Monday 17 Jun 2024
By
main news image

This article first appeared in Digital Edge, The Edge Malaysia Weekly on May 27, 2024 - June 2, 2024

Southeast Asia is increasingly vulnerable to climate change, marked by heatwaves, coastal flooding and devastating impact on agriculture. Technologies such as artificial intelligence (AI), climate analytics and nature-based solutions (NbS) are now being used to predict and mitigate these threats effectively.

Of the 11 countries in the region, four — Indonesia, Myanmar, Vietnam and the Philippines — stand out as being particularly vulnerable, being ranked among the world’s top 15 most at risk from natural disasters, according to the United Nations University Institute for Environment and Human Security (UNU-EHS) in its World Risk Index 2023. This index, based on a statistical model developed in 2011, evaluates the potential risk of 193 countries facing humanitarian disasters driven by extreme natural events and the adverse impacts of climate change.

Seeking to bolster Malaysia’s preparedness for climate emergencies, impact-driven placemaking entity Think City, Boston Consulting Group (BCG) and World Wide Fund for Nature Malaysia (WWF-Malaysia) have joined forces. This collaboration aims to address the pressing need for region-specific discussions and dedicated efforts in Southeast Asia, ensuring the country can effectively tackle the challenges posed by climate change.

Together, the parties set up the Southeast Asia Climate Adaptation and Resilience (SEACAR) Alliance last year to address the gaps in discourse on adaptation and resilience (A&R) in the region following the 27th Conference of the Parties (COP27) in 2022.

Southeast Asia is a major trading region with the rest of the world, which generated an average annual export value of US$1.3 trillion and import value of US$1.2 trillion from 2011 to 2020, noted SEACAR in its “The Might of Nature and the Power of Technology Charting a Climate-Resilient Southeast Asia” report. 

Adding that “a vulnerable Southeast Asia is not just a regional concern — it means a vulnerable globe”. 

The finding revealed only 25 out of 127 cities in the region have clear adaptation plans and the US$10.42 billion received from both the public and private sectors between 2000 and 2019 only represents 1.6% of what is required annually for these cities. 

The vulnerability of the region is only going to be exacerbated as the number of cities are expected to increase by another 200 by 2050. Which is why the alliance believe there is enormous potential to leverage the world’s oldest and newest solutions: NbS — which can potentially save developing countries US$393 billion in cost by 2050 — and technology in the form of AI to help conduct advanced analytics.

“A&R efforts are fundamentally about averting potential challenges, underscoring the importance of well-informed decision-making. Climate AI is instrumental in making informed decisions for climate adaptation and resilience efforts,” says Dave Sivaprasad, Southeast Asia lead for climate and sustainability and managing director and partner at BCG.

Climate AI refers to the application of AI and advanced analytics to climate data, enabling insights and actions in climate adaptation efforts. It involves using technology to simulate different climate scenarios, quantifying the costs of inaction and prioritising resources effectively for A&R projects and programmes.

Climate AI aids in processing extensive climate data, says Sivaprasad, which translates complex information into actionable insights. This empowers decision-makers with a clearer understanding of climate-related challenges and opportunities, facilitating strategic choices.

“Moreover, climate AI enables the generation of concrete, data-driven recommendations for A&R action. It identifies high-priority intervention areas, potential partnerships with private entities holding vulnerable assets and highlights international funding opportunities to ensure that adaptation measures are not only strategic but also impactful in addressing the unique challenges of each region,” says Sivaprasad.

In the SEACAR study, BCG observed sea-level rise four times the global average and increased frequency of severe tropical cyclones. By leveraging climate AI and analytics, the consulting firm quantified specific statistics, such as 78% of the vulnerable population concentrated in four municipalities out of 21 municipalities in Southeast Asia are highly exposed to climate risks, which could result in an estimated US$3 billion in GDP loss. This kind of data-driven insight allows for targeted interventions, addressing both social and economic impacts.

The SEACAR team at COP28 (Photo by SEACAR)

Integrating nature-based solutions with AI

Harnessing AI alongside climate modelling can be a game-changer in unravelling the complexities of quantifying the diverse benefits of NbS for climate adaptation.

“The process involves the precise quantification of various future climate scenarios using vast amounts of data and their multipronged impacts as well as the various benefits NbS can provide in addressing the potential climate risks,” says Lavanya Rama Iyer, director of policy and climate change at WWF-Malaysia.

