Update hydraulic failure mortality for places with frozen soils

[JessicaNeedham]

This PR address issue #1492.
Hydraulic failure mortality (hmort) occurs when btran is less than a soil moisture threshold. If soils are frozen then btran is zero. To prevent hmort of plants at high latitude during winter there is an if statement that prevents hmort being calculated if any soil layer is frozen. However, in regions with permafrost this essentially turns off hmort entirely, even in summer months when plants may actually be water stressed.

This PR adds additional logic so that we still calculate hmort if soil layers with 90% of the root biomass are not frozen.

Collaborators:

@mvdebolskiy @rosiealice @kjetil

Expectation of Answer Changes:

This will change hmort rates of plants at high latitudes.

Checklist

If this is your first time contributing, please read the CONTRIBUTING document.

All checklist items must be checked to enable merging this pull request:

Contributor

• The in-code documentation has been updated with descriptive comments
• The documentation has been assessed to determine if updates are necessary

Integrator

• FATES PASS/FAIL regression tests were run
• Evaluation of test results for answer changes was performed and results provided
• FATES-CLM6 Code Freeze: satellite phenology regression tests are b4b

If satellite phenology regressions are not b4b, please hold merge and notify the FATES development team.

Documentation

• Technical Note update:
• User's Guide update:

Test Results:

Not tested

[JessicaNeedham]

I'm wondering if there should be a new function that gets the maximum soil layer index with 90% of root biomass so that we don't have to duplicate that chunk of code before each call to mortality_rates?

[JessicaNeedham]

For reference, here is global btran by season from a recent historic run.

DJF:

MAM:

JJA:

SON:

[JessicaNeedham]

Here are differences from a 5 year run (with hmort scalar set to 0.6 for all PFTs). Changes don't seem to be having a huge impact.

[JessicaNeedham]

We could also simplify things a lot by evaluating only the active layers. i.e. calculate hmort if

( ( minval(bc_in%t_soisno_sl(1: bc_in%max_rooting_depth_index_col) - tfrz ) > soil_tfrz_thresh ) ) then

[JessicaNeedham]

Here's fineroot carbon by soil layer (historic run - mean of last year)

And active layer thickness by month (1850 simulation - only one where this was turned on)

[JessicaNeedham]

Using active layer thickness is much simpler and seems to be the same as calculating the maximum soil layer with 90% of the root biomass. These are 5 year runs from bareground - with hmort scalar set to 0.6 for all PFTs.

[JessicaNeedham]

A 50 year run with the default parameter file (fates_mort_scalar_hydrfailure = 0.1 for all PFTs) shows little difference in hydraulic failure mortality from the current logic (mean over last ten years). Mostly because there doesn't seem to be much hydraulic failure mortality at high latitudes anyway with this parameter file and forcing.

[mpaiao]

@JessicaNeedham thanks for submitting this PR! I think your implementation makes total sense and is simple and elegant. It is also consistent with how btran is calculated.

[JessicaNeedham]

Thanks @mpaiao ! I am also testing a version now where only 50% of the active layer need to be thawed before we allow hydraulic failure. Discussing with @rosiealice we thought maybe the whole active layer having to be thawed was a bit restrictive.

[JessicaNeedham]

With the condition that we calculate hydraulic failure when 50% of the soil layers in the active layer are thawed (~0.6 m with CLM soil layers) we see increased mortality - especially in NE China, SE Russia. Cold-deciduous broad leaf and needleleaf tree PFTs.

Here are changes in LAI as a result:

[rgknox]

@JessicaNeedham , could we also perform this logic based off of the fraction of non-frozen depth, or the fraction of non-frozen fine-root biomass? That might buffer this calculation from artificial model constructs like how we decide on layering.

[JessicaNeedham]

@rgknox yes it also occurred to me we could get quite different behaviour here if the HLM layering scheme changed. I might have missed something, but it looked to me like FATES doesn't actually see the thickness of each HLM soil layer, which makes it hard to do this by depth, rather than number of soil layers. Making it by root biomass works, and that was what I originally wrote. It is a bit clunkier but maybe safer (see this first commit).

[rgknox]

Recommend setting 0.5 to a named constant like "hmort_thaw_frac_threshold" or whatever you deem appropriate

[JessicaNeedham]

@rgknox - this is back to being based on root biomass. Tests running with NorESM.

[JessicaNeedham]

I ran two simulations where soil layers with either 25% or 75% of root biomass thawed before hydraulic failure mortality is allowed. All PFTs had hydraulic failure mortality set to 0.25. There is very little difference except for at the highest latitudes.

Hydraulic mortality rate:

Hydraulic mortality carbon flux: