Experts Map Climate Change Drivers in Indonesia Using Combined Modeling Approach

An open-access study published in Discover Environment on 22 October 2026 has applied an integrated modeling framework to map how fourteen climate change drivers relate to one another in Indonesia, and the results place governance—not technology—at the base of the system. Researchers led by Asep Marfu combined two established decision-analysis techniques, Interpretive Structural Modeling (ISM) and the Decision-Making Trial and Evaluation Laboratory (DEMATEL), to build a hierarchy of climate drivers and quantify the causal influence flowing between them. The work is exploratory and rests entirely on expert judgment, but its central message is striking: the factors usually treated as solutions, such as renewable energy adoption, appear at the top of the hierarchy as dependent outcomes, while regulatory weakness and institutional capacity sit at the foundation.

The methodological pairing is the study’s core contribution. ISM takes pairwise judgments about how factors influence one another and converts them into a structured hierarchy, identifying which elements are foundational and which are resultant. DEMATEL complements this by scoring the intensity of causal links, aggregating expert ratings into a direct-relation matrix, normalizing it, and computing a total-relation matrix that captures both direct and indirect effects. From that matrix the researchers calculated two metrics for each driver: prominence, the sum of influence given and received, and relation, the net difference between them. A positive relation marks a cause factor that pushes the system; a negative relation marks an effect factor that is pushed by it. All computations were implemented in Python 3.13.5 using NumPy, Pandas, and Matplotlib.

The evidence base came from a purposively selected panel of twelve Indonesian experts drawn from academia, government, non-governmental organizations, and the private sector, each with at least seven years of professional experience in climate change, environmental management, sustainability, or public policy. Data were collected through structured questionnaires and online interviews across two rounds of a modified Delphi procedure. In the first round, experts independently assessed relationships among the proposed drivers; aggregated, anonymized results were then returned to the panel so members could reconsider their judgments in light of group responses. Consensus was deemed achieved when at least 80 percent of experts agreed on the direction of a relationship, and only judgments that remained stable across rounds were retained in the final matrices.

The fourteen drivers themselves were not chosen arbitrarily. A systematic review of peer-reviewed literature published between 2020 and 2025, IPCC assessment reports, and Indonesian policy documents including the Nationally Determined Contribution and the National Action Plan for Climate Change Adaptation produced an initial pool of 26 candidate variables. A three-stage screening process then applied three inclusion criteria: relevance to climate change in the Indonesian context, frequency of occurrence in the literature and policy documents, and potential systemic influence on other climate-related factors. Variables receiving less than 80 percent expert agreement, or overlapping conceptually with broader variables, were dropped. Excluded candidates included climate finance availability, technological innovation, carbon pricing, green infrastructure, circular economy practices, and disaster preparedness—factors the experts judged to be policy instruments or consequences of deeper processes rather than fundamental driving variables.

The ISM analysis partitioned the fourteen retained drivers into four structural levels, and the composition of those levels carries the study’s most consequential implication. Weak environmental regulations (V10), climate governance and institutional capacity (V13), public environmental awareness (V12), the economic development model (V11), and deforestation (V1) occupy the lowest level, marking them as the foundational drivers that shape the behavior of the entire system. An intermediate tier comprising industrial emissions, fossil fuel dependency, and four other variables transmits and amplifies that influence upward. At the highest level sit renewable energy adoption (V14), agricultural practices (V7), and population growth (V9)—variables the model treats not as independent levers but as cumulative outcomes of upstream interactions.

The DEMATEL results reinforce that picture with quantitative symmetry. Seven variables classified as causal and seven as effect, with deforestation showing the highest positive relation value and therefore the strongest causal push in the system. Governance-related variables—weak environmental regulations, public environmental awareness, and climate governance and institutional capacity—also fell in the causal group, influencing the broader system more than they are influenced by it. Renewable energy adoption recorded the lowest relation value, confirming its position as a dependent outcome rather than a primary driver. In the impact-relation map, governance, deforestation, industrial emissions, and fossil fuel dependency clustered in the causal quadrant with high prominence, while renewable energy adoption, agricultural practices, and population growth sat in the effect quadrant.

