News From The IPCC - Data Science And A Changing Climate

It’s Earth Month and the Inter-governmental Panel on Climate Change (IPCC) just released the third installment in the series of reports on the state of the planet. The first report dealt with the science of climate change and the second one looked at its impacts on society and the planet. The third report was released a few days ago and looked at our response to climate change as societies and what we can expect as a result.

IPCC reports have been released approximately every decade since 1990 - the current one is the 6th installment. Several hundred scientists collaborate on the technical and scientific assessments - reviewing the latest research published globally, combining multiple scientific areas and disciplines - in order to develop as comprehensive a picture as possible of the state of the planet. Just as an example, the second report on impacts to the planet draws from 34,000 studies and involved 270 authors from 67 countries. 

The first report, released in December 2021, evaluated the latest published research on climate science. The areas that were covered were - how do we measure the Earth’s climate, how can we measure the impact of human behavior since the industrial age on different Earth systems, what are the short-term and long-term drivers of climate change, what is the state of major components of the Earth system (primarily the water cycle, carbon cycle, biogeochemical cycle, oceans and cryosphere), can we link global models to regional models and what is happening with extreme events around the world.

The second report, released in February 2022, explored the impacts on human societies and ecosystems as a result of a changing planet. The main areas covered were - measuring the extent of the impacts on the planet, identifying how the planet will continue to change as the climate changes, what adaptation and mitigation strategies are possible and how do they play out in different parts of the globe, and how can the risks associated with climate change be managed effectively.

The third report, released in March 2022, examined societal responses to climate change to date, their effectiveness and what’s needed in the future to ensure that the planet remains hospitable to humanity and our activities. The primary questions asked were - what commitments were agreed on and how effective are they in mitigating climate change, what are the recent advances in knowledge across sectors and systems including energy, urban and other settlements, transport, buildings, industry, and agriculture, forestry and other land use, how can these advances and technological developments be used to effectively mitigate or adapt to climate change, and what are the social, governmental and behavioral changes that are most likely to keep climate change at manageable levels. 

From a modeling perspective - as we’ve discussed in our previous set of blog posts - all these questions can be divided into three main categories. 
The first category is data collection and data assimilation - which is where all measurement and linkage questions can be grouped. When trying to understand planetary processes, scientists have looked at past data (paleoclimate and biotic records), current records (including direct measurements) and predicted possible outcomes in the future based on the best understanding of how planetary processes behave. In most scientific studies looking at the planet we have data at different spatial and temporal scales from a variety of sources - ice cores, tree rings, direct atmospheric measurements of different gases, measured ocean temperatures, historical records, model simulations of past, current and future conditions. Typically these data are correlated to determine the patterns, if any, exist. The main questions that get answered here are 1) Are there any correlations between parameters? 2) How certain are these correlations given the differences in spatial and temporal scales? 3) How significant are the changes over the last few decades - i.e. are we seeing a greater signal compared to background noise?

The second category deals with understanding the processes behind the changes that are observed. This is where several combined modeling techniques get used - physics-based or process-based models, standard spatial and temporal statistical models, and machine learning models. Of course, statistical and machine learning techniques are used in the first category as well - a classic example being satellite measurements of Arctic sea ice - but they are used and combined to a greater degree while exploring and understanding the drivers and links between planetary systems. Here, scientists focus on building and understanding the processes that drive different planetary systems and how sensitive they are. For example, a biogeochemical model looks at the flow of nitrogen through the planet - the processes include nitrogen application to land, nitrogen movement in plants, soil and water systems, nitrogen flow in different water and atmospheric bodies. The models include the different physical, biological and chemical processes occurring as well as the interactions between all of them. The results are an estimate of how much nitrogen there is in different ecosystems, their impacts on wildlife and what is likely to happen as conditions change. 

The questions that typically get asked here are - how accurate is our current understanding of planetary systems, where are the gaps in knowledge, how has our understanding changed over time, what are the planetary systems that appear to be most sensitive to climate change and which processes are driving that sensitivity? As an example, extreme events such as the floods, fires, storms and droughts that occurred in 2021 are evaluated, an understanding of how likely it is that these events will continue and perhaps worsen is developed, and strategies and technologies to minimize the risk are quantified.

The third category deals with the interactions between individuals, society, planetary processes and natural systems. This is where modeling techniques such as network analysis and graph theory, combined statistical and physics-based models, and framework assessments are used. This is where the knowledge gained from the previous two categories is used as the basis for understanding what actions need to be taken and the impacts they have. The questions that are typically asked here are - how stressed is the system and can our actions mitigate the stresses, where are the stresses being felt globally and within different societies, are there different pathways or routes that can be created to improve societal and ecosystem health and viability? For example, let’s take a look at buildings and their role in climate change and social well being. As human beings, we need buildings for several purposes - to live in, to manufacture and create in, to store and grow food - among other uses. However, the way buildings are designed and built as well as the systems that are used in them can contribute to increased emissions and waste or can help mitigate climate change. A classic example is the use of green roofs in buildings to reduce the level of heat present, thus minimizing the use of air conditioning and greenhouse gas emissions as well as improving the well-being of people in the building. 

All three reports had a lot of very interesting takeaways on the state of the planet and what we can do about it. Next time, we’ll dive into the major results and discuss how modeling and data science have shaped climate science over the last decade.

What our community are reading

Moonshots, Models, IoT and Machine Learning in Agriculture

Our online community space is now live!

How much water should an email consume? Data centers and water use