Uncertainty in baseline CO2 emissions for South Africa

Client: United Nations Environment Programme

Period: September 2013–December 2014

Project leader: Bruno Merven

Project team: Bruno Merven, Bryce McCall, Ian Durbach (UCT Statistics Department)

The objective of this project is to quantify the uncertainty associated with key model inputs to develop a probability distribution of baseline emissions for South Africa over the 2015-2050 period. This objective is to be met in two phases. In the first phase, the most important and uncertain input parameters were selected for uncertainty analysis, and the associated uncertainty was described. In the second phase the uncertainty in inputs is propagated via an E3 model of South Africa (South African TIMES Model - SATIM) to obtain the probability distribution for the baseline emissions of South Africa, over the period of interest.

Projecting this far into the future is an extremely, perhaps impossibly, complex task. We use a combination of methodological approaches to do this, triangulating between these approach in an attempt to arrive at some kind of consensus projections. The approach followed here is to assess uncertainty on a small number of key drivers influencing the energy system, and hence GHG emissions associated with it. We assess distributions over possible values that these drivers can obtain in the future, and pass these values to the E3 model. For each combination of possible inputs, the model returns outputs for quantities like GHG emissions. By submitting many possible inputs to SATIM, a range of possible outputs is obtained. This process takes the form of a Monte Carlo simulation.

Outputs: Report due April, long abstract accepted at IEW 2015, 2 publishable papers expected in 2015.