The authors argue that this convergence between the two methods strengthens the findings: variables identified as foundational in the ISM hierarchy also emerged as causal in DEMATEL, a consistency they read as evidence that governance reform and institutional strengthening represent the most effective leverage points for systemic climate mitigation. On this reading, renewable energy expansion depends on prior improvements in governance quality, institutional coordination, and environmental regulation—a framing that departs from conventional climate narratives that treat clean energy deployment as the natural starting point for action. The study instead suggests that downstream interventions are likely to be more effective once upstream structural conditions improve.

The Indonesian context gives the analysis its urgency. The paper describes Indonesia as among the countries most vulnerable to climate impacts owing to its geography as a tropical archipelago with diverse ecosystems, citing historical data indicating approximately a 25 percent increase in the amplitude and frequency of extreme rainfall across northern regions such as Kalimantan, northern Sumatra, parts of Sulawesi, and Papua, and drought intensity rising by up to 60 percent in drier southern regions including southern Sumatra, Java, and Nusa Tenggara. Those trends, the authors note, are projected to persist under future Shared Socioeconomic Pathways and Representative Concentration Pathways scenarios. High rates of deforestation, agricultural expansion, and uncontrolled urbanization are described as exacerbating the crisis, while existing mitigation and adaptation programs are characterized as fragmented and lacking synergy.

The policy implications drawn from the model are specific. The authors recommend prioritizing cross-ministerial coordination among the environment, energy, agriculture, and forestry ministries and regional governments; improving regulatory enforcement and harmonizing national and regional climate policies; strengthening land-use governance through stricter forest protection, better monitoring of illegal land conversion, and tighter integration of climate objectives into spatial planning and permitting; and gradually reducing fossil-fuel subsidies while accelerating renewable infrastructure investment and grid modernization. Downstream variables such as renewable energy adoption, they argue, should be viewed as outcomes of successful governance reform rather than initial intervention targets, with the integrated framework serving as a decision-support tool for prioritizing actions under limited financial and institutional resources.

The study’s limitations are stated plainly, and they matter for how far the conclusions should travel. The analysis relies wholly on expert judgment, which introduces subjectivity; different expert panels could produce different structural relationships even with the two-round Delphi process and 80 percent consensus threshold. The panel of twelve experts, while consistent with the 10-to-20 participant range common in ISM and DEMATEL research, may not capture the full diversity of perspectives across Indonesia’s climate governance system. The findings are specific to the Indonesian context and should be generalized to other countries with caution, and the model offers a static snapshot of causal relationships that does not account for how climate systems evolve dynamically over time. The authors recommend future work with larger and more diverse panels, cross-country validation, and integration with complementary approaches such as MICMAC, system dynamics, scenario analysis, or GIS-based modeling. Within those bounds, the study offers developing-country policymakers a transferable template for asking a deceptively simple question before spending on any climate intervention: which factors actually drive the system, and which merely respond to it.

Subject of Research: Structural and causal analysis of climate change drivers in Indonesia using an integrated ISM–DEMATEL expert modeling framework

Article Title: Integrative analysis of climate change drivers using ISM DEMATEL approach for structural decision making

Article References: Marfu, A., Tanubrata, D., Pratiwi, D. I., Rahmayanti, H., Hasanah, U., & Astra, I. M. (2026). Integrative analysis of climate change drivers using ISM DEMATEL approach for structural decision making. Discover Environment, 4(1), Article 593. https://doi.org/10.1007/s44274-026-01139-8

Image Credits: AI Generated

DOI: 10.1007/s44274-026-01139-8

Keywords: climate change, Indonesia, ISM, DEMATEL, deforestation, environmental governance, renewable energy, expert panel, systems modeling, climate policy, multi-criteria decision-making, Integrative