Model changes
While the previous implementation of tidal hydrology for a microtidal marsh used a sinusoidal function to represent tidal variations, modeling vegetation responses to tidal cycles across a broader range of coastal systems requires the ability to drive the model with site-specific tidal hydrology and salinity concentrations. We added the capability to read in an external forcing file containing a time series of water level heights and salinity concentrations (add github link?). The tide file may be created with measured water levels and salinity concentrations, water levels calculated from tide constituents, or hypothetical values for scenario simulations, thereby facilitating flexibility in simulating different site conditions. This approach builds the groundwork for coupling ELM to water level and salinity boundary conditions determined by ocean and river components of E3SM in future coupled simulations.
To represent the influence of salinity on vegetation growth, we used a Gaussian function (Equation 2, Li et al., 2021)
\(f_{i}(s)=\ e\ \frac{{-(s-\mu)}^{2}}{{2y}^{2}}\) Equation 2
In this function, s represents salinity, µ represents the salinity concentration at which maximum growth occurs (optimal salinity for a PFT), and ʎ represents the salinity tolerance. The functionfi(s) yields a value between 0 and 1, andfi(s) is multiplied by w­ii for each layer to decreaset proportionately to the salinity concentration (Fig 1a). ELM does not represent salinity variations with depth in this study; the model is forced with a single salinity concentration applied to the entire marsh column, assuming that salinity in the marsh is equal to that in the tidal channel.
To represent the effects of flooding on vegetation growth, we used a linear function that limits carbon uptake by the proportion of vegetation that is submerged (Equation 3).
\(f_{i}\left(z\right)=\left\{\par \begin{matrix}0,\ \ z\geq H\\ \frac{H-z}{H},0<z<H\\ 1,z\leq 0\\ \end{matrix}\right.\ \) Equation 3
The inhibition of growth due to inundation,fi(z) , is calculated by the plant height Hand the surface water height z . Water levels above the plant height set fi(z) to zero, while water levels below the marsh surface set fi(z) to 1. When 0 < z < H , the function inhibits root water uptake in proportion to the fraction of plant height that is submerged (Fig 1b). As with the salinity response,fi(z) is multiplied by the root water uptake resistance in each layer to represent the impact on root function. The improved model thus represents the soil water stress via a transpiration function that includes salinity and flood stress (Equation 4).
\(\beta_{t}=\ \sum_{i}{w_{i}\text{\ r}_{\text{i\ }}f_{i}\left(s\right)}f_{i}(z)\)(Equation 4)