Asia Climate Risk Screener: A Zero-Backend Browser Tool for Asset-Level Physical and Transition Climate-Risk Screening
DOI:
https://doi.org/10.31224/8047Keywords:
climate risk, physical risk, transition risk, decarbonisation, power sector, Asia, open data, climate finance, decision support, zero-backend web application, evaluationAbstract
Asset-level climate-risk analytics remain priced for large institutions, leaving many financiers, insurers and plant owners without an accessible first-pass screen. This paper presents the Asia Climate Risk Screener, a single self-contained web page that screens physical and transition climate risk for more than 15,000 geolocated power stations across Asia — all operating Asian coal plants in the GEM census, the wider satellite-estimated fossil fleet of gas and oil assets, plus the solar, wind, hydropower, nuclear, geothermal and bioenergy fleet — using only open satellite and climate data. Selecting any asset builds a plain-language risk profile in seconds from four live public feeds — six decades of measured weather (ERA5), river-discharge records since 1984 (GloFAS), climate-model projections to 2050 (CMIP6), and satellite-derived solar and wind resource (NASA POWER) — joined to a bundled satellite-emissions asset layer (Climate TRACE). Feature engineering condenses multi-decadal daily weather and river records into decision-ready signals, and a transparent heuristic scoring model flags heat stress, flood trends and transition exposure, with explicit "feed unavailable" states rather than silent gaps. I evaluate the hazard-data pipeline against documented Asian climate disasters using a two-arm recall-and-specificity harness, detrended and graded at the 95th percentile. Cross-matching the coal fleet against an independent facility census confirms 96% coverage. The system is zero-backend — no server, account or API key — is deployed as a static site, and is released openly.
Downloads
Downloads
Posted
License
Copyright (c) 2026 Ka Wah Philip Ng

This work is licensed under a Creative Commons Attribution 4.0 International License.