ESG & Data Driven Climate Impact

 Applications of Data Science for Climate Change Management

1. Introduction 

One of the greatest problems of the 21st century is anthropogenic climate change, which can have devastating effects on the environmental, social, and governance (ESG) scales. Climate data is already used for meteorological predictions and health analysis of atmospheric particles, such as PM2.5[1]. In ESG, there are new applications being developed for climate data for risk management, asset evaluation and control measures for financial impacts from climate change. The environmental sector focuses on conservation of natural resources through climate change emissions reduction, natural resource management, and waste management [2]. With the growth of sustainability consciousness, many companies are investing in the environmental sector, particularly risk prevention and post-calamity insurance. Social and governmental management of environmental factors requires data to be used for climate modelling, evaluations and re-analysis of past data, which can contain several petabytes and thus is being evaluated with cloud-based approaches [3]. This data can be analysed for actionable insights in financial decision-making in ESG, as well as disaster prevention and management related to climate change, and is used by regional ESG firms for paying clients looking for sustainable investment.

2. Data Generation & Accumulation 

1. Variety: The Global Climate Observing System (GCOS) is a collective database for Essential Climate Variables (ECV), as there is such a diverse range of variables for different layers of the biosphere (which the ECV model sub-divides into land, ocean and atmosphere) [4]. This wide range of climate data tends to be in numeric form from multiple units, recorded with various kinds of instruments, from satellites to barometers. This structured numerical data is combined with climate science models to generate evaluations, yet the various instrumentation techniques suggest the data can either be static (manual or interval recordings) or dynamic (real-time observations) depending on the monitoring system, which relies on social infrastructure. The environmental data is only a part for ESG evaluations, as the financial decisions require consideration of the specific type of data dependent on the purpose, such as local or global scale, the type of natural risk presented, the scale of time, technological advances for tackling these risks, the current client’s investment model, etc., hence, the structured climate data is collaborated with semi-structured data like low-carbon patent analysis, policy decision making, and transition risk analysis [5][6]. 

2. Velocity: The velocity of climate data is dependent on numerous factors, such as the geo-spatial infrastructures and social development index, and with a growth in ESG, this data is becoming more valuable. Climate observatories can process up to 20 terabytes of data daily, which is highly valuable for financial stakeholders such as insurance companies for risk management related to climate and weather [8]. With an estimated 100,000 weather observatories globally, and a growing number of ESG firms, the data providence for climate change management is growing geographically, chronologically, and economically [9][10]. 

3. Volume: The volume of climate data is dependent on the specific area being studied (a particular atmospheric particle, surface run-off in a particular region, temperature rise effects on a particular region, etc). This will also include variables such as regional population density, geographical factors, and infrastructures. For example, regional data for climate change adaptation in the Barents Region (1.4 million km²) is only 100 GB, whilst the Database for Policy Decision making for Future climate change (d4PDF) in Japan (377,915 km²) is 3 PB [11][6]. For ESG evaluation of technological advances such as low-carbon patent analysis include databases of 65 million patents, and social development indicators are collected from reserves of United Nations data, which contains 32 databases with 60 million records [5][12]. Such large and varied datasets between climate and socio-economic spheres thus require to be processed with clear objectives in relation to regional and corporate targets.

4. Veracity: The linkage of climate data with financial data presents issues of robustness of the evaluative models. Even though climate data is generally structured, semantics and syntax of the data labelling at a global scale provides a standardization issue for climate modelling and risk evaluations [7]. As the global to local impact of climate data requires downscaling, it can degrade the quality of the dataset, thus challenging the robustness of future risk modelling outputs [7]. Furthermore, semi-structured meta-data is analysed for patents and policies, which require robust AI development for successful implementation. This analysis however needs to consider errors in environmental communication, such as greenwashing and silent-green [13]. Thus, there are many error-managing analytics requirements for increasing the reliability of ESG data, and standardization consideration must be provided for climate data in order to achieve actionable climate change management. 

3. Valuation and Actionable Insights from Data Processing Life-Cycle 

Data Processing Life-Cycle for Climate ESG Analysis: 

• Data gathering is done by data sources such as the GCOS or local observatories, with focuses on geography, ECVs, and is combined with social development data, technological patents and evaluations, company or governmental policies collected by the UN, corporate documents, and IP services. 

• Data selection and quality control is completed via standardization of semantics and syntax of climate data, scaled or converted within a high-quality scale. Robustness of semi-structured data should be processed for reliability via AI and ML techniques for the meta-data present in documents for IP, policies and documents. 

• Data is processed and transformed into ESG models for different purposes (evaluating efficiency of a firm, ESG ranking, auditing, insurance gap and risk analysis, etc) [14].

• Depending on the purpose for data analytics, the data can output the required parameters (environmental productivity, ranking output in relevance to competitors, auditing results, insurance needs in specific sectors, etc). Game theory models and statistical evaluations from the climate data can provide risk and insurance financial analytics for the firms.

