Recent developments in JULES-crop

Recent developments in JULES-crop
Karina Williams, Jemma Gornall with thanks to Anna Harper,
David Scoby, Tim Arkebauer, Tristan Quaife, Debbie Hemming,
June 28th, 2016
JULES-crop
I
JULES-crop is a generic parameterisation of crops, which has been
in JULES since version 4.0.
I
Each crop is treated as a separate surface tile.
I
Rate of crop development is determined by temperature and
parametrised by the crop development index (DVI).
sown
emerges
-1
0
c Crown copyright. Met Office
flowers
DVI
1
harvested
2
Crop carbon pools
I
NPP is distributed to leaf, root, stem, harvest and reserve carbon
pools according to DVI.
I
Carbon is moved between carbon pools to simulate the
remobilisation of stem reserves and leaf senescence.
I
LAI is calculated from leaf carbon, canopy height is calculated from
stem carbon
Maize
fraction of NPP partitioned to organs
1.0
0.6
0.4
0.2
0.0
0.0
c Crown copyright. Met Office
root
stem
leaf
harv
0.8
0.5
1.0
DVI
1.5
2.0
Recent improvements to JULES-crop
New parameters
I
improved parametrisation of senescence (mu_io, nu_io,
sen_dvi_io)
I
initial_carbon_io: carbon in crops at emergence
I
t_mort_io: temperature of second soil layer which triggers harvest
(if crop has already flowered)
I
yield_frac_io: fraction of the harvest carbon pool which is added
to the crop yield diagnostic
New diagnostics
I
harvest_trigger: Condition which triggered crop harvest
I
harvest_counter: Can be used to keep track of the number of
crop harvests within a time period
c Crown copyright. Met Office
Tuning using Mead FLUXNET sites
There are three FLUXNET sites at Mead, Nebraska: irrigated maize,
irrigated maize/soybean rotation, non-irrigated maize/soybean rotation.
Sukyer et al 2004, Verma et al 2005 (photo credit: fluxnet.ornl.gov).
c Crown copyright. Met Office
Improved parametrisation of senescence
I
Old parametrisation: a fraction 0.05 of leaf carbon is moved to the
harvest pool per day after DVI = 1.5.
I
New parametrisation: a fraction
ν
min [µ (DVI − DVIsen ) , 1]
of leaf carbon is moved to the harvest pool per day after
DVI = DVIsen .
This reduces to the previous parametrisation when DVIsen =1.5,
µ=0.05, ν=0.0.
c Crown copyright. Met Office
(1)
Senescence: old parametrisation
nu=0.0, sen_dvi=1.5
0.20mu * (dvi - sen_dvi) ** nu, mu=0.05, nu=0.0, sen_dvi=1.5
green leaf carbon
0.15
0.10
0.05
1.5
2.0
0.00
0.0
0.5
1.0
DVI
1.5
2.0
nu=3.0, sen_dvi=0.5
* (dvi - sen_dvi) ** nu, mu=0.02, nu=3.0, sen_dvi=0.5
0.20mu
Coloured lines:
Mead FLUXNET observations for maize.
Black lines: simulated.
c Crown copyright. Met Office
0.15
0.05
1.5
2.0
0.00
0.0
0.5 tuned 1.0new parametrisation
1.5
2.0
Senescence:
DVI
=3.0, sen_dvi=0.5
0.20mu * (dvi - sen_dvi) ** nu, mu=0.02, nu=3.0, sen_dvi=0.5
green leaf carbon
0.15
0.10
0.05
1.5
2.0
0.00
0.0
0.5
1.0
DVI
1.5
Coloured lines: Mead FLUXNET observations for maize.
Black lines: simulated.
c Crown copyright. Met Office
2.0
Initial amount of carbon in the crop
Mead FLUXNET observations for maize
100
Abovegroundc / kg C m-2
10-1
10-2
10-3
10-40.0
0.2
0.4
DVI
0.6
0.8
Initial carbon was previously hard-wired to 0.01 kg C m−2 which for maize is
equivalent to aboveground carbon of approximately 0.005 kg C m−2 .
c Crown copyright. Met Office
Soil temperature which triggers harvest
Green: grid boxes where crop was harvested because temperature in second soil
layer went below t mort=8◦ C in at least one year in the global runs in Osborne
et al 2015.
This plot was created using the new diagnostic harvest_trigger.
c Crown copyright. Met Office
Yield fraction of harvest pool
In JULES-crop, the harvest carbon pool includes
I
the economically valuable part of the plant (maize: kernel)
I
other reproductive parts of the plant (maize: cob, husk, ear shank,
silk)
I
yellow leaves
The yield fraction converts the mass of the carbon in the harvest
pool on the day of harvest into carbon mass of yield.
c Crown copyright. Met Office
Yield fraction of harvest pool
260/2013
kernel fraction of ( yellow leaf mass + repromass )
0.80
y=mx+c, m=0.09, c=0.60
0.78
266/2013
241/2012
0.76
284/2011
277/2011
0.74
272/2011
0.72
268/2013
277/2008
267/2007
270/2007
267/2014
260/2007
0.70
0.68
1.2
1.3
1.4
1.5
1.6
1.7
( yellow leaf mass + repromass ) in kg m-2
1.8
The Mead FLUXNET observations (maize) are labelled with the day of year
and year that the observation was taken (last measurement day before harvest
in each year).
c Crown copyright. Met Office
Canopy structure factor
Mead FLUXNET observations for maize before flowering. Colours show the
cosine of the zenith angle.
Left: FAPAR observations with diffuse radiation fraction between 0.2 and 0.3.
Right: model FAPAR using observed diffuse radiation fraction (0.2-0.3) and
assuming uniform canopy (a=1) using sellers.py (Quaife 2016).
c Crown copyright. Met Office
Canopy structure factor
Mead FLUXNET observations for maize before flowering. Colours show the
cosine of the zenith angle.
Left: FAPAR observations with diffuse radiation fraction between 0.2 and 0.3.
Right: model FAPAR using observed diffuse radiation fraction (0.2-0.3) and
assuming a non-uniform canopy with a=0.65 using sellers.py (Quaife 2016).
c Crown copyright. Met Office
Soil moisture stress
Soil moisture stress factor (fsmc) based on
soil moisture obs and an exponential root
distribution with e-fold depth 0.5m.
Mead FLUXNET obs for maize. Site is silty clay loam (θw = 0.25, θc = 0.39).
c Crown copyright. Met Office
Soil moisture stress
Two new options for soil moisture stress have been added to
JULES vn4.6:
I
Soil moisture stress factor can be calculated based on the
average soil properties in the root zone (as opposed to being
weighted by root mass density)
I
Soil moisture threshold at which plant starts to experience
stress can be pft-dependent
But there is still lots more to do in this area → new Process
Evaluation Group on improving the parametrisation of soil
moisture stress on vegetation. Email Anna Harper or me for more
information.
c Crown copyright. Met Office
AgMIP
JULES is currently participating in two parts of the Agricultural Model
Intercomparison and Improvement Project:
I The AgMIP Maize Evapotranspiration Study
I The AgMIP Global Gridded Crop Model Intercomparison Phase 2
LAI against day of year for obs (points), agmip-maize-ET-v0.2beta (blue) and agmip-maize-ET-v0.3beta (green).
c Crown copyright. Met Office
JULES-crop users meeting
I
1st-2nd August, Met Office
I
Share experiences of using JULES-crop
I
Everyone welcome!
I
Email [email protected] for more information.
c Crown copyright. Met Office