The goal of the Cozero Forecasting Model
The goal of the Cozero forecasting model is to give you an idea of how your company’s CO2 emissions might develop in the future based on different scenarios that the world might take. This allows you to set reasonable targets and assess the gap between your targets and the potential future development. The model provides insights into your company’s future emissions development, departing from the assumption that you would not proactively take any climate actions. In such a “do-nothing-scenario”, your company could still benefit from the economies’ general decarbonization effects i.e. greener energy systems.
To account for your company’s climate actions and GHG reductions, please refer to the —> Cozero strategy builder.
Model set-up
Model set-up
The Cozero forecasting model factors in two main drivers that influence the future development of your corporate carbon footprint (see schematic overview below):
Projected growth of your company: If your company grows, it is likely that without further carbon action, your emissions will increase. Therefore, the first part of our model is a statistical model predicting the growth of your company, and directly translating that into an increase in CO2e emissions. The model is based on general GDP predictions as well as your company’s past revenue if available.
Emission intensity scenarios: It is likely that emission intensities (e.g. the amount of emissions produced by consuming 1 kWh of electricity) will decrease in the future due to the world’s effort to decarbonize. In order to predict the impact on your company, we take external data on how fast emission intensities of different business activities will decrease and bring it together with your company’s log data on Cozero. For example, if most of your emissions are produced by consuming electricity, decarbonization is likely to happen rather fast. In contrast, if most of your emissions are produced by flights, decarbonization will happen much slower. The underlying data sets for this part of the model are from the 2017 IEA Energy Technology Perspectives Report (IEA ETP, 2017).
Schematic overview of the forecasting model
Settings
In the following, we explain how the settings influence the outcome of the model to instruct you on how to apply it to your company’s needs.
Region
The region determines which general growth will be taken as prior assumption for the statistical model predicting the growth of your company. Therefore, it makes most sense that you select the region in which you create most revenue, not necessarily the headquarters of your company.
Scenario
The scenarios describe different potential future developments of emission intensities, depending on how much effort the world takes in decarbonizing. We offer three different scenarios:
Global warming well below 2°C until 2100: This is a very optimistic scenario assuming that the world will achieve very fast decarbonization, likely limiting global warming to well below 2°C. Please refer to the Beyond 2°C Scenario in IEA ETP 2017 for further details.
Global warming of 2°C until 2100: This scenario assumes that the world takes a pathway to a global warming of 2°C by 2100. Please refer to the 2°C Scenario in IEA ETP 2017 for further details.
Global warming of more than 2°C until 2100: This scenario assumes a slower decarbonization, i.e. the goal of limiting global warming to 2°C is likely to not be reached in this scenario. It mainly builds on the carbon reduction commitments made in the Paris agreement. Please refer to the Reference Technology Scenario in IEA ETP 2017 for further details.
Company growth model
Company growth model
Cozero growth forecast
To forecast the growth of your company, we use a statistical model that takes into account the general forecasted GDP growth as well as data about the past development of your company. The general forecasted GDP growth serves as a so called prior, i.e. a default assumption, in the model while your company’s data serves to tailor the model towards the specific development of your company. Therefore, the more data points we have on the past development of your company the less the model relies on the general GDP growth. In contrast, in the extreme case of not knowing anything about the past development of your company the model purely predicts according to the general forecasted GDP growth.
In detail, this is the input data that we use to project the growth of your company:
GDP growth: This is the growth of GDP projected by IEA ETP 2017 for the region you selected as Your main revenue region in the forecasting settings.
Past growth of your company:
If you fill in the section Growth indicator - Past data of the forecasting settings we use the data provided by you. In order to account for the specific needs of your business you can choose between different types of growth indicators (e.g. revenue, fuel consumption). Note that we convert the data to percentages relative to the reference period, i.e. the absolute values are not important, it’s just important that they reflect the percentage growth of your company.
If you don’t provide any data we use the past emissions from Cozero Log as a proxy for the growth of your company up until the selected reference period. This is a weak proxy though, i.e. we’d highly encourage you to provide data for the growth indicator to get a more accurate forecast.
If no data on past emissions is available the model defaults to a forecast based purely on GDP growth.
Custom growth forecast
In order for you to be able to tailor the forecasting model to your needs, we allow you to add your own forecast of your company’s growth in the section Growth indicator - Future data of the forecasting settings. In this case we replace our growth forecast given by the values that you provide up until the last year you provided. If you leave gaps in the data (e.g. you fill in a value for 2025 and 2030 but not in between) we interpolate the values in between linearly. After the last year of the input you provide we fall back to our growth model, adjusting it with the values you provided, and then forecasting the further development until the end of the graph.
Forecast consolidation across the organizational structure
Forecast consolidation across the organizational structure
To illustrate the forecast consolidation logic, think of your organizational structure as a family tree with parents and children.
The logic follows two main rules:
Top-down inheritance: If a lower level business unit (child) does not have its own customized forecast settings, it inherits its forecast from a higher level business unit (parent).
Bottom-up aggregation: A child business unit that has its own customized forecast settings, will not inherit its forecast from its parent business unit, but send its custom forecast data points to its parent, to be included on the consolidated higher level.
The top-down inheritance reduces the set-up time for organizations, since just one set of forecast settings on the top level can be used by the entire organization. The forecast model generates data per business unit, so a child unit will take only the portion of their parent’s forecast relevant to them. Thereby, all business units can immediately use Act with a forecast particular to their business unit.
The bottom-up aggregation of forecast data enables organizations to incorporate differing growth assumptions for different divisions of their business, while still having all this data reflected in the overarching organization’s forecast.
Important note: For the inheritance and aggregation logic to work, the selected reference period years of the different business units must be the same.
Consider the graphic and the following examples:
1. When a forecast is created for business unit A
All business units use the forecast settings of business unit A, until they create their own
2. When a forecast is created for business unit B
Business units B, B1, and B2 use the forecast settings of business unit B
Business unit A’s forecast gets updated with new forecast data points from business unit B
3. When a forecast is created for business units C1 and C2
Business unit C and A’s forecasts get updated with the new data points from business units C1 and C2
Business units B and its children remain unchanged because business unit B has its own forecast settings