introduction to biogeochemical modelling (elements of...
TRANSCRIPT
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
Introduction to biogeochemical modelling(elements of system theory and language ..) (C. Solidoro )
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
A model is a representation of reality
(it captures essential features/particular aspects of reality)
tools for:
PREDICTION
UNDERSTANDING
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
two words about MODELS……Knowledge, in every field, is the result of a number of steps:
gathering of empirical information, also through collection direct observation
empirical understanding of phenomena (i.e. identification of relationships among vairables).
theoretical understanding of relationships amongvairables, (hypothesis of cause- effect mechanisms, identification of conceptual models, quantitativeunderstanding of phenomena (set up of a THEORY)
collection new data for corroboration /falsificationmodel
application of the model to new problems
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
different purposes, different representations ..
different models ..
statistic modelsdeterministic modelsstochastic models
toy modelsprocesses oriented models
numerical model
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
REALSIM
ACCURACY
Accurately describe specific situationsMay perform worse in other situationsTypically detailed
Account for known relationshipsMay be complex
GENERALITYApplicable in most situationsDon’t predict specific outcomes wellTypically “simple”Good for exploring relationships
(functionality)
(performance)
toy
proc orien
numeric
applied models
theoreticalmodels
functionalmodels
Modified from Levins (1968)
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
BGC models usually are:
deterministic
process oriented
(simple),
spatially resolved (coupled to transport models)
GENERALITY ACCURACY
REALSIM
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
DETERMINISTIC MODELS
Basic idea:1. a system (the piece of reality we are interested
in) can be efficiently described by a finite number of variables (state variables),
2. both the system and its evolution in time can be explained as a function of interactions (input, boundary conditions, controls) among the system and the remaining part of the universe
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
fitofito
Timescale
virus
zoofish
zoo
Space scale
small scales implie that in most case studies we have spatial heterogeinity and need of transport model
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
1m-1 years
50m – 10 km
Up to mesoscale
Dickey 2001
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
Coupling hydrodynamical and biological modules
[ ]>′′<−Θ∇∇+Θ∇⋅−=∂Θ∂ θukUt ii
igeneric passive tracer
[ ] ( )..,,, IITTqukUt ii
i ′+′+′+Θ+>′′<−Θ∇∇+Θ∇⋅−=∂Θ∂ θθ
generic active tracer
[ ] ( )..,,, ITqukUt ii
i Θ+>′′<−Θ∇∇+Θ∇⋅−=∂Θ∂ θ
proper time scale (x-y vs z; biology “slow”)
( )..,,,2 ITqz
wz
Kz
kUt
isi
iviHhi
i Θ+∂Θ∂
+⎥⎦
⎤⎢⎣
⎡∂Θ∂
∂∂
+Θ∇+Θ∇⋅−=∂Θ∂
standard formulationCrise et al. 2001
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
Biogeochemical ModelSet of Advection-Diffusion-Reaction equations:Conservation laws with source and sink terms
biophys tc
tc
tc
∂∂
+∂∂
=∂∂
)()()( cRcDcAtc
biophysphys ++=∂∂
Lineartransport term
Non linearreaction term
physic biology
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
0D application
(Rinaldi, Gatto, ….)
( )..,,, ITqt
i Θ=∂Θ∂
‘toys model’ ma anche applicazioni
( )..,,, ITqz
wz
Kzt
isi
iv
i Θ+∂Θ∂
+⎥⎦
⎤⎢⎣
⎡∂Θ∂
∂∂
=∂Θ∂orrizont. omogeno 1D
(many ! DCM jamart,77, varela 92,94)
‘process oriented model’
( )..,,,2 ITqkUt iHhi
i Θ+Θ∇+Θ∇⋅−=∂Θ∂
vert. omogeno 2D
(not very frequent)
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
Venice Lagoon
Adriatic Sea
Lido inlet
Malamocco inlet
Chioggia inlet
Can. d. Navi
Can. Giudecca
Is. S.ErasmoSIL
DES
LUS
OSE
TAG
NAV
MON
CUO
Venice
Can. S.Nicolò
Can. Treporti
ARIARadiazione
luminosaScambiotermico
Temperaturaacqua
Scambioattraverso lebocche
MAREADRIATICO
Det. N
Det. PDet. C
Zoo Fito Oxy
Inputdallagrondalagunare
BACINOSCOLANTE
diporto
RP
N-NOx-
N-NH4+
2N
P NSedimento
C
DeposizioniatmosfericheN, P
)()()( cRcDcAtc
biophysphys ++=∂∂
Solidoro et al 1995, 2004, 2005
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
Numerical Integration
physic submodel
COUPLING
forcing and parameters
biological submodel
new state
Operator splittingtechnique: forcings and
parameters are input foreach of the two
subsystems for the derivative computation in the proper time step. The
derivatives are coupledafterwards, allowing
different time steps forthe two processes and
therefore optimising the computational resources
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
( )..,,, ITq ΘThe BIOLOGICAL term
AmmoniaNitrate
Phosphate
Detritus
Netplankton106:16:1
Picoplankton106:16:1
Zooplankton48:12:1
VariableC:N:P ratio
PP
NN
DD
[N,P,D]
BGC meansN-P-C cycles ..
