Building a Climate Data Platform: Balancing Downscaling Resolution, Stakeholder Needs, and Service Delivery

Climate change adaptation requires actionable information at scales relevant to decision-making. We present the development of a climate data platform (https://ccrab.rcec.sinica.edu.tw/) that integrates downscaled climate projections to deliver accessible climate services for diverse users in Taiwan. The platform architecture employs advanced downscaling techniques to transform global climate model outputs into high-resolution datasets, coupled with user-friendly visualization and data access tools that bridge the gap between climate science and practical application. Beyond research applications, the platform addresses growing demand for climate risk data in financial sectors, providing standardized projections that support Task Force on Climate-related Financial Disclosures (TCFD) reporting requirements and climate risk assessments for businesses and financial institutions.

A critical challenge in developing effective climate services lies in meaningful stakeholder engagement. Understanding the diverse needs of decision-makers across sectors, from water resource management to agricultural planning and disaster risk reduction, requires sustained dialogue and iterative co-design processes. This engagement is complicated by the technical complexity of climate data, varying levels of climate literacy among users, and the need to balance scientific rigor with practical usability.

Determining optimal spatiotemporal resolution presents a fundamental technical and practical challenge, particularly acute in regions with steep topographic features such as Taiwan, Japan, and the European Alps. In these mountainous terrains, climate variables can vary dramatically over short distances due to elevation gradients, orographic effects, and valley-plain transitions. While stakeholders often request the finest possible resolution to capture these local variations, computational constraints, data storage limitations, and uncertainties inherent in downscaling methods necessitate careful trade-offs. The challenge intensifies when complex topography creates microclimates that even high-resolution models struggle to represent accurately, which is a critical issue for Taiwan, a small island country with rough terrains. We discuss our approach to identifying appropriate resolutions for different applications and regions, considering both scientific validity and stakeholder requirements, while acknowledging that higher resolution igher accuracy in topographically complex areas. The platform ultimately aims to provide climate information that is both credible and usable for adaptation planning and climate risk assessment.