> For the complete documentation index, see [llms.txt](https://docs.alphageo.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.alphageo.ai/climate-resilience-suite-crs/climate-resilience-suite-overview/limitations-assumptions-and-data-transparency.md). # Limitations, Assumptions, and Data Transparency At AlphaGeo, we aim to provide robust, transparent, and financially relevant insights into climate risk. While our methodology is grounded in the best available science, financial modeling, and engineering evidence, it is important for users to understand the inherent **limitations, assumptions, and uncertainties** in our approach. This ensures results are interpreted with appropriate context and caution. *** ### 1. Climate data and projections * **Current-condition baseline:** The 2025 climate indicators primarily use ERA5-Land reanalysis averaged over 2020–2025. Reanalysis reconstructs past conditions using models informed by observations; it carries uncertainty and represents a multi-year period. Other hazard layers retain their own source-specific baseline periods. * **Coverage gaps:** Coastal gaps are filled from nearby ERA5-Land cells. Remaining gaps, including small islands, use NASA NEX-GDDP-CMIP6 2025 ensemble values under SSP2-4.5. The selected 2025 physical inputs are shared across scenario requests. * **Baseline-to-projection comparisons:** Future climate indicators use NASA NