Dynamic Evaluation of climate and health field-building and system-strengthening: a protocol for the Advancing Research for Climate and Health global initiative

Climate change poses urgent and unevenly distributed risks to health, while research on climate and health remains underdeveloped and under-representative of low- and middle-income country contexts. The Advancing Research for Climate and Health (ARCH) global initiative addresses this gap by supporting a regional network of researchers focused on strengthening climate-resilient primary health care and health systems. Because ARCH is designed to build a nascent interdisciplinary field and support systems-level change across diverse contexts, conventional monitoring and evaluation approaches are unlikely to provide the adaptive foresight, learning, accountability and strategic feedback required. This paper presents the design and application of a Dynamic Evaluation approach developed for ARCH to support collective learning, adaptation and responsible scaling, and scaling science in complex climate-health systems. Drawing on the International Development Research Centre (IDRC)’s work to cultivate the field of Scaling Science, the ARCH initiative implements DE by cultivating a mindset of collective and inclusive learning across multiple levels, using a decentralized, participatory process through working groups and aligning with a nested Theory of Change. The DE framework is structured around four core components: purposes, priorities, people and processes. These components define learning purposes that go beyond conventional learning and accountability functions, identify "springboard" learning questions to guide evaluative enquiry, establish participatory learning and governance arrangements across local, regional and global levels, and create adaptive processes for foresight reflection, sense-making and accountability. The framework also recognizes the ethical and capacity considerations required to support inclusive participation and adaptive decision-making in complex systems. Drawing on examples from ARCH's regional hubs (West and Central Africa, East and Southern Africa, Asia, Latin America and Caribbean, and the Middle East and North Africa), we illustrate how DE operationalizes a continuous cycle of conjecturing, reflection, sense-making and accountability. Dynamic Evaluation is a novel approach that moves beyond extractive, top-down, snapshot evaluations toward the cultivation of a functioning learning system across hubs, projects and programme governance spaces. In ARCH, DE is designed to strengthen the collective, ongoing learning “evaluative muscle” required to understand emerging change, use evidence to support impact at scale, adapt strategies in complex climate-health systems, distribute learning across local, regional and global actors, and connect learning to decision-making and accountability by making visible progress, uncertainty, trade-offs and adaptation. However, established literature on communities of practice, boundary crossing and adaptive development also highlights risks: learning processes may remain performative, become compliance exercises, fail to bridge institutional and disciplinary boundaries, or be constrained by funder expectations, incentives, time and capacity. The value of ARCH DE will therefore depend not only on the existence of learning processes but also on whether these processes create the trust, autonomy, feedback loops and decision space needed for evidence to shape strategy, relationships and scaling choices.