Lavanya says it can also assess other aspects or co-benefits such as carbon sequestration capability, reducing health stressors and impact on local biodiversity. Carbon sequestration is the process of capturing and storing carbon dioxide from the atmosphere to mitigate climate change by reducing greenhouse gas emissions.

One example of a successful NbS project is the Bishan-Ang Mo Kio Park in Singapore, which transformed a man-made storm drain into a naturalised canal within the Kallang River. “The naturalised park provides ecosystem services such as flood control, soil erosion control and mosquito control and water treatment, and has increased biodiversity by 30%,” she says.

Beyond quantification, AI models contribute to decision-making by analysing trade-offs between different benefits associated with NbS adaptation measures. This analysis provides decision-makers with insights into potential conflicts or synergies among various aspects.

“Moreover, AI models exhibit adaptability by integrating diverse datasets related to NbS adaptation measures. This integration encompasses information on mitigation, adaptation, biodiversity and other relevant factors, offering decision-makers a nuanced perspective,” she says.

But is AI really the best way to approach climate challenges?

Sivaprasad thinks so. NbS and AI offer a robust strategy across planning, financing and implementation, providing a comprehensive approach to climate action.

“In planning, AI collaborates by modelling climate impact and incorporating existing green and grey infrastructure. This joint effort enhances the baseline assessment, evaluating ecosystem health and service provision,” he says.

AI integrates data on NbS benefits, from mitigation to biodiversity, into models for return on investment (ROI) calculation and trade-off analysis during financing, Sivaprasad explains.

“AI, in implementation, will be able to optimise solutions through simulations. This helps with the technical assessment of NbS projects for factors like cost, sustainability and co-benefits.”

In the realm of climate resilience, AI can aid with value-based prioritisation, or the practice of ranking options or tasks according to their perceived importance or alignment with overarching goals and values.

“Value-based prioritisation relies on several key criteria and metrics to guide decision-makers effectively. One significant metric to consider is the cost of inaction,” says Hamdan Abdul Majeed, managing director of Think City.

According to the SEACAR report, the cost of inaction is particularly high for Southeast Asia, which is primarily made up of low- and middle-income nations. The annual damage bill for natural disasters on essential assets like power and transport could reach a staggering US$18 billion in these countries.

The cost of inaction metric provides a framework for decision-makers to make informed choices and allocate resources efficiently in the pursuit of climate resilience.

This evaluation involves a comprehensive understanding of the consequences that may arise from inaction and weighs them against the costs incurred by taking proactive measures, says Hamdan.

Forecasting climate scenarios and resource limitations

Forecasting climate scenarios encounters inherent limitations due to the intricate and interconnected nature of climate systems, characterised by numerous variables. This complexity poses a formidable challenge in achieving accurate long-term predictions.

To address these limitations, AI can leverage advanced analytics and machine learning algorithms, enabling scenario simulations that consider a myriad of factors and potential uncertainties.

“One of the key strengths of AI lies in its adaptive learning feature. This ensures a continuous enhancement of models over time as they evolve with the infusion of new data,” says Sivaprasad.

He adds that, by constantly adapting and learning from real-world observations, AI models contribute to refining future climate scenarios, enhancing their accuracy and reliability in addressing the uncertainties inherent in climate forecasting.

The question of equitable distribution of the benefits of AI and NbS remains, especially in regions with limited resources in Malaysia. Lavanya stresses the importance of an inclusive adoption process, involving all stakeholders at all stages.

Aligning with the principles of Locally Led Adaptation (LLA), she suggests two principles that should be adopted. One being developing decision-making at the lowest appropriate level.

The principles for LLA, developed by the Global Commission on Adaptation at the 2021 Climate Adaptation Summit, aim to guide the adaptation community in shifting towards locally owned adaptation efforts, fostering a community of practice to share progress and lessons learnt.

“[The country needs to] prioritise resource allocation based on the specific needs and vulnerabilities of regions with limited resources. This aligns with devolving decision-making to the lowest appropriate level, empowering local institutions and communities,” says Lavanya.

Moreover, the structural inequalities faced by women, youth and children, as well as the disabled, the displaced, indigenous peoples and marginalised ethnic groups, should be addressed as well.

Integrating these principles into the implementation of AI and NbS initiatives ensures the benefits will reach all segments of the population and meets the needs of those who are already marginalised, particularly those in regions with limited resources, to foster resilience and sustainability, says Lavanya.

Integrating NbS and climate AI opens up a field of potential, specifically in establishing benchmarks and frameworks. Currently, efforts are underway to develop frameworks and benchmarks for assessing the success of these initiatives.

