UCSC Ocean Modeling and Data Assimilation Website
The Ocean Modeling Group at UC Santa Cruz is involved with many research projects focused on a variety oceanographic problems. Although our main focus is on physical oceanography, we also work on many interdisciplinary activities involving chemical and biological oceanography. Over time, this website will be populated with sections describing the people in the group and ongoing projects. At the present time we have links to our near-real-time ocean hindcast/nowcast system, and our historical reanalyses, both of which make available output from our data-assimilative ocean model for the California Current System.
A general description of the various models developed and implemented by the UCSC Ocean Modeling team. All models described below use the Regional Ocean Modeling System (ROMS).
Data assimilative analyses
1) A set of two data assimilative reanalyses for the California Current System (CCS) have been developed and are referred to as CCSRA31 and CCSRA13. The numbers 31 and 13 refer to the number of years they span (1981-2010 and 1999-2011, respectively), and the two models differ in the atmospheric products used to force the model CCSRA31 uses output from the European Center for Medium Range Weather Forecasting (ECMWF), specifically ERA3 and ERA Interim products, whereas CCSRA13 is forced by a multi-nested configuration of the Coupled Ocean Atmosphere Mesoscale Prediction System (COAMPS). Boundary conditions derive from the Simple Ocean Data Assimilation (SODA) product. State estimates of the ocean physical state (temperature, salinity, velocity, and sea surface height) are based on 4-Dimensional Variational Assimilation (4D-Var) methods, using overlapping 8-day cycles and assimilating sea surface temperature (SST), sea surface height (SSH), and in situ temperature (T) and salinity (S). Model initial conditions, boundary conditions, and surface forcing are adjusted. The model domain spans Baja, California, to just south of the Strait of Juan de Fuca and offshore to 134 W at 1/10 degree resolution (approximately 10 km), with 42 terrain-following vertical levels. Model output is served through the UCSC THREDDS server. Descriptions of the assimilation can be found in Moore et al. (2011) with description of the product and underlying analysis in Neveu et al. (2016).
2) A data assimilative estimate of physical ocean state is computed in near real-time (NRT) from 2010 to the present on the same grid as the CCSRA runs and using 4D-Var with model initial conditions updated on 4-day assimilation cycles. This product recently underwent revision and therefore consists of model configurations.
a. 2010 – 7/30/2023: Boundary conditions for this model are obtained from the Hybrid Coordinate Ocean Model (HYCOM), and surface forcing derives from COAMPS as with CCSRA13. In addition to SSH, SST, and in situ T and S, this product also assimilates surface ocean currents derived from High Frequency RADAR (HFR) observations. Model output is served on the UCSC THREDDS server. A description of the early implementation of this product can be found in Moore et al. (2013).
b. 7/31/2023 – ~2 months prior to present (i.e., delayed time): Boundary conditions are obtained from the GLORYS reanalysis product served by the Copernicus Marine Environmental Monitoring Service (CMEMS), and surface forcing derives from the ERA5 product from the European Center for Medium Range Weather Forecasting center (ECMWF). This model assimilates SSH, SST, SSS, in situ T and S, and surface ocean currents derived from High Frequency RADAR (HFR) observations. This product is delayed relative to the present due to delays in the GLORYS reanalysis product availability and Q/C’d in situ T and S through the regional moorings, gliders, and the global EN4 database served by the UK Met Office. Model output is presently served on the UCSC THREDDS server as a continuous product with (a), though there is a discontinuity on 7/31/2023.
c. Previous ~2 months prior to present – present (i.e., near real-time): Boundary conditions are obtained from the GLORYS analysis-forecast system and surface forcing derives from the North American Mesoscale Model (NAM). This model assimilates SSH, SST, in situ T and S, and surface ocean currents derived from High Frequency RADAR (HFR) observations. In situ T and S are obtained from regional moorings, regional gliders, and the Argo program. Model output is presently served on the UCSC THREDDS server as a continuous product with (a), though there is a discontinuity on approximately 2 months prior to the present, when the delayed time product (b) supersedes and replaces this output.
3) A version of the NRT system (a) that includes biogeochemical fields is also available for the period 2010-2023. This model is on the same 1/10 degree resolution grid as CCSRA and the NRT physical system. The biogeochemial model is based on the North Pacific Model for Understanding Regional Oceanography (NEMURO). It includes two phytoplankton components, three zooplankton components, nitrate, ammonium, silicate, dissolved and particulate nitrogen and particulate silica. Satellite estimates of surface chlorophyll are assimilated using 4D-Var in 4-day assimilation cycles with an assumed constant carbon to chlorophyll ratio. State estimates from this product are accessible on the UCSC THREDDS server. A description of the methodology used in this product can be found in Song et al. (2016a, 2016b, and 2016c) as well as Mattern et al. (2018).
Non-data assimilative models, also referred to as forward and/or nonlinear models
1) Physical-biogeochemical hindcast (1995–2020) at 1/30° (~ 3 km) horizontal resolution for the U.S. west coast 30–48°N. Physical properties are generated using ROMS and biogeochemical properties are generated using a version of the NEMURO model that has been adapted at UCSC and is called NEMUCSC. Like NEMURO, this model includes three limiting macronutrients (nitrate, ammonium, and silicic acid), two phytoplankton functional groups (nanophytoplankton and diatoms), three zooplankton size-classes (microzooplankton, copepods, and euphausiids), and three detritus pools (dissolved and particulate organic nitrogen and particulate silica) (Fiechter et al., 2018, 2020), and the NEMUCSC version also simulates carbon and oxygen cycling (Cheresh and Fiechter, 2020; Cheresh et al., 2023). This ROMS hindcast is forced at the surface with atmospheric fields from the ERA5 reanalysis and at the model open ocean boundaries by the GLORYS global reanalysis. NEMUCSC is similarly forced by daily averaged shortwave solar radiation from ERA5 and the GLORYS global hindcast for nutrients and dissolved oxygen at the open ocean boundaries. A subset of the coupled model output is available on the California Ocean Observing Systems data portal (https://data.caloos.org/#module-metadata/fb6cf7e9-eab9-4eae-aaee-26e93f9cc326)
2) Since 2022, the National Oceanographic and Atmospheric Administration (NOAA) has produced daily 4D-Var state estimates for the physical ocean properties within the CCS, referred to as the West Coast Operational Forecast System (WCOFS). The model domain is considerably larger than the UCSC CCSRA domain, and at higher resolution (4 km), with 40 vertical terrain-following levels. UCSC has built two high resolution nests with 5x multipliers to obtain a 160 m grid focused on the Monterey Bay and surrounding regions. The first nest is driven at the boundaries by the WCOFS product that is rerun locally, and the three domains are linked in a two-way nesting configuration in which information propagates in both directions. While WCOFS is data assimilative, the Monterey Bay domain is nearly free running only constrained by a weak relaxation to WCOFS fields. This product is served on the UCSC THREDDS server. At present, there is no publication that describes its features.
3) Various other forward models and configurations have been produced for different research projects and purposes. These are described briefly on the Projects tab.
We gratefully acknowledge financial support for various aspects of our research from the Central and Northern California Ocean Observing System (CeNCOOS), the National Oceanographic Partnership Program (NOPP) , the Office of Naval Research (ONR), and the National Science Foundation (NSF), the National Oceanographic and Atmospheric Administration (NOAA) , the Gordon and Betty Moore Foundation and the Simons Foundation.