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
.. The microbial loop ..
phyto
zoo
nut
det
doc
bac
µzoo
don
azam, fenchel, thingstand, rassoulzadegan
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
Sinking vs recycling Fasham 1990
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
The mistivourous food web ( a continuum ..)
Legendere & Rassoulzadegan 1995
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
It is relatively straightforward toformulate more complex models toinclude explicitly differentfunctional groups of phytoplankton, zooplankton and bacteria, and toinclude regulation by multiple nutrients such anitrate, ammonium, silica, and iron.
Denman 2003
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
Proto-zooplankton Meso-zooplankton
Macro-zooplankton
DON(N
d)
DON(N
d) Phytoplankton
(P)
Phytoplankton(P) Nitrate
(No)
Nitrate(N
o)
Amonium(N
r)
Amonium(N
r)
Bacteria(B)
Bacteria(B) Zooplankton
(Z)
Zooplankton(Z) Detritus
(D)
Detritus(D)
Net-PhytoplanktonP2
Net-PhytoplanktonP2 Meso-Zooplankton
Z1
Meso-ZooplanktonZ1 Amonium
(Nr)
Amonium(N
r)
Nitrate(N
o)
Nitrate(N
o) Detritus
(D)
Detritus(D) DON
(Nd)
DON(N
d)
Pico-PhytoplanktonP1
Pico-PhytoplanktonP1 Protozoos
Z2
ProtozoosZ2 Bacteria
(B)
Bacteria(B)
Disolved inorganicnitrogen
Disolved inorganic
phosphorous
Phytoplankton
N P
Dyazotrophs
N P
Zooplankton
N P
Detritalnitrogen
Detritalphosphorous
Sinking Sinking
Dinitrogen
Fennel
FUNCTIONAL TYPES
DYNAMIC GREEN MODEL
GENERALITY ACCURACY
REALSIM
Pico-heterotrophs
Pico-autotrophs
Phyt
opla
nkto
nN
2-fix
ers
Phyt
opla
nkto
n ca
lcifi
ers
Phyt
opla
nkto
n D
MS-
prod
ucer
s
Mixed-phytoplanktonPhytoplankton silicifiers
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
Diatoms
Flagellates
Picophytoplankton
L PhotoadaptationLP(1)
Large Phyto.LP(4)
LP(2)
LP(3)
Oxygen
Carbon dioxide
O Dissolved GasesO(2)
O(3)
PhosphateNitrateAmmonium
Silicate
N Inorganic NutrientsN(1)
N(3)
N(5)
Red. EquivalentsN(6)
N(4)
ZMicrozooplankton
(s.s.)Heterotrophicnanoflagellates
MicrozooplanktonZi
(5)
Zi(6)
Carnivorous
Omnivorous
Z Mesozooplankton
Zi(3)
Zi(4)
B
Bacteria (aerobic
and anaerobic)
Bacterioplankton
Bi
Dissolved
Particulate (detritus)
R Organic Matter
Ri(1)
Ri(6)
Diatoms
Flagellates
Picophytoplankton
P PhytoplanktonPi
(1)
Large Phyto.Pi(4)
Pi(2)
Pi(3)
Vectors (Functional Group or Ordinary State Variables)
Organic matter flow (C,N,P,Si)
Inorganic nutrient flow (N,P,Si)
Gas exchange
Benthic-Pelagflow
Scalars (Ordinary State Variables)
Z
BFM
(ersem)
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
Street & phelps 27 Riley, 46 Steele, 60…..…..La Querelle 06
( … but problems still exist )
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
It is relatively straightforward to formulate more complex models …..