“Key benchmarks for success include reduction in loss and damage from disasters, improved environmental indicators, increased economic benefits from NbS interventions, enhanced social resilience in communities and efficient knowledge sharing in NbS and climate AI efforts,” she says.

As projects progress and data accumulates, Lavanya sees refined benchmarks and assessment methods are likely to emerge, ensuring a robust and effective monitoring and evaluation for future initiatives.

Moving forward

According to a report by AI for the Planet Alliance, produced in collaboration with BCG in 2022, 87% of climate and AI leaders from both public and private sectors acknowledged the efficacy of advanced analytics and AI in combating climate change.

Moreover, 43% of organisations have a defined vision for utilising AI in their climate change efforts, indicating a widespread interest in harnessing the potential of this transformative tool. Over 50% of industrial goods companies expressed a vision for AI integration, contrasting with 30% for consumer companies.

In Malaysia, the integration of AI in climate action is nascent but burgeoning, driven by growing recognition of its potential benefits and increasing availability of AI expertise. The adoption of AI holds promise across sectors such as renewable energy, agriculture and disaster management.

Artificial intelligence can optimise energy production and distribution from sources such as solar and wind power, while predicting energy demand. In agriculture, AI-powered tools are being employed to improve crop yields, manage water resources and detect plant diseases.

Additionally, it is proving valuable in disaster management, aiding in the prediction of and preparation for natural disasters such as floods and landslides.

“Several factors contribute to the receptiveness of certain regions and sectors to AI adoption, including the availability of data, government support through policies and funding, private sector investment and increasing public awareness of AI’s potential to address climate change,” says Hamdan.

He sees a promising future for AI in climate action in Malaysia. With continued investment, collaboration and innovation, the technology has the potential to play a significant role in helping the country achieve its climate goals, he believes.

“To ensure the success of integrated NbS and climate AI initiatives in Malaysia, climate AI and modelling must go hand in hand. A localised model is crucial for greater accuracy and efficiency of success,” says Lavanya.

Critical to this effort is also the implementation of monitoring and evaluation methods, such as bio-indicators and ecosystem services valuation as well as social vulnerability and capacity assessments.

These methods track changes in key environmental indicators, including water quality, air pollution and biodiversity, while assessing the economic value of services provided by NbS, such as flood protection and carbon sequestration, she says.

Impact assessments, further enhanced by AI-powered scenario modelling, simulating the long-term effects of NbS interventions under different climate change scenarios, are equally important.

“Utilising Earth Observation (EO) data from satellites and drones as monitoring tools, along with participatory monitoring and evaluation involving local communities, ensures culturally appropriate interventions that meet their needs,” says Lavanya.

The Ministry of Natural Resources and Environmental Sustainability has placed focus on A&R, evident in the dedicated discussions at COP28 and ongoing formulation of the country’s inaugural National Adaptation Plan. But to expedite the adoption of NbS and AI, strategic communication efforts are crucial to articulate their compelling merits and garner broader support.

While the government’s undeniable potential to spearhead this transformative initiative is acknowledged, Lavanya, Hamdan and Sivaprasad concur that a comprehensive narrative highlighting the synergies between AI and NbS remains pivotal. Additionally, positioning AI as a facilitator for private sector involvement creates a new dimension, encouraging widespread participation in addressing environmental challenges.

The primary goal of the report and initiative is to issue a compelling call to action, going beyond raising awareness of critical climate challenges. The SEACAR Alliance aims to catalyse collective action in response to the escalating climate challenges in Southeast Asia.

The establishment of the SEACAR Alliance in 2023 stems from a shared commitment to climate action. BCG contributes its expertise in collaboration with the government and the private sector, Think City provides on-the-ground insights and WWF-Malaysia leverages its extensive global network and expertise on nature.

 

Satellite imagery and machine learning

Malaysian rainforests’ hyperdiversity means each forest patch varies. Restoration is not a one-size-fits-all. Projects must target specific native species based on the unique pre-degradation makeup of each patch to effectively restore these complex ecosystems.

In an interview with Digital Edge, Dr Dzaeman Dzulkifli, executive director of Tropical Rainforest Conservation and Research Centre (TRCRC), speaks on how satellite imagery and machine learning can facilitate the restoration and management of forests in Malaysia.

TRCRC was established in 2012 to restore tropical rainforests and address the critical rate of biodiversity loss in the country. The centre reconnects forest fragments through the reforestation of degraded patches using native trees sourced from nurseries as well as seedlings from rescued threatened plants.