• Using the analytics outputs, firms can decide to take appropriate measures for environmental risk mitigation, sustainable investment or business strategies, decreasing emissions or energy conservation, etc. These evaluations for decision making have further impacts on the environment, and an iterative process can be established to maintain trends towards sustainability, and corrected if away from targets.  

ESG assets, with a market value projection of 40.5 trillion USD in the finance industry in the past year, are predominantly growing in the USA and Europe, although it’s picking pace in Asia [15]. With 86% of millennials attentive towards sustainable investments, it’s essential to see a global transition to more environmentally conscious economic models [16]. Collaboration between financial investments and climate predictions is reliant on data analytics on multiple levels and scales, thus the strategies of ESG firms provide a competitive market of growth for big data applications. 

References

1. “Data Science Answers Big Questions like 'How Does Climate Change Affect Your Health?'.” Harvard Business School Digital Initiative, May 15, 2020. https://digital.hbs.edu/data-and-analysis/data-science-answers-big-questions-like-how-does-climate-change-affect-your-health/#:~:text=Data%20science%20can%20tell%20us,to%20predict%20their%20future%20occurrences.  

2. “ESG Investing and Analysis.” CFA Institute. Accessed February 4, 2021. https://www.cfainstitute.org/en/research/esg-investing#:~:text=What%20Is%20ESG%20Investing%3F,material%20risks%20and%20growth%20opportunities.  

3. Setchell, Helen. “ERA5: The New Reanalysis of Weather and Climate Data.” ECMWF, January 7, 2019. https://www.ecmwf.int/en/about/media-centre/science-blog/2017/era5-new-reanalysis-weather-and-climate-data.  

4. “Introduction.” Getting your hands-on Climate data: Introduction. Accessed February 4, 2021. https://nordicesmhub.github.io/climate-data-tutorial/01-introduction/index.html.  

5. “Climate Data and Metrics.” MSCI. Accessed February 4, 2021. https://www.msci.com/our-solutions/esg-investing/climate-solutions/climate-data-metrics.  

6. Nakagawa, Y., Onoue, Y., Kawahara, S. et al. Development of a system for efficient content-based retrieval to analyze large volumes of climate data. Prog Earth Planet Sci 7, 9 (2020). https://doi.org/10.1186/s40645-019-0315-9 

7. Dresen, Martin. “What Kind of Data Is Needed to Identify Climate Impacts? How Can Data Be Managed and Organized through Data Catalogues?” The GIZ-project “Inventory of Methods for Adaptation to Climate Change,” September 2011. https://doi.org/https://www.adaptationcommunity.net/download/climateinformation/ci-tools/GIZ-Deskstudy_Metadata&Catalogue_07-Sep-2011.pdf.  

8. Nash, Kim S. “How to Profit From the Ultimate Big Data Source: The Weather.” CIO. CIO, May 24, 2013. https://www.cio.com/article/2385814/how-to-profit-from-the-ultimate-big-data-source-the-weather.html.  

9. Brandon, John. “How to Collect and Analyze Data from 100,000 Weather Stations.” CIO. CIO, June 16, 2015. https://www.cio.com/article/2936592/how-to-collect-and-analyze-data-from-100000-weather-stations.html.  

10. “ESG Ratings Are Not Perfect, but Can Be a Valuable Tool for Asset Managers.” KPMG. KPMG, October 6, 2020. https://home.kpmg/cn/en/home/insights/2020/10/esg-ratings-are-not-perfect-but-can-be-a-valuable-tool-for-asset-managers.html#:~:text=Currently%2C%20there%20are%20roughly%2030,the%20major%20agencies%20listed%20above.  

11. Benestad, Rasmus, Kajsa Parding, Andreas Dobler, and Abdelkader Mezghani. “A Strategy to Effectively Make Use of Large Volumes of Climate Data for Climate Change Adaptation.” The Norwegian Meteorological institute, July 2017. https://doi.org/https://doi.org/10.1016/j.cliser.2017.06.013.  

12. “UNdata.” United Nations. United Nations. Accessed February 4, 2021. https://data.un.org/.  

13. In, Soh Young, and Ki Young Park. “When Do Firms Oversell or Undersell Environmental Sustainability?: An Empirical Analysis of Sustainability Communications.” SSRN, October 19, 2018. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3264923.  

14. “ESG REPORTING DATA MANAGEMENT SYSTEM.” ESG360. Accessed February 4, 2021. http://www.esg360.com.hk/esg-reporting-data-management-system.html.   

15. Baker, Sophie. “Global ESG-Data Driven Assets Hit $40.5 Trillion.” Pensions & Investments, July 2, 2020. https://www.pionline.com/esg/global-esg-data-driven-assets-hit-405-trillion.   

16. “A Look at ESG Reporting and Sustainable Investment in Hong Kong.” Bloomberg.com. Bloomberg Professional Services , June 24, 2019. https://www.bloomberg.com/professional/blog/look-esg-reporting-sustainable-investment-hong-kong/.