Why many BGC models like to be simple (rough?)
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
However, the number of parameters thatmust be specified from observationsincreases approximately as the square of the number of compartments and quicklysurpasses our ability to constrainthem properly from observations. Moreover, ecosystem models often becomeunstable for small changes in parametervalues, and increasing complexity may notlead to increased stability.
Denman 2003
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
the number of parameters will usually begreater than nember of fluxes
FASHAM 28 parPFT 100-200 ??
Probablity of Instability, abrupt changes, chaosincrease …. EVEN for FORCED model
May 73, Smale 76, Kuztnesov 93
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
…calibration is not the answer
Not all paraemters can be identified by calibration(fitting vs experiemtnal observation)
problems are:over and/or underdetermination (# eqs vx # unknown)no exact solutionexistence of multiple optimal solutions
(there are more than one combianation of parameterswhich give the same fit. This is WITHIN equations and cannot be solved)
hnnVm +
=NfSolidoro 2003Matear 2001Vallino 2000
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
How complex a model should be still is anopen question.
Costanza & Sklar (1985)
The main recommendation is that the use of a single ultimate’ecosystem model is ill-advised, while the omparative and confirmatory use of multiple ‘minimum-realistic’ models isstrongly recommended. (Fulton 2003 ?)
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
In addition we want to be 3D …
… Deades long integration of transport + 50 variables for grid point over 100 meter gridcovering thousands of square kilometers, …
GENERALITY ACCURACY
REALSIM
?
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
Attempts at simplify (paraemterization) exist ..
Steele 98
implicit treatment of the microbialloop. the proportion oftotal grazing thatflows directly tomesozooplankton, can vary accordingto nitrateavailability or total phytoplankton
.. but also Armstrong, Denman …
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
kintetics (choice functinal forms)
structure identification (box and arrows)
First parameter choice
Sensitivit analysis
calibration
corroboration (validation)
applications
After structure identification one has tochoose kinetics and paraemters
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
Standard complexitypelagicmodel
Cossarini e Solidoro 2006
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
[ ]PhyPhyPhyPhy sgrazingmortrespgrowth
dtPhyd ink−−−−=
[ ]ZooZooZoo excretmortgrazingKeff
dtZood
−−⋅= equations in the water quality submodel(mass balances)
[ ] [ ] { } nitrificdecaydecayexcretresprNNHgrowthr
dtNHd
SedNDetNZooPhynctot
Phync −+++⋅+⋅⋅−=++44
[ ] [ ] denitrifnitrificNNOgrowthr
dtNOd
tot
phync −+⋅⋅−=−−33
[ ] { } SedPDetPZooPhypcPhypc decaydecayexcretresprgrowthrdt
POd+++⋅+⋅−=
−34
[ ] ( ) DetCDetNPhytoZooZoo sinkdecaymortmortgrazingKeffdt
DetCd−−++⋅−= 1
[ ] ( ){ } DetNDetCZooPhyZoonc sinkdecaymortmortgrazingKeffrdt
DetNd−−++⋅−⋅= 1
[ ] ( ){ } DetPDetPZooPhyZoopc inksdecaymortmortgrazingKeffrdt
DetPd−−++⋅−⋅= 1
[ ]SedChyPDetC decaysinksink
dtSedCd
−+=
[ ]SedNhyPncDetN decaysinkrsink
dtSedNd
−⋅+=
[ ]SedPhyPpcDetP decaysinkrsink
dtSedPd
−⋅+=
[ ] [ ] [ ]( )OxyDOKrearnitrificardecaydecayrespgrowthrdtOxyd
satnoSedCDetCPhytoPhyoc −⋅+⋅−−−−⋅=
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
{ } [ ]PhyFFFFGPgrowth PNTIPhy ⋅⋅⋅⋅⋅= max
⎟⎟⎠
⎞⎜⎜⎝
⎛−
= opt
k
II
opt
kI e
IIF
1
( )( )
)()(
max
maxmax
opt
opt
TTmTT
