mingkaijiang / quasi_equil_analytical Goto Github PK

Quasi-equilibrium analysis of ecosystem C, N and P interactions

License: GNU General Public License v3.0

R 67.32% Python 6.35% Makefile 0.17% C 23.29% C++ 2.87%

quasi-equilibrium-analysis ecosystem simulation approximation-algorithms

quasi_equil_analytical's Introduction

Quasi-equilibrium analysis of plant response to eCO2

This is a repository for evaluating carbon response to eCO2 by incorporating different model assumptions and evaluating their consequences dynamically and analytically. More background information please read McMurtrie and Comins (1993).

Code structure:

Script Run_prorgams.R contains the master-level executable commands. Folder Functions includes all analytical functions and the post-processing scripts Folder GDAY includes numerous sub-folders:

analyses include all the analyses output
code include the simplified GDAY code (in language C)
met_data contains driving met data
outputs contains all the gday simulation output data
params contains the parameter files
post_processing contains the R code to post-process the simulations
pre_processing contains the R code to pre-process the simulations (e.g. generating met data)
simulations contains the gday executive program and the python code to run the simulations

GDAY model framework

Key references:

(more to be added) Quasi-equilibrium analysis: Comins and McMurtrie (1993) Kirschbaum et al. (1994) Kirschbaum et al. (1998) etc.

Key model assumptions: Reich et al. (2008)

GDAY model:

Contact Mingkai Jiang for more information ([email protected]).

quasi_equil_analytical's People

Contributors

Stargazers

Watchers

Forkers

datafusion18

quasi_equil_analytical's Issues

import pandas

When running gday, the import pandas as pd returns error:

Traceback (most recent call last):
File "./quasi_equil_annual_spin_up.py", line 236, in
main(site, SPIN_UP=True)
File "./quasi_equil_annual_spin_up.py", line 230, in main
import translate_GDAY_output_to_NCEAS_format as tr
File "/Users/mingkaijiang/Documents/Research/Projects/Quasi-equilibrium-test/Git/GDAY/code/scripts/translate_GDAY_output_to_NCEAS_format.py", line 15, in
import pandas as pd
ImportError: No module named pandas

same result with and without CWD pool

Problem description:
plant biomass and productivity are not affected by the additional CWD pool. In theory, cwd pool should store some nutrient so plant productivity should decrease. By checking gross mineralization alone, nutrient is indeed reduced, but net mineralization remains the same. I think this is the problem of using SOM NC and PC ratios to control the litter release rates. Need to investigate in a more detailed manner.

N uptake ~ root biomass and mineral N

O-CN and GDAY assumes different relationships to control the N uptake response. O-CN assumes that N uptake is a saturating function of mineral N, whereas GDAY assumes it is a saturating function of biomass. The former has been implemented in analytical solution. The latter is a bit more complicated because it yields a 2nd order polynomial equation (i.e. two possible NPP values).

Add folder structure diagram

To add a folder structure diagram for illustration purpose in readme file.

.DS_Store file

To delete all the .DS_Store files

Different programming language

Not an issue per se, but need to (at least try to) come up with a cohesive coding system that links all different parts of the projects together:
GDAY in C;
wrapper file in python;
analytical solution and post-processing scripts in R;
Also need to use linux language to make the gday files.

clean the process

Several steps can be improved:

GDAY simulation output can be translated on the run by calling a translation function; currently I am storing all the gday output and then translate from monthly into annual, very wasteful of disk space and time.
GDAY simulations in each sub-directory can use parallel run to speed up the process.

Slow initialization in GDAY

Even initiating GDAY using some equilibrium values doesn't warrant a faster spin-up process, need to check if the cfg file misses anything, or try to solve why the spin-up process takes such a long time to finish (>15,000 model years).

analytical solution for priming effect

The analytical solution for priming effect is not outlined yet (i.e. equations not derived). However, root exudation has been put into the analytical solution. Need to test the sensitivity of the quasi-equil behavior to various parameters assumed in the analytical solution (i.e. omegae, omegaw, ae and vx). Also, the priming effect needs a medium term.

Reverse calculation of pf from G for long-term constraint

The analytical solution between photosynthetic constraint and Nitrogen constraint gives a long-term equilibrium G term (production). In theory, there exists a phosphorus level (pf) that co-limits production. Currently, an implementation was done to use G constrained by photosynthesis and N cycling to back calculate leaf P for both very long-term and long-term. This yields two different leaf P levels for the same G, and yet I don't have a better solution.