11:45 AM - 12:00 PM
[MIS15-11] Model reproducibility of the warming event at the Dome Fuji station, inland of East Antarctica, in JARE44
Keywords:Antarctica, In-situ observation, Reanalysis, Case study
Due to the steep slope of the inland ice sheet and persistent polar anticyclone, the marine moisture into the high altitude and far interior is prevented, resulting in most precipitation being a small amount of ice-precipitation under clear-sky such as diamond dust over the Antarctica inland area (e.g., Schlosser et al., 2010). However, developing blocking-high over the ocean causes moist and warm air intrusion and makes intense precipitation events in the inland areas over Antarctica (e.g., Hayasaka et al., 2013). From February 2003 to January 2004, the 44th Japan Antarctic Research Expedition (JARE44) observed surface meteorology, radiation budget, and vertical backscatter profile at Dome Fuji station (77.32°S, 39.70°E, 3810m a.s.l.). In the period, 11 intense precipitation events occurred in 18 days, which opaqued half of the total precipitation amount (Fujita & Abe, 2006). Therefore, the extreme precipitation event is important for understand of the surface water balance over the icesheet of Antarctica. The present study, at first, compares in-situ observation data such as temperature, cloud cover, or radiation budget for JARE44 with reanalysis products. We evaluate the model reproducibility and then use the long-term results of products with good accuracy to assess the strength of the JARE44 event. The present abstract focuses on the comparison between observation and ERA5 reanalysis product based on four-dimensional variational (4D-Var) data assimilation of the Integration Forecasting System operated by ECMWF (Hersbach et al., 2020). The horizontal resolution is 0.25°, and the spatial resolution is 1 hour. It is known that ERA5 meteorology or radiation balance reproduces well globally (e.g., Zeng et al., 2023). Although data assimilation was not performed at Dome Fuji, ERA5 reproduced surface meteorologies and clouds well. Only for downward longwave radiation, the diurnal variation could not be reproduced. Although the correction for the dome's secondary radiation was applied, observation errors could occur. Except for one event, the model reproduced the precipitation trend. However, the amount was approximately half of what was observed. Also in the comparison of vertical profiles, ERA5 produced thick cloud and intense precipitation when ceilometer detected optically thick cloud. Both for the annual mean and event periods, the frequencies of cloud existence were well reproduced compared with the ceilometer and visual observation. 500 hPa geopotential height of ERA5 indicates that blocking-high causes of moist air intrusion for the period of extreme precipitation event. Although the influx of water vapor was very small in absolute values, it was extremely strong in relative values. During the intense event, in early November, a +300% compared to the average was considered a very strong influx. Since the reliability of ERA5 could be confirmed, we evaluated precipitation events using long-term results from 1979 to 2024. The sum of the average daily precipitation and standard deviation for 2003 was 0.12 mm/day. Days with precipitation higher than this threshold were considered precipitation events. In 2003, JARE44 period, the total annual precipitation was 16.1 mm, with 7.7 mm during the event period. In contrast, the 45-year average annual precipitation was 20.2 mm, with 11.5 mm during the event period. Although 2003 was a year with low precipitation, this can be explained mainly by the precipitation during the event period. The most substantial warming during the JARE44 period occurred during the early November event, where the daily average temperature increased by 22K over two days. The temperature during this event ranked 37th in history for the spring period from September to November, while the precipitation ranked 16th in history, indicating that it was an extreme event.