Capping infiltration seems not to account for rain or flows from neighboring cells

[lrntct]

Original report by pirmin borer (Bitbucket: PBorer, ).

---

In line 398 and 423 the infiltration rates are capped with the method cap_inf_rate(float dt_h, float h float infrate)

It seems that the capping of the infiltration rate does only account for the maximum of waterdepth which is available for infiltration. It does not account for the rain which falls during that time-step, neither the water which will flow from other cells. This has the effect that the amount of water is only correctly calculated where enough ponding occurs such that the infiltration is not limited by the water height. Actually the infiltration is twice capped, once by the method cap_inf _rate but also in apply_hyrology() where the new water height is set to :

h_new = max(arr_h[r, c] + (rain - inf - etp - capped_losses) * dt_h / 1000., 0.)

The effective infiltration rate could be calculated by

(1) arr_inf_out[r,c] = max(1000 * arr_h[r,c] / dt_h + qn * dx + qs * dy + rain - etp - inf- capped_losses, 0)

It is unclear to me if at the point of applying the hydrology, the water flow from other cells already has been added to the array arr_h. If so, then qn*dx + qs *dy are not necessary in (1)

This is just an idea of how the inflow from rain and other cells could be used to calculate the effective infiltration rate. What do you think?

Thank you for your answer.

[lrntct]

I believe this issue could be resolved by reducing the hydrology time-step. If hydrology is computed at every timestep, then the last step inflow from neighbouring cells will be taken into account. Which might be closer to the reality.