“Using a combination of different species with a wide range of genetic composition will have a lot of positive downstream repercussions such as increasing its resilience to climate change and weather extremes that are on the current rise,” says Dzaeman.

Planting the wrong trees can crowd out native species that are crucial for a healthy ecosystem. This simplification reduces overall biodiversity, which then weakens services such as clean water filtration, flood control and carbon sequestration.

“So, we’ve been working on the MYFarmTrees app where the main idea is, with the geolocation and [information] you have, to collate a list of tree species that can be planted or used within a patch of forest that you want to restore,” he says.

The platform is funded by the Global Environment Facility, implemented by the International Union for Conservation of Nature and executed by the Alliance of Bioversity International and the International Center for Tropical Agriculture. The platform provides decision-support tools from seed selection to tree planting and maintenance and helps engage with smallholders and rural community members in their restoration efforts.

Even people with no background in forestry can use the app to pinpoint their land on a map and get recommendations on suitable tree species based on geolocation data. The app gathers information on suitable tree species from various sources, making previously scattered knowledge accessible in one place.

“So, from there, we’re going to have a list of different trees along with the different sorts of services those trees provide such as carbon sequestration, biodiversity support, food production, watershed protection including ecosystem profiles, and suitability for what sort of habitats such as riparian zones,” says Dzaeman.

Thinking about wider accessibility for communities in rural areas, he is looking into solar power for better accessibility for devices to be charged and be used off the grid. “The nice thing about it is it can work offline because of the GPS system in your phone,” he adds.

However, the biggest challenge is the resolution of the images. “When you scan a forest, the commercially available mapping resolution will give about a 30m resolution. Some of the higher-quality ones, maybe about 5m resolution and some of the really powerful resolutions are 1m, but they are costly,” he says.

While satellite imagery is pricey due to subscription fees and its launch cycles, Dzaeman is exploring light detection and ranging (LiDAR), a drone-mounted laser scanning technology, to gather detailed data on smaller areas like forests.

“And then that information you get, you can correlate or check or use it as a reference with more conventional satellite imagery. You need to kind of like overlay the two different textiles,” he says.

Now, the focus lies in establishing clear standards for environmental practices. With differentiating satellite imagery analytics using different methodologies, it is important for Malaysia to standardise the methodology and use the most updated and highest-quality images to make sure that it is beneficial for everyone in Malaysia, says Dzaeman.

 

AI carbon management software

Managing and organising data for carbon emissions can be quite complex, especially when you have to manually sift through a massive amount of disorganised data on a shared drive to calculate your company’s carbon footprint. Not only would the data quality be unreliable, but it also increases the risk of making mistakes in your efforts to reduce emissions.

“The challenges with managing or getting calculations for your carbon emissions right is because of data collection. If you’re lucky, you might have everything in one place, but most of them don’t,” says Eong Tat Ooi, CTO and co-founder of Pantas Climate Solutions.

Simplifying data collections and extraction, Pantas’ AI carbon management software helps companies better manage and reduce their carbon emissions through the power of AI.

“The platform is also able to perform data identification to match with the proper emissions factors to do the calculations based on the data companies provide to suggest ways to reduce their carbon emissions,” says Max Lee, CEO and co-founder of Pantas.

“We train our natural language processing (NLP) to understand invoice descriptions and group similar items. It then applies the appropriate emission factor to each category, leading to faster, more accurate carbon footprint calculations,” says Eong.

Moreover, the platform also automatically generates annual reports for companies to review their objectives in carbon reduction. “After [giving suggestions on carbon reductions], we will also get our banking partners to come in and provide the financing solutions which are usually cheaper, because the impact is measurable through our software,” says Lee.

“By creating application programming interfaces (API) connections to your enterprise resource planning (ERP) system, we can automatically pull in all your purchase invoices and data. This eliminates the need for manual data entry, saving you time and ensuring accuracy,” says Eong.

The AI carbon management software uses the API to access data stored in the company’s ERP system. With this information, the AI carbon management software can calculate the emissions associated with your company’s carbon emissions and provide accurate insights for carbon reduction efforts.

Pantas’ AI carbon management software adheres to the payment card industry compliance even though they don’t process payments for data security. They have external security experts regularly scan their systems to identify and address vulnerabilities, emphasising on preventing insecure data from entering the system and maintaining its quality.

Save by subscribing to us for your print and/or digital copy.

P/S: The Edge is also available on Apple's App Store and Android's Google Play.

      Print
      Text Size
      Share