[M-IS15] Paleoclimatology and Paleoceanography

[MIS15-29] High-accuracy calibration of paleo-environmental proxies using high-spatiotemporal resolution hindcast simulation

Tomoya Omata¹, *Takashi Nakamura¹, Atsuko Yamazaki^2,3, Taro Komagoe³, Tsuyoshi Watanabe^4,3 (1.School of Environment and Society, Tokyo Institute of Technology, 2.Graduate School of Environmental Studies, Nagoya University, 3.KIKAI Institute of Coral Reef Sciences, 4.Graduate School of Science, Hokkaido University)

Keywords:hindcast simulation, ocean model, downscaling, water temperature, Kikai Island, Sr/Ca

Scleractinian corals provide high-resolution (monthly to weekly) climate and environmental records of the shallow-water tropical and subtropical oceans. Environmental parameters, such as temperature and salinity, are extracted from the coral skeleton as environmental proxies, such as skeletal δ¹⁸O, Sr/Ca and Mg/Ca. However, in order to convert the proxies to actual environmental parameters, the environmental proxies should be calibrated with observed data. This is because the proxies usually do not follow theoretical values due to biological activities, so-called vital-effect. Therefore, to convert the proxies to environmental parameters and to extend proxies extracted from the coral core to long-term environmental data, the environmental data on where coral cores are collected is required. However, available on-site past environmental data is quite limited.
To overcome the problems, we conducted downscaling from global/regional products to the Kikai Island area by multi-nested approach by using an ocean circulation model based on the Regional Ocean Modeling System (ROMS; Shchepetkin and McWilliams, 2005) implemented in the Coupled-Ocean-Atmosphere-Wave-Sediment Transport (COAWST) Modeling System (Warner et al., 2008; 2010). An AMAMI1 is the coarsest computational grid which is 1/36° (~3 km) grid resolution. The AMAMI1 was downscaled from Global HYCOM (https://www.hycom.org/) between 1994 and 2015. The AMAMI1 results are downscaled to AMAMI2 domain which is 1/180° (~600 m) resolution and covering the entire Amami Island and Kikai Island, then that is downscaled to KIKAI1 domain which is 1/900° (~120 m) resolution and covering whole Kikai Island. Finally that is downscaled to KIKAI2 domain which is 1/2700° (~40 m) resolution for focusing Shiomichi Bay in Kikai Island by offline one-way nesting. Since the Global HYCOM products do not include tide, the tidal forcing by NAO.99Jb tidal solution (Matsumoto et al., 2000) was added to the boundaries produced by Global HYCOM products. Atmospheric forcing datasets are used Dynamical Regional Downscaling Using the JRA-55 Reanalysis (DSJRA-55; Kayaba et al., 2016) from 1994 to 2006, and Japan Meteorological Agency’s Meso-Scale Model Grid Point Value (MSM-GPV) from 2006 to 2015. To incorporate the impact of freshwater discharge from the terrestrial side of Kikai Island on Shiomichi Bay, the rates of river discharge and submarine groundwater discharge were estimated based on field measurement and simulation by the Soil and Water Assessment Tool Plus (SWAT+), and these rates were input into the KIKAI2 domain as freshwater point sources. By using this multi-nested model, we can obtain a long-term (1994 to 2015) high-spatiotemporal-resolution dataset of temperature, salinity and water velocity around the Kikai Island area.
These long-term temperature data were used to calibrate the Sr/Ca of a coral core collected in Shiomichi Bay, Kikai Island. Due to the seasonal variation in skeletal growth rate in higher latitude corals, it is generally difficult to reproduce correct seasonal water-temperature patterns from the Sr/Ca of the coral core. To overcome this difficulty, the dependence of growth rate on temperature was elucidated, and a conversion function from Sr/Ca data with distance from core-top and coral growth rate to temperature with time stamp was developed. Using this conversion function, a more accurate time series water temperature with seasonal patterns reproduced from coral Sr/Ca was successfully obtained.