2:00 PM - 2:15 PM
[AOS16-02] Improvement of Carbon River Inflow Processes in Tokyo Bay through Satellite Data Assimilation into Ecosystem Models
Keywords:Data assimilation, Carbon cycle, GCOM-C/SGLI, Tokyo bay, Remote sensing reflectance
Coastal areas adjacent to urban areas, such as Tokyo Bay, are heavily influenced by human activities, which flow into these regions through rivers and have a significant impact on the circulation of materials within the bay. Tokyo Bay receives inflows from several rivers with relatively high discharge, such as the Arakawa and Tama Rivers. Therefore, it is crucial to understand the amount and dynamics of organic matter inflows from rivers to elucidate the carbon cycle in Tokyo Bay.
However, in many inner bay basins there is a paucity of data on water quality during high flow events compared to normal conditions, leading to significant uncertainties in the estimation of river loads during large flow events.[1] Furthermore, indirect methods based on water level-discharge curves, which are most used to estimate river discharge, have an uncertainty of about 13%.[2]
In this study, we developed a data assimilation method aimed at improving the accuracy of river boundary conditions during flood events by using the surface observations of Tokyo Bay obtained from the high-resolution ocean colour satellite GCOM-C/SGLI as input to a non-hydrostatic three-dimensional flow-ecosystem model (Ecological hydrodynamic simulation system by PARI: EcoPARI) [3][4][5] .
Since the surface concentration and extent of organic carbon-bearing suspended sediment inputs from rivers can be derived by satellites, we can expect estimation of riverine carbon loads by satellite data assimilation becomes feasible. However, in coastal areas, variations in in-water inherent optical properties (IOP) due to river plumes and internal production lead to significant uncertainties in the water quality estimation. Therefore, in this study, we incorporated a simplified in-water radiative transfer model into an ecosystem model to simulate the in-water optical environment, and used remote sensing reflectance (Rrs) obtained by satellite observation after the atmospheric correction as a model constraint. This approach reduces the uncertainty associated with satellite water quality estimation in the complex water and establishes a more versatile and robust data assimilation method.
In the accuracy validation using observation data set of public water areas, the estimation accuracy of total organic carbon (TOC) and dissolved organic carbon (DOC) concentrations in the estuarine region improved. This suggests that the data assimilation method developed in this study is effective in improving river boundary conditions in coastal areas.
[1] Ayako SAKAI, Yasuo NIHEI, Keisuke EHARA, Miho USUDA, Kyosuke SHIGETA & Satoshi OOTSUKA. (2008). NUTRIENT AND COD LOADS IN THE EDO, ARA, TAMA AND NAKA RIVERS UNDER FLOOD FLOW CONDITIONS. PROCEEDINGS OF HYDRAULIC ENGINEERING, 52, 1117-1122.
[2] Junzo Sago. (2008). Comprehensive Assessment of Hydrological Uncertainty and Its Utilization Strategies. J. Japan Soc. Hydrol. And Water Resour., 21(5), 353-360.
[3] Hafeez, M. A., Nakamura, Y., Suzuki, T., Inoue, T., Matsuzaki, Y., Wang, K., and Moiz, A. (2021): Integration of weather research and forecasting (WRF) model with regional coastal ecosystem model to simulate the hypoxic conditions. Science of the Total Environment, 145290.
[4]Hafeez, M. A., and Inoue, T. (2021): Determination of flow characteristics of Ohashi River through 3-D hydrodynamic model under simplified and detailed bathymetric conditions. Water, 13, 3076.
[5]Matsuzaki, Y., and Inoue, T., 2022: Perturbation of boundary conditions to create appropriate ensembles for regional data assimilation in coastal estuary modeling. Journal of Geophysical Research-Oceans, 127(4), e2021JC017911.
However, in many inner bay basins there is a paucity of data on water quality during high flow events compared to normal conditions, leading to significant uncertainties in the estimation of river loads during large flow events.[1] Furthermore, indirect methods based on water level-discharge curves, which are most used to estimate river discharge, have an uncertainty of about 13%.[2]
In this study, we developed a data assimilation method aimed at improving the accuracy of river boundary conditions during flood events by using the surface observations of Tokyo Bay obtained from the high-resolution ocean colour satellite GCOM-C/SGLI as input to a non-hydrostatic three-dimensional flow-ecosystem model (Ecological hydrodynamic simulation system by PARI: EcoPARI) [3][4][5] .
Since the surface concentration and extent of organic carbon-bearing suspended sediment inputs from rivers can be derived by satellites, we can expect estimation of riverine carbon loads by satellite data assimilation becomes feasible. However, in coastal areas, variations in in-water inherent optical properties (IOP) due to river plumes and internal production lead to significant uncertainties in the water quality estimation. Therefore, in this study, we incorporated a simplified in-water radiative transfer model into an ecosystem model to simulate the in-water optical environment, and used remote sensing reflectance (Rrs) obtained by satellite observation after the atmospheric correction as a model constraint. This approach reduces the uncertainty associated with satellite water quality estimation in the complex water and establishes a more versatile and robust data assimilation method.
In the accuracy validation using observation data set of public water areas, the estimation accuracy of total organic carbon (TOC) and dissolved organic carbon (DOC) concentrations in the estuarine region improved. This suggests that the data assimilation method developed in this study is effective in improving river boundary conditions in coastal areas.
[1] Ayako SAKAI, Yasuo NIHEI, Keisuke EHARA, Miho USUDA, Kyosuke SHIGETA & Satoshi OOTSUKA. (2008). NUTRIENT AND COD LOADS IN THE EDO, ARA, TAMA AND NAKA RIVERS UNDER FLOOD FLOW CONDITIONS. PROCEEDINGS OF HYDRAULIC ENGINEERING, 52, 1117-1122.
[2] Junzo Sago. (2008). Comprehensive Assessment of Hydrological Uncertainty and Its Utilization Strategies. J. Japan Soc. Hydrol. And Water Resour., 21(5), 353-360.
[3] Hafeez, M. A., Nakamura, Y., Suzuki, T., Inoue, T., Matsuzaki, Y., Wang, K., and Moiz, A. (2021): Integration of weather research and forecasting (WRF) model with regional coastal ecosystem model to simulate the hypoxic conditions. Science of the Total Environment, 145290.
[4]Hafeez, M. A., and Inoue, T. (2021): Determination of flow characteristics of Ohashi River through 3-D hydrodynamic model under simplified and detailed bathymetric conditions. Water, 13, 3076.
[5]Matsuzaki, Y., and Inoue, T., 2022: Perturbation of boundary conditions to create appropriate ensembles for regional data assimilation in coastal estuary modeling. Journal of Geophysical Research-Oceans, 127(4), e2021JC017911.