Network Analysis PPR values in .csv file not clear

[sheilah]

I have been looking at the PPR estimated by Ecosim and Ecopath (for both EwE 5 and EwE6) and it is not clear to me what is being reproduced in the network.csv file. If you look at the value reproduced in the network.csv file for PPR it does not reproduce any of the values for PPR in the Ecopath model for the first year. Can you please check that?

[jeroens]

It is confirmed that the value produced in Network Analysis

Ecosim network analysis > Indices with PPR > CSV, Catch PPR

differs from

PPR > For harvest of all groups > PPR (total)

It is presumed that these two values should be identical.

[joeb]

The values for PPR estimated by Ecosim in the network.csv file can be different than the PPR values estimated by Ecopath. Ecopath outputs are computed from the Ecopath inputs, which is effectively time step zero. The Ecosim outputs start at the end of the first time step. So the first value is at time step one and may differ from the Ecopath PPR calculated from time step zero.

Having said that the way the Network Analysis computes EE for input to PPR seems wrong. See cEcoNetwork.PrepareUlanowForCallFromEcosim() aprox. line 3949. 

SimEE(i) = 1 - (m_esdata.loss(i) - m_esdata.Eatenof(i) - m_esdata.FishTime(i)) / (SimPB(i) * BB(i))

loss(ngroups) is sum of all biomass loss in the system computed by Ecosim, by group.

Eatenof(ngroups) is the biomass lost to predation on this group.

FishTime(ngroups) is F rate (fishing mortality per unit of biomass) not catch.

loss() and Eatenof() are both total biomass FishTime() is the rate of loss due to fishing not the catch.

If it is supposed to be catch it should be FishTime() * Biomass()

[jeroens]

First for .. loop in cEcoNetwork.PrepareUlanowForCallFromEcosim seems wrong. Villy, please validate:

For i = 1 To m_epdata.NumLiving

SimB(i) = BB(i)
SimPB(i) = m_esdata.loss(i) / BB(i)
SimCatch(i) = BB(i) * m_esdata.FishTime(i)
SimEE(i) = 1 - (m_esdata.loss(i) - m_esdata.Eatenof(i) - m_esdata.FishTime(i)) / (SimPB(i) * BB(i))

It is presumed this should be

SimEE(i) = 1 - (m_esdata.loss(i) - m_esdata.Eatenof(i) - SimCatch?(i)) / m_esdata.loss(i)

[joeb]

The above change, using catch instead of F in the calculation of SimEE has been committed for version 6.2 but I have left the ticket open as a reminder

[jeroens]

Moved to next release; we need Villy and he's not here, but there is no point stalling the release for this issue