New Milestone in Indonesia's Education and Innovation Journey