4:45 PM - 5:00 PM
[AAS03-24] Verification of NWP at Bangaladesh Meteorological Department
Keywords:JICA, Bangladesh, QPE, verification NWP
In Bangladesh, severe weather disasters such as cyclones and their associated storm surges, as well as monsoon floods, frequently occur. To enhance the country’s overall meteorological capabilities, the Japan International Cooperation Agency (JICA) is implementing a technical cooperation project titled "Project for Strengthening the Capacity of Weather and Climate Services in Bangladesh," with the Bangladesh Meteorological Department (BMD) as the primary counterpart. In the previous presentation (Saito et al., 2024 Autumn Meeting, P339), we introduced an initial condition verification of the numerical weather prediction operated by BMD. This time, we will report on verification efforts aimed at comparing quantitative precipitation estimation (QPE) using radar and automated weather observation systems within the framework of the aforementioned technical cooperation project.
In 2023, the meteorological radar in Dhaka was upgraded through JICA's grant aid cooperation, and 225 Automatic Weather Stations (AWS) deployed by the World Bank began operation. Under Outcome 2 of JICA's technical cooperation project, efforts have commenced to generate composite images of weather radar reflectivity every 10 minutes and to create Quantitative Precipitation Estimates (QPE) calibrated using AWS data. This report introduces a case from October 6, 2024, when precipitation was observed during the period after October 2024, for which hourly precipitation intensity estimates from the Dhaka radar became available. The left side of Figure 1 shows the three-hour accumulated precipitation observed by the Dhaka radar at 06 UTC on October 6, 2024. The right side displays the three-hour accumulated precipitation forecast by the WRF-18km model, initialized at 00 UTC on October 6, 2024, with FT=6, incorporating both convective and grid-scale rainfall. The predicted precipitation area significantly deviates from the observed precipitation, resulting in a threat score of only 0.11 for a threshold of 1 mm.
As of the time of writing, only dry-season cases have been analyzed, but it is expected that more validation cases will become available once the rainy season begins. Future plans include comparisons with the WRF-9km model.
In 2023, the meteorological radar in Dhaka was upgraded through JICA's grant aid cooperation, and 225 Automatic Weather Stations (AWS) deployed by the World Bank began operation. Under Outcome 2 of JICA's technical cooperation project, efforts have commenced to generate composite images of weather radar reflectivity every 10 minutes and to create Quantitative Precipitation Estimates (QPE) calibrated using AWS data. This report introduces a case from October 6, 2024, when precipitation was observed during the period after October 2024, for which hourly precipitation intensity estimates from the Dhaka radar became available. The left side of Figure 1 shows the three-hour accumulated precipitation observed by the Dhaka radar at 06 UTC on October 6, 2024. The right side displays the three-hour accumulated precipitation forecast by the WRF-18km model, initialized at 00 UTC on October 6, 2024, with FT=6, incorporating both convective and grid-scale rainfall. The predicted precipitation area significantly deviates from the observed precipitation, resulting in a threat score of only 0.11 for a threshold of 1 mm.
As of the time of writing, only dry-season cases have been analyzed, but it is expected that more validation cases will become available once the rainy season begins. Future plans include comparisons with the WRF-9km model.