Mimi Framework

`acei` and `aeei` parameters in FUND

Hi there!

Just wanted to clarify what the parameters acei and aeei were in FUND, and how they are calculated? I am guessing that acei is growth rate of (carbon) emissions intensity, and aeei is growth rate of energy intensity?

However, if so, it is my understanding that carbon emissions intensity is usually calculated as (CO2 emissions / GDP), whereas the following code (from scenconverter.jl in FUND) suggests that carbon intensity is calculated based off energy intensity, which is in turn calculated as (energy use / income):

v.energint[t, r] = p.energuse[t, r] / p.income[t,r]

            v.emissint[t, r] = p.emission[t, r] / p.energuse[t, r]

            if !is_first(t)

                v.scenaeei[t, r] = -(v.energint[t, r] / v.energint[t - 1, r] - 1.) * 100.

                v.scenacei[t, r] = -(v.emissint[t, r] / v.emissint[t - 1, r] - 1.) * 100.

I see in the FUND model the scenconverter.jl script isn’t actually included, and rather it reads in the parameters directly from a data file, so I just wanted to double check whether the parameters read into the model from the csv data files are calculated using the above definition or another?

For context, I am asking this question because I’m trying to read in a different CO2 emissions vector, and want to make sure that aeei and acei are consistent with this exogenous vector.


@davidanthoff ideas here?

acei is Autonomous Carbon Efficiency Improvement and aeei is Autonomous Energy Efficiency Improvement. We use the Kaya identity decomposition, and in that framework carbon/emission intensity is defined as per unit of energy, not per unit of GDP. DICE uses a different decomposition, i.e. it skips the whole eneryg part and has an emission intensity that goes directly from GDP to emissions. I think, though, that most models in the scenario world actually use the Kaya framework.

And yes, the scenconverter component can be used if one has input data that is the levels in the Kaya identity, and then it converts things on the fly into the rates of change form that FUND requires as an input.

1 Like

Thanks, David! To use the scenconverter component, should I just add that as a component at the beginning when I’m building the model and reconnect the variables outputted from that component to the relevant input parameters in the scenariouncertainty / population / socioeconomic components? Or is there a better way of doing it?

@tammyt123 I think what you’re describing sounds correct, recalling that you’ll need to place it in the right spot in the list of components, do you agree @davidanthoff ?

@tammyt123 we can stay in touch if/when you do that, I haven’t used it before and am not sure it’s tested, so my only concern would be that you’ll run into a problem in the beginning with missing data or something. But perhaps it’ll work smoothly!

1 Like