Poster Title:Leveraging High-Resolution Data and AI to Optimize Cloud Seeding Operations in Thailand
Full Name:Sarawut Arthayakun
Affiliation / Institution:Department of Royal Rainmaking and Agricultural Aviation
Co-Author
Full Name:Pakdee Chantraket
Affiliation / Institution:Department of Royal Rainmaking and Agricultural Aviation
Co-Author 2
Full Name:Parinya Intaracharoen
Affiliation / Institution:Department of Royal Rainmaking and Agricultural Aviation
Co-Author 3
Full Name:Aroonroth Sricharounchot
Affiliation / Institution:Department of Royal Rainmaking and Agricultural Aviation
Co-Author 4
Full Name:Marut Ratmanee
Affiliation / Institution:Department of Royal Rainmaking and Agricultural Aviation
Co-Author 5
Full Name:Chaiyaphorn Nilarat
Affiliation / Institution:Department of Royal Rainmaking and Agricultural Aviation
Co-Author 6
Full Name:Patathip Meesangngern
Affiliation / Institution:Department of Royal Rainmaking and Agricultural Aviation
Co-Author 7
Full Name:Rungthip Nuangtawee
Affiliation / Institution:Department of Royal Rainmaking and Agricultural Aviation
Abstract:“Right Place, Right Time” is a key objective of the Department of Royal Rainmaking and Agricultural Aviation (DRRAA)’s cloud seeding operations. To accomplish this objective, data serves as the backbone of operational processes, ranging from planning to real-time weather monitoring and post-operation evaluation. DRRAA is actively developing datasets related to water demand and weather conditions, aiming to increase data accuracy and achieve finer spatial and temporal resolutions. More than sixty data layers have been established, and the total volume of data has increased nearly fifteen-fold over the past three years. This growth is attributable to DRRAA’s research endeavors and collaborations with various agencies. The following section focuses on the development of three principal data systems. (1) Agricultural Rainwater Demand Data System: This system provides information for determining target regions for rain enhancement, particularly in rainfed areas, which account for approximately 78% of Thailand’s total agricultural land. The target regions are classified into areas requiring water and those that are adequately supplied. Crop water requirements for economically important crops were calculated based on 500-meter satellite imagery interpreted by the Geo-Informatics and Space Technology Development Agency and updated every two weeks. Moreover, ongoing research aims to increase spatial resolution to 10 meters and update frequency to a weekly basis. Additionally, local-level demand data reporting from service recipients and volunteers has been refined to provide details down to specific coordinates. (2) Weather Data System: The core data is supplied from eleven DRRAA weather stations, each equipped with a dual-polarization weather radar, an upper-air sounding system, a microwave radiometer profiler, and other synoptic instruments. This system provides data for operational planning, monitoring cloud development, and evaluating seeding activities. DRRAA is currently developing a Radar QPE system, a nowcasting model, and a short-range forecasting model to estimate rainfall resulting from cloud seeding missions and to identify potential cloud seeding areas. Additionally, high-spatiotemporal-resolution weather forecast data from the Hydro-Informatics Institute were employed for precise analysis, and the decision framework was transitioned from a rule-based system to an AI-driven model, achieving an accuracy of 73.29%. (3) Weather Modification Data Standard: This standard is crucial for facilitating seamless data integration and accommodating future data. To achieve this, DRRAA developed a weather modification data standard based on a discrete global grid system. All weather modification data were reprocessed to comply with the standard, reducing the grid size from 10×10 km to 1×1 km. According to a study, DRRAA found that the discrepancy between the total area of standardized agricultural grids and survey data was less than 1%. As a result, the estimation of areas benefiting from cloud seeding is more closely aligned with actual conditions. In conclusion, DRRAA is transitioning toward a data-driven operation by 2027. This transition aims to enhance the spatial and temporal accuracy of weather modification, leading to more cost-effective operations. However, further research and development are necessary, particularly to refine suitable operation methods, such as developing simulation models and region-specific practices. Additionally, infrastructure upgrades are essential to support future research and technological advancements.