Climate models

We need to improve climate models to make better policies

Representative photo: Chris Liverani/Unsplash

  • As COP27 approaches, the spotlight is once again turning to crafting a compelling national narrative on climate change – based on what India can do and what India needs.
  • There are many independent estimates of India’s needs, ranging from $1 trillion to invest over the next decade to $10 trillion needed to reach net zero by 2070.
  • How are targets and estimates like these determined and what do they mean for the country?
  • Quantified goals can usually be attributed to scenario studies generated by powerful analytical tools called energy saving models.
  • They follow a seemingly elegant process – but if you look closely you may notice questions about the purpose of their results and the extent to which they can inform policy.

As COP27 approaches, the spotlight is once again on building a compelling national narrative on climate change – based on what India can credibly do and what India actually has. need. Five hundred gigawatts of renewable energy capacity to be installed by 2030; 50% of installed capacity will come from non-fossil sources by 2030; net zero emissions to be achieved by 2070 – many national goals set the stage for such a narrative.

Similarly, there are numerous independent estimates of India’s needs, ranging from $1 trillion to be invested over the next decade to $10 trillion needed to reach net zero by 2070. objectives and estimates of this type determined and what do they mean? for the country?

Quantified goals can usually be attributed to scenario studies generated by powerful analytical tools called energy saving models. Although these studies can take different forms, the essential premise is the same: a modeller makes certain assumptions about the climate economy, technology and policies, and incorporates these assumptions into the model in the form of scenarios. The model then performs calculations and generates results for each scenario; these typically relate to a country’s projected emissions, the costs of those policies, and combinations of energy demand and supply.

It sounds like a very elegant process – but how conclusive are the results, and what else do we need to know about these models to design policies?

Unclear and deterministic processes: There is a wide variety of models that have been developed for use in climate policy making. Each modeler may use a certain model, based on varying sets of assumptions about the country’s future socio-economic opportunities and needs – for example, whether India is likely to grow at an annualized rate of 5.5% or 6%, or whether the cost of solar energy will drop by 45% or 50%, over the next decade – and then each model can handle these assumptions differently depending on its structure.

Because of these varied processes, different models can say very different things about how India’s climate and development goals can be achieved. While this diversity of assumptions and results can help capture a wider range of uncertainties in India’s future paths, a challenge lies in the credibility of the assumptions, the clarity with which the approaches are communicated, and the circumspection studies about these uncertainties. Often uncertainties may not be highlighted and the results of modeling studies may be treated deterministically as a fait accompli.

Insufficient focus on challenges and implications: In addition to this, most models focus on technological solutions to mitigate emissions – offering recommendations on renewable energy, hydrogen, carbon capture, demand side management, etc. They are ill-equipped to comment on the political or behavioral aspects of implementing these solutions – the distributional impacts of phasing out coal on jobs, for example, or the management of just transitions, the management of resources, or the exploration of alternative trends of urbanization or economic growth is possible.

Through this narrow scope, the adoption of modeled outcomes may risk locking India into a techno-economic path and turning scenarios into self-fulfilling prophecies, where outcomes that should be economically feasible – regardless of socio-political constraints – are adopted in the national objective. -setting.

These confounders pose a problem for policy makers, who look to models for realistic outcomes and are faced with conflicting basic messages about what might be necessary and what might be possible – under the very specific conditions under which these models work – and proceed to design policies based on limited information in the face of an uncertain future. Moreover, with this approach, it is very difficult to envision what the implications of these findings might be for India’s future development. It is therefore not surprising that this approach entails the risk of policies whose implementation has only limited success.

We need a deeper, more considered approach to modeling. Policy makers need to be able to better understand what these modeling studies are qualified to say, whether their statements are defensible, and the conditions under which they release their statements and what information they are unable to capture.

There is a potential two-step approach to achieving this. First, modeling studies must be carefully evaluated for their structure (how well they are put together) and second for their implications (how lessons are interpreted). The first step is to ask whether the choice of model is appropriate, whether the inputs are credible, how robust the scenarios are constructed, how the study accounts for uncertainties, and whether the results of the study are validated.

In the second stage, the studies must be carefully assessed for what they say – or imply – about the socio-economic development models that are locked in, how the energy transition will be managed, what emissions are projected, what are the needs for investment, how the study thinks about social equity and natural resource impacts, and what this will mean for India’s energy security.

A better understanding of these two things – the structure and implications of modeling studies – can help decision-makers have more information about how to use these modeling studies in a nuanced way and what additional information to bring when they design policies. This can lead to better policy design and create the necessary policy space to deal with the many uncertainties present in India’s development future when implementing policies.

Such an approach can also serve as a checklist of factors to consider when designing future modeling studies and can thus encourage greater transparency and rigor in the way studies are set up, so that they be clearer about what they are saying, as well as recognizing what they are unable to say.

The famous British statistician George Box said: “All models are wrong, but some are useful. Given the importance of well-designed climate policies and the central role of models in their design, it is essential to better construct and interpret future modeling studies, in order to improve their usefulness and application.

The authors are grateful to Navroz Dubash and Sonali Verma for their contributions to this article, and to Eri Ikeda and Mandakini Chandra for their continued contributions to this project.

Aman Srivastava is a fellow at the Center for Policy Research in New Delhi. Kaveri Iychettira is an Assistant Professor at IIT Delhi.