Verification of carbon decomposition (DAYCENT)

Inés Delsman Valderrama (CREAF), Miquel De Cáceres (CREAF)

2026-05-12

Introduction

This document presents an exercise to verify the implementation of the carbon decomposition process of DAYCENT into MEDFATE. The verification is done against the outputs of CENTURY (ver. 4.7) corresponding to simulation of Duke and Harvard forests.

Methods

Initial soil carbon content

• Surface structural (strucc(1)) was applied as the initial carbon content for the Leaves pool. Similarly, belowground structural (strucc(2)) was applied to the FineRoots pool. The remaining litter pools (snags, small branches, large wood and coarse roots) are set to zero initial value.

• We took the soil organic carbon (SOC) initial values from the CENTURY site.100 input file, and the surface and belowground metabolic data from the first row of the output file of our CENTURY simulation of the Duke site.

MEDFATE pool	CENTURY parameter	DUKE	HARVARD
Leaves	strucc.1.	72	311
FineRoots	strucc.2.	58	223
SurfaceMetabolic	metabc.1.	28	89
BelowgroundMetabolic	metabc.2.	42	477
SurfaceActive	SOM1CI(1,1)	65	105
BelowgroundActive	SOM1CI(2,1)	60	120
SurfaceSlow	SOM2CI(1,1)	850	1400
BelowgroundSlow	SOM2CI(2,1)	1100	1800
BelowgroundPassive	SOM3CI(1)	1200	1200

Decomposition parameters

Litter decomposition parameters

Species	LeafLignin	WoodLignin	FineRootLignin	Nleaf	Nsapwood	Nfineroot	LeafLigninN
DUKE	20.0	25	20	5.000000	1.744186	10	40.000
HARVF	20.3	25	8	8.474576	4.658385	10	23.954

Maximum decomposition rates

The maximum (base) decomposition rates for the different soil pools in MEDFATE (i.e. control parameter baseAnnualRates) were adapted to match the maximum rates from CENTURY (these can be found in the fix.100 file).

Element in baseAnnualRates	CENTURY pool	CENTURY parameter	Value
Leaves	Surface structural	DEC1(1)	3.9
FineRoots	Belowground structural	DEC1(2)	4.9
SurfaceMetabolic	Surface metabolic	DEC2(1)	14.8
BelowgroundMetabolic	Belowground metabolic	DEC2(2)	18.5
SurfaceActive	Surface active	DEC3(1)	6
BelowgroundActive	Soil active	DEC3(2)	7.3
BelowgroundPassive	Soil slow turnover	DEC4	8^{-4}
SurfaceSlow	Surface intermediate	DEC5(1)	0.03
BelowgroundSlow	Soil intermediate	DEC5(2)	0.07

Maximum decomposition rates for dead wood pools (small branches, large wood and coarse roots) were taken from elements DECW1, DECW2 and DECW3 of tree.fix.

Element in baseAnnualRates	CENTURY parameter	DUKE	Harvard
SmallBranches	DECW1	1.5	1.5
LargeWood	DECW2	0.5	0.5
CoarseRoots	DECW3	0.6	0.6

Rate of mixing between surface slow and belowground slow compartments was taken from tree.100:

Element in control	CENTURY parameter	DUKE	Harvard
annualTurnoverRate	TMIX	0.11	0.11

Soil texture and pH

We took the the sand and clay values (and multiplied them by 100 to obtain the fractions) and the pH value from the CENTURY site.100 file.

Soil property	CENTURY parameter	DUKE	Harvard
Percent sand	SAND	28	80
Percent clay	CLAY	12	5
Soil pH	PH	5	5.16

Environmental variation

Variation of environmental conditions for carbon decomposition were drawn from CENTURY outputs. Specifically, soil temperature was taken from CENTURY variable stemp, whereas CENTURY variable asmos.1. was used to estimate soil relative moisture.

Litter input

Litter inputs (senescence of leaves, small branches, large wood, fine roots and coarse roots) are estimated using CENTURY output corresponding to variation in live carbon pools in the forest system (i.e. rleavc, fbrchc, rlwodc, frootc and crootc) and the CENTURY parameters specifying monthly death rates for those live pools (i.e. leaf death rates leafdr(x) for each month and wooddr(2-5) for the remaining four pools, and wooddr(1) in case of a deciduous forest).

Simulations

MEDFATE simulations were run using function decomposition_DAYCENT().

Comparison

The monthly results from the CENTURY simulation are transformed to daily data to make it comparable to the results from MEDFATE. The results are compared in terms of total soil C stock, soil organic matter pools (SOM pools), (leaf and fine root) litter pools and dead wood pools.

Results

Total soil carbon

SOM pools

Structural (leaf and fine root) and metabolic pools

Dead wood pools