opt
mT e
TTTTF −⋅
−⋅
⋅⎥⎥⎦
⎤
⎢⎢⎣
⎡
−−
= αα [ ]totNtot NKN +=NF
[ ] [ ][ ]−− += 34
34 POKPOF PP
[ ] [ ]−+ += 34tot NHN NO
[ ] ⎟⎟⎠
⎞⎜⎜⎝
⎛∫ ⋅+⋅−
⋅=k
Phy dzPhyKselfkKest
sunk eII 0
( )maxm ,TminT T=
[ ] [ ][ ]OxyKOxyF oxyoxy +=
[ ];10det DetXFQKdecdecay oxyDetX ⋅⋅⋅=
[ ]( ) 103 119.0 QKdeniNOdenitrif ⋅−⋅= −
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
Mostly empirical, with a posteriori ‘phyisological’ derivation
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
GPmax 0.18 [h-1] max growth rate for phyto (optimal values of T , I and nutrients)KmPho 0.006 [h-1]death rate for phytoplanktonKrPhy 0.003 [h-1]respiration rate for phytoplanktonKN 0.05 [mg N/L] halfsaturation constant for nitrogen assimilationKP 0.01 [mg P/L] halfsaturation constant for phosphorus assimilationTopt 27 [° C] optimal temperature for phytoplankton growthTmaxx 41 [°C] inhibition temperature for phytoplankton growth ? 0.15 [°C-1] Lassiter e Kearnes exponential coefficient rnc 0.15 mg N/mg C N/C ratio in phytoplankton (Redfield ratio)rpc 0.023 mg P/mg ] N/C ratio in phytoplankton (Redfield ratio)Light parametersIopt 50000 [lux]optimal light for phytoplankton growthKselfPhy 4. [mg C-Phy/L] self-shading coefficient for phytoplanktonKest 1.0 [m-1] shading coefficient Parameters of zooplankton dynamicKgr 0.04 [h-1] max grazing rate Kgphyto 1. [mg C-Fito/L] halfsaturation constant for grazing formulationKmZoo 0.006 [h-1]death rate for zooplankton KeffZoo 0.5 [dimensionless] grazing efficiency KeZoo 0.002 [h-1]excretion rate for zooplankton Parameters of nitrogen dynamicKnit 0.0043 [h-1] nitrification rate at 20°CKdenii 1.6 [mg NO3-/l/h] denitrification rateParameters of sediment and detritus dynamicsKdecDet 0.0048 [h-1] decay rate of organic detritus at 20° C
…….
28 parametri
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
1. choose subset parameters Sensitivity analysis ,Tuning importanceRanking of parameters
what’s the purpose ?What about data ?What about error in data?
2. Define cost function
IF you can .(data assimilation,Adjoint, simplex brute force..)
3. Get parameter values by fitting
There is not sucha thing as a model, but only a model for a purpose
Tuning is possible for subsets of parameters
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
..validation …
..capability to fit new data
Need to define a skill function
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
authors Dim Area Temporal Scale
Spatial Scale
Physical submodel
Biological submodel Coupling Objective
Prod I
Prod. II
Nutr Det/ Doc
Bact DO T L Nutr Turb
Morel & Andre 91 0D West Med
Year Basin - 1 - - - - - - L - - Pr. Prod. From CZCS
Antoine et al. 95 0D East Med
Year Basin - 1 - - - - - - L - - Pr. Prod. From CZCS
Bethoux 91 box Med Sea Asymptotic Basin box - - N,P - - - - - - - Nutrient Balances Sarmiento et al 88 Box Med Sea Asymptotic Basin Box - - PO4 - - DO - - - - Nutrient Balances &
Anoxia Klein & Coste 84 1D NW Med Days 50m Mellor
Yamada 2 - - Nut - - - - - Nut - Influence Wind
Nut. as passive tracer Varela et al 92 1D NW Med Days 200 m Mellor
Yamada 2 2 1 Nrid
Nox - - - T L Nut - Influence Focings on
DCM Zakardijan &
Prieur 94 1D West Med months 200m - 1 1 Nrid
Nox - - DO - L Nut Turr Influence Turbulence
Allen et al. 98 1D Adriatic year 200 m Mellor Yamada 2
E R S E M T L Nut - Ecosystem Study
Tusseau et al 97 1D Ligurian Sea Year 200 m Termal
2 (physiological)
- Nrid Nox Si
DetSi Dom1 (mono) Dom2 (poly)
Bac - T L Nut - Calibrated. Used as Bound Condition for
3D Simulations. Levi et al. 98 1D Ligurian Sea Year 100m Box 1 2 Nrid
Nox
Doms Doml Dets Detl
Bac T L Nut - Fluxes Eval.
