Urgent Assistance Request: Unusual Elevation of Nutrients in FVCOM-FABM-ERSEM Model Simulations #148
Description
Hi everyone,I am writing to seek your expert guidance regarding a critical issue encountered in my recent numerical simulations using the coupled FVCOM-FABM-ERSEM model. Specifically, I have observed unexpected and persistent elevations in key nutrient concentrations (nitrate, phosphate, and silicate ) throughout multiple model runs, which deviate significantly from both observational data and expected biogeochemical dynamics. This anomaly is undermining the reliability of my results, and I am eager to resolve it with your insights.
Background & Model Setup
To provide context, my model configuration is designed to simulate coastal biogeochemical processes ina small semi closed bay at a spatial resolution of 100 m and temporal resolution of "0.9 s"(EXTSTEP_SECONDS = 0.9, ISPLIT = 10). The setup includes:
FVCOM version [ "4.3"] for hydrodynamic simulations, forced by [ "wrf atmospheric data,and tidal boundary conditions from TPXO9,and temperature and salinity boundary conditions from ROMS model"].
The in-situ measurements of nutrient concentrations and river discharge was not used yet.Because I want to figure out the reason behind the observed increase in nutrients by the model simulation, all external nutrient sources were disabled. This was achieved by deactivating river forcing (RIVER_NUMBER = 0) and turning off boundary nudging for FABM variables (OBC_FABM_NUDGING = F).
Model Validation: The performance of the FVCOM simulation was evaluated against the in-situ data, and the model demonstrated a satisfactory fit.
FABM-ERSEM (v[ fabm-1.0.3) as the biogeochemical module, with parameterizations forphytoplankton, zooplankton, organic matter decomposition, and nutrients. Utilized the fabm-ersem-15.06-L4-noben-docdyn-iop.yaml configuration file from the testcases folder to execute the FVCOM-FABM-ERSEM model (did not change any parameters there), with only two additional parameter instances added: pCO2a (set to 420) and silt_absorption (set to 0.01). The corresponding YAML and NML configuration files are attached.
Cold start was selected.
Anomaly
The nutrient elevation manifests as follows:
Magnitude & Spatial Pattern: Nitrate concentrations up to 3000 mmol P/m^3μmol L⁻¹ are 3000 times higher than observed climatology across the entire domain, with particularly strong signals near the boundary. total phosphorus increased from 210^10 to 2.510^13. During the simuliation, Phosphate, nitrate and silicate show similar trends. (I attached the outputs of the model here ).
Impact on Biogeochemistry: Elevated nutrients coincide with dissolved inorganic carbon O3c decreased from 2200 (mmol/m^3) to 0 (mmol/m^3). ["Diatom chlorophyll a varies from 0 to 45 mg m⁻³)"]
Troubleshooting Steps Taken (Without Success)
To diagnose the issue, I have systematically tested potential causes, including:
· Hydrodynamic Drivers: Verified flow patterns, mixing depths, and advection using tracer experiments (e.g., passive dye release), which showed reasonable transport timescales consistent with observations. The temperature, salinity, ocean speed and direction is ok.
· Initial/Boundary Conditions: Sensitivity tests with reduced nutrient concentrations at boundaries/rivers "insitu measured data, lowered by 25% , 50% , 100% , and completely turn off"] , but the nutrients still keep increasing. O3c keep decreasing in all simulation.
Is anyone have any idea why the nutrients keep increasing even without any nutrients input. Could this arise from "incorrect parameterization of sediment nutrient release, or incorrect parameterization in the ERSEM remineralization subroutine"? I tried changed sR1( the maximum turn-over rate of DOM (1/d), default = 1.0 ) as 0.5, but the nutrients still go up.
1. Diagnostic Tools: Are there specific FABM-ERSEM diagnostics (e.g., tracer budgets, process-specific flux terms) or FVCOM output variables (e.g., vertical eddy diffusivity) that would help pinpoint the source?
2. Literature/Case Studies: Are there known instances of similar nutrient imbalances in FVCOM-FABM-ERSEM applications, and how were they resolved?
Supporting Materials
I have attached:
· Some [png files] with time-series plots of nutrient concentrations at key stations and spatial maps of the anomaly.
· A [v17_run.nml] summarizing model configuration details (parameters, forcing data, and boundary conditions), and fabm.yaml summarizing the parameters in ERESM model.
Thank you sincerely for your time and expertise. Resolving this issue is critical for my ongoing research on coastal eutrophication and climate change impacts, and I am happy to provide additional information or run specific sensitivity tests at your suggestion.
I look forward to your response and am available for a virtual meeting at your convenience.
Best regards,
Have a nice day
chao