Technology for Forecasting the Global Environment and Human Society

2.1 Efforts to forecast extreme weather

Extreme weather events, such as typhoons and linear rainbands, have been occurring frequently. To improve accuracy in extreme weather forecasting, it is important to observe the interaction between the atmosphere and ocean as well as improve the accuracy of numerical forecasting models. Ocean temperature and humidity as well as ocean surface and underwater temperature need to be directly observed to improve forecasting accuracy, but the number of sensors is insufficient, making observations along the path of typhoons very difficult. The area directly under a typhoon is very difficult to observe because of the harsh environment in which storms and high waves occur. To conduct real-time ocean observations directly under typhoons, we developed the atmospheric and ocean autonomous observation system “Seiuchi-san” that can autonomously record observations for long periods without human intervention. In July 2022, we commenced observations offshore of Okinawa in collaboration with OIST (Fig. 1). We observed the area near the center of Typhoon Hinnamnor in August 2022. For future work, we will analyze the experimental data to improve accuracy of typhoon forecasts and research real-time continuous observation techniques in the ocean near Japan to improve the accuracy of extreme weather forecasts. We will attempt to broaden the scope of observations in the ocean near Japan as well as in distant oceans where typhoons occur through joint research with the Japan Agency for Marine-Earth Science and Technology.

Fig. 1. Atmospheric and ocean autonomous observation system (Seiuchi-san) and dropping it into the sea.

2.2 Efforts to model marine ecosystems

Marine ecosystems are constantly transforming due to the effect of air temperature, water temperature, and changes in sunlight intensity, which is exacerbated by climate change. The results of human activities, such as agricultural runoff carrying various nutrients and chemicals, flow into the ocean through rivers. It is also difficult to observe ecosystems and quantify changes when we do not fully understand the role of individual organisms in nutrient cycling. Therefore, we aim to observe and model marine ecosystems, including plankton and other microorganisms, fish, and mammals, using new observation technologies that use satellite IoT technology, and to forecast marine ecosystems using simulations of environment changes. We are currently studying ecosystem-cycle forecasting technology that models ocean-ecosystem circulation processes and predicts the effects of climate change and human activities, fish-ecosystem monitoring technology that restores the ecosystem balance by constructing indicators of fish well-being, and microorganism-diversity-modeling technology that models the diversity of millions of microbial species and forecasts changes due to climate change (Fig. 2).

Fig. 2. Marine ecosystem model.

3.1 ESG-intelligence scientific methodology

We collect global information on politics, economy, society, technology, and environment using artificial intelligence (AI) including text mining technology, etc., and scientifically analyze them as information that contributes to the formulation of management strategies regarding ESG. For example, we are collecting and analyzing information necessary to formulate business strategies related to carbon neutrality and energy, in light of the latest efforts to decarbonize around the world and the energy crisis caused by the conflict in Ukraine. To create multiple possible future scenarios, important influencing factors are extracted on the basis of the collected information and set up as branch points in the creation of future scenarios. Figure 4 shows an example of manual analysis with the conventional method. For the same information sources as in the conventional method, by applying AI text mining technology to the aforementioned series of analysis processes, we are attempting automatic information collection and analysis so that the bias of manual information analysis can be eliminated and branch points can be extracted objectively.

Fig. 4. Attribution analysis using conventional method.

3.2 Corporate well-being assessment methodology

To evaluate the contribution of information and communication technology (ICT) to achieve a sustainable society, NTT developed the Gross Social Feel-good Index (GSF Index) [2] for quantitatively evaluating the positive and negative effects of ICT services and solutions in terms of their impact on the global environment, society, and economy (Fig. 5). The corporate well-being assessment methodology is an extension of the GSF Index to establish a method for quantitatively evaluating the sustainability of the global environment and human society in terms of societal and corporate well-being. This methodology is an attempt to evaluate the contribution of corporations toward a well-being society that people desire, particularly for multiple stakeholders and considering non-financial impact. In FY2021, in a joint research project with Kyoto University, we conducted a quantitative evaluation of employee satisfaction in companies by identifying priority issues on the basis of findings from a questionnaire survey.

Fig. 5. Overview of the GSF Index.

3.3 ESG forecasting methodology

The ESG forecasting methodology is used to establish a set of scientific-prediction models to quantitatively predict the results of implementing a single ESG-related management strategy in multiple possible futures using the branch points obtained using the ESG-intelligence scientific methodology. The set of scientific-prediction models consists of a macroeconomic model that predicts economic, social, and environmental changes at the macro level by using input-output analysis [3] and a micro-social model that predicts changes in individual behavior related to environmental impact. By combining these models, the risks and effects of implementing corporate ESG-related management strategies can be predicted that cannot be predicted manually on the basis of changes in social conditions output from the models. The results of the quantitative forecasts are output as scenarios, and the forecast results are evaluated using the corporate well-being assessment methodology. We aim to develop scientific technology that will be useful in the formulation of ESG-related management strategies that can lead to a better future for companies and society as a whole.