Secondary Forest Tracking and Fixes to External Seed Rain

This release tag contains a feature and a bug fix. Charlie Koven added the ability to discretely track both primary and secondary forest patches, paving the way to a broader expansion of human dimensions interfacing with FATES. Rosie Fisher implemented a fix to correctly account and track external seed input.

Fixes to leaf area and demotion/promotion. Crown area conservation during fusion

This tag contains a set of changes oriented around achieving stable long-term simulations. This includes a feature addition from Charlie Koven that maintains conservation of crown area during cohort fusion. This also includes a set of fixes: 1) demotion and promotion logic, 2) litter fluxes during demotion termination, 3) ensuring minimum patch areas are being imposed, 4) ensuring that fusion and numerical imprecisions to not generate massive LAI via adjusting target plant sizes and 5) reverting to 30 leaf layers and a 2 layer canopy by default. Fixes were implemented by Ryan Knox, Charlie Koven, Jackie Shuman and Greg Lemieux.

Carbon Isotope and Updated Photosynthesis Smoothing

This release adds a new feature in tracking carbon isotopes, provided by Liang Wei. Rosie Fisher, Anthony Walker, Alistair Rogers and Jackie Shuman also teamed up on an update to the smoothing function in the photosynthesis scheme. Charlie Koven also provided updated in-line text and descriptions for how the history diagnostics are generated.

Leaf age discretization within cohort

This release contains a feature that discretizes the leaf biomass pool into age classes within each cohort. This makes use of the vector enabled memory structures in parteh. The discretization (binning) is controlled by the user through the parameter file. vcmax and leaf longevity are now 2d parameter arrays containing the leaf age dimension. Leaf longevity of each bin defines the turnover time governing transfer between bins. This feature was developed by Chonggang Xu, Ryan Knox, Shawn Serbin, Liang Wei, Charlie Koven, Rosie Fisher and the NGEET development team.

Height based sorting logic and modified static array sizes

This release updates the logic to the cohort sorting logic by using height instead of diameter for cohort order. Along with this, the way ranked promotion and demotion was handled, was updated to allow cohorts of the same height to be treated equally. Static array sizes for maximum cohorts, maximum numbers of canopy layers and maximum number of leaf layers were adjusted to 100, 3 and 20 respectively (from 160, 2 and 40). These changes were mostly implemented by Charlie Koven and Ryan Knox.

Ecosystem dynamics compatible FATES-HYDRO

This release provides the core functionality that enables FATES-HYDRO to work with a dynamic ecosystem. This functionality accounts for the fate of water in dying plants, growing plants, and newly recruited plants. This release also has added in a first pass at restart capabilities with FATES-HYDRO, although initial testing shows restarts with dynamic FATES-HYDRO is incomplete. Also note, that while FATES-HYDRO has shown stable and publication worthy results thus far, this is still considered unsupported, "beta level" code. Those using it should consider it still under development as of this release, and should consult with Chonggang Xu and Bradley Christoffersen. Authors are Chonggang Xu, Bradley Christoffersen and Ryan Knox, with consultation from Yilin Fang, Junyan Ding and others on the FATES development team.

Fixes to fire mortality diagnostics, patch disturbance area checks and trimming

This release adds a modest set of feature changes and several small bug fixes. The small feature additions improve the way that mortality is calculated for fire, by saving these diagnostics to a tracking array. The bug fixes are as follows: 1) Many site level diagnostics were not being saved in restarts, thereby the first day following a restart was contributing bogus information to the diagnostics, 2) the net assimilation budget had not been correctly calculating night-time respiration for trimming calculations (the prognostic carbon balance was not affected) , and 3) in very rare cases, patches with nearly zero sizes were creating problems during patch generation and creating problems during carbon balance checks, so checks were implemented. Contributions from Jackie Shuman, Charlie Koven, Ryan Knox, and others.

Remove Lasslop wind effects on fire Rate of Spread

This release contains contributions from Jackie Shuman. Per her pull request:
Within SFMain remove the conditional clause that reduces fire rate of spread for high wind speeds. Within parameter files removed the parameter for fire_wind_max that triggers the conditional.
This showed an improvement in mode simulated fire behavior in past work (Lasslop et al 2014), but current research demonstrates there is not a clear link between high winds and a reducing effect on rate of spread. Andela et al 2018 ESSDD https://doi.org/10.5194/essd-2018-89
This PR acts to remove this Lasslop et al (2014) reducing effect, and revert back to original Thonicke ROS wind effect.

Provisions for Coastal Grasses

This tag contains key changes that enable simulation of a coastal salt marsh. Key features are 1) an allowance that when maximum height/dbh reached, all the carbons goes to the bstore for grasses; 2) when plant dies, bstore goes to reproduction for grasses; 3) recruitment only happens at the growing seasons so that the newly developed cohorts are viable; 4) add a new capability to simulate the impact of salinity and sulphide through btran; and 5) a
new parameter file for Spartina.
These changes were developed by Joel Rowland and Chonggang Xu, with help from Rosie Fisher, Charlie Koven and Ryan Knox.

Hydro, Recruitment and Radiation Refactors

This tag contains mostly updates, bug fixes and refactors. Chongang Xu and Yilin Fang submitted updates to FATES-hydro related to parameters and commenting. This also included a pass through the code to remove calls to the dble() intrinsic (replaced by real(,r8)). Charlie Koven added a diagnostic that tracks recruitment across size classes. Ryan Knox refactored the restart system to bypass the saving of 3D arrays, dimensioned by pft x canopy-layer x leaf-layer.