Solidoro et al. 98 1D Tirrenian Sea year 100 m diffusive 2 1 Nrid Nox PO4
DetN DetP DetC
- DO T L Nut - Sensitivity Analisys
Napolitano et al. 97 1D Ionian, Rhodus Gyre
year 200 m Mellor Yamada 2.5
1 1 Nrid Nox
Det - - T L Nut - Ecosystem Study
Pinazo et al 96 3D Gulf of Lions 20 days 250 Km Full 2.5
1 - Nrid Nox Q
DetN DetC - DO T L Nut - Influence Forcing & Nut. Cycle
Civitarese et al. 96 3D Ionian Sea Year 200 Km MOM 1 - N D - - T L Nut - Nutrient Balance Tusseau et al. 98 3D Gulf of Lions Year 200 Km General
Circulat. 2
(physiological)
- Nrid Nox Si
DetSi Dom1 (mono) Dom2 (poly)
Bac - T L Nut - Fluxes Eval.
Bergamasco et al. 98
3D North Adriatic months 100 Km POM (nesting off
line)
1 - N - - DO T L Nut - Fluxes, Nesting , Influence Biology
Zavatarelli et al. 98 3D Adriatic year 800 Km POM (MY 2.5)
ERSEM T L Nut - Ecosystem Study
Crise et al 98; Crispi et al. 99
3D Mediterranean years basin MOM 1 - N DetN - - T L Nut - Ecosystem Study
Crispi et al. 98 3D Mediterranean years basin MOM 2 1 Nrid Nox PO4
DetN DetP DetC
- DO T L Nut - Ecosystem Study
BGC have shownto be reliableenough tobe increasinglyused in application …
Arhonditsis & Brett 2003
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
APP MED SEA
Crise et al 2001
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
Understanding system functioning
Crise et al 2001
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
APP VENEZIA
2041.63612 145
35109
6
DIN9.5
Phyto Zoo
Losses1099 Recycling
102 5.024833
36122
43
69
DIN Phyto Zoo
Losses
69 3.4 0.2769427 17
234 7
SouthernSubbasinVol. 0.15 kmSurf. 90 km
3
2
DINPhyto Zoo
Losses
Recycling
1
491
1265
982
1324
601
682
1431
1034
21
1017
471
828
317
63580
1
388
1029
47
CentralSubbasinVol. 0.21 kmSurf. 110 km
3
2
NorthernSubbasinVol. 0.32 kmSurf. 186 km
3
2
Adr
iatic
Sea
Dr a
inag
e B
asin
0.4
Recycling
Solidoro et al 2004
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
-0.7384.73
0.842
0.340.34
2.52
0.2194.07
0.944
1.260.05
2.05
4.97
0.058
2.02
9.595
1.685
-3.96
-1.097
-0.044
-0.119
Fitoplancton0.024 mg N/l
NO3-
0.46 mg N/l
NH4+
0.06 mg N/l
Zooplancton0.0023 mgN/l
Detrito N0.013mgN/l
Sedimento N0.041mg N/m2
0.3611
+ AQUACULTURE+ control Nutrient load+ DA
Solidoro et al 2001, Pastres et al. 2001, Cossarini et al. 2004
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
base for end-to-end conceptualization marine ecosystem
virus
whale
fitozoo
fish
Space scale
Timescale
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
…another chapter of the story …
ECEM’07 WorkshopBiogeochemical processes and fish dynamics in food web models forend-to-end conceptualisation of marine ecosystems.Theory and use of Ecopath with Ecosim.
Evolution of Mediterranean Ecosystem Models
Green = Box,Conceptual,reviews 6
Yellow = 1D BGC N 12
Grey = 1D BGC multinutrient 4
Magenta = 1D BGC ERSEM 7
Turquoise = 3D BGC 19 (6 ERSEM)
White = Data assimilation 9