About
DeepCyc by Reask is a next-generation climate-adjusted tropical cyclone (TC) catastrophe model purpose-built for a changing climate. Unlike traditional cat models that rely on historical data alone, DeepCyc uses AI-driven atmospheric simulations to generate more than 100 million physically realistic stochastic events, each conditioned on real and projected climate states — from current ENSO phases through to 3°C+ global warming scenarios. The model delivers global coverage across all tropical cyclone basins with a consistent probabilistic framework at an exceptionally fine 1×1 km spatial resolution, including terrain-adjusted 3-second wind gust simulations that account for local topography and surface roughness. Its event catalogue spans 1900–2100, offering both 20,000-year and 100,000-year return-period sets for extreme tail-risk analysis. DeepCyc integrates with Reask's Climate-Based Risk Adjustment (CBRA) tool and can be benchmarked against building design wind speed standards, fitting naturally into existing actuarial and engineering workflows. Use cases range from catastrophe model benchmarking and seasonal risk assessment to regulatory stress testing and climate-scenario pricing for property and specialty lines. Trusted by industry leaders in insurance and reinsurance, DeepCyc represents a step-change in how the (re)insurance industry quantifies, prices, and communicates evolving tropical cyclone risk.
Key Features
- Climate-Aware Stochastic Catalogue: Generates over 100 million physically realistic tropical cyclone events conditioned on observed and projected climate states, from historical baselines through ENSO phases to 3°C+ warming scenarios.
- Global Coverage at 1km Resolution: Delivers a consistent probabilistic view of TC risk across all ocean basins worldwide at a 1×1 km spatial resolution, including terrain-adjusted 3-second wind gusts accounting for topography and surface roughness.
- Extended Temporal Range (1900–2100): Provides event sets spanning the full 20th and 21st centuries with 20k- and 100k-year catalogues, enabling robust tail-risk and return-period analysis.
- Seamless Workflow Integration: Integrates with the Climate-Based Risk Adjustment (CBRA) tool and can be benchmarked against building design wind speed standards for use in actuarial, engineering, and regulatory workflows.
- AI-Driven Atmospheric Simulations: Uses machine learning grounded in real atmospheric physics to simulate how tropical cyclone behaviour evolves under different climate conditions, connecting all TC basins into one coherent framework.
Use Cases
- Insurance and reinsurance portfolio stress-testing under multiple climate warming scenarios to quantify evolving catastrophe loss potential.
- Climate-adjusted pricing of property and specialty lines exposed to tropical cyclone risk in any global basin.
- Regulatory and internal capital modelling using long return-period event catalogues (20k- and 100k-year) for solvency and resilience assessments.
- Engineering and site-level wind hazard analysis using high-resolution terrain-adjusted gust outputs for building design and infrastructure resilience planning.
- Seasonal and near-term risk assessment by conditioning the event catalogue on current ENSO state to inform dynamic underwriting and portfolio management decisions.
Pros
- Physically Realistic and Climate-Consistent: Events are generated using AI simulations rooted in actual atmospheric behaviour, making the model more reliable than simple statistical extrapolations for future climate states.
- Exceptional Spatial Resolution: The 1km resolution with terrain adjustment provides granular, site-level wind hazard estimates that outperform coarser cat models for property-level underwriting and engineering assessments.
- Full Climate Scenario Coverage: Supports risk quantification from present-day conditions through long-term warming pathways, enabling insurers to stress-test portfolios and price climate transition risk.
- Globally Consistent Framework: All TC basins are modelled within a single coherent system, eliminating inconsistencies that arise when combining outputs from multiple regional models.
Cons
- Enterprise-Only Pricing: Access requires a demo request and is targeted at large insurers, reinsurers, and institutional risk managers, placing it out of reach for smaller organisations or individual researchers.
- Domain Expertise Required: Effective use of DeepCyc requires strong background knowledge in catastrophe modelling, climate science, and actuarial workflows, limiting accessibility for non-specialists.
- Limited Public Documentation: Detailed technical specifications, pricing tiers, and integration guides are not publicly available and require direct engagement with Reask's sales team.
Frequently Asked Questions
What is DeepCyc and who is it for?
DeepCyc is Reask's global tropical cyclone catastrophe model designed for insurers, reinsurers, catastrophe modellers, and risk managers who need to quantify, price, and stress-test TC risk under current and future climate conditions.
How many events does DeepCyc generate?
DeepCyc generates over 100 million physically realistic stochastic tropical cyclone events, with catalogues available at 20,000-year and 100,000-year return periods for extreme tail-risk analysis.
What climate scenarios does DeepCyc support?
DeepCyc is conditioned across a range of climate states including current ENSO phases (El Niño, La Niña, neutral) through to long-term warming pathways of 1.5°C, 2°C, and 3°C+ above pre-industrial levels.
How does DeepCyc integrate with existing workflows?
DeepCyc outputs can be integrated with Reask's Climate-Based Risk Adjustment (CBRA) tool or used to benchmark against building design wind speed standards. Its API-accessible data fits into actuarial, engineering, and risk management pipelines.
What spatial resolution does DeepCyc provide?
DeepCyc delivers hazard outputs at 1×1 km resolution globally, including terrain-adjusted 3-second wind gust simulations that account for local topography and surface roughness conditions.
