Convergency issue: soil moisture, soil water potential and leaf water potential calculation

[Crystal-szj]

Hi @MostafaGomaa93 @prajzwal08 @Yunfei-Wang1993 I listed the problems we found:
First, we found the transpiration (Trans) and the root water uptake (RWU) are not always the same.
Second, under extremely dry conditions, there is always a lot of transpiration even if there is no precipitation (or soil moisture).

• complex value of Ksoil when the leaf water potential didn't reach to convergence

[prajzwal08]

@Crystal-szj , @Yunfei-Wang1993 , @MostafaGomaa93 There is a complex value appearing in the calc_rsoil function when Se becomes negative—this happens when soil moisture content (SMC), i.e., SoilVariables.Theta_LL(1:end - 1, 2), drops below the residual soil moisture content $\theta_{r}$. We need to check the update process for SoilVariables.Theta_LL, ensuring it does not fall below $\theta_{r}$.

Another issue is that leaf water potential (PSI) should only be initialized to zero at the start and then remain a continuous value. I’ve implemented this in my local run. @MostafaGomaa93 and @..., I will make sure to push the changes after resolving the issue with negative Se.

[bobzsu]

Hi all, as discussed at last meeting, the van Genuchten equation does not work and is not valid when the soil moisture is below the residual soil moisture. Therefore you need to set it to the residual soil moisture if an iteration gives a lower value. Otherwise a wrong value, in this case a negative one, is given for soil water potential. The equation is a curve fitting that going to infinity if the soil moisture approaches ZERO. I suggested to find another function based on the Maxwell-Boltzmann distribution for particle size and link it to ideas in quantum physics - nothing in physical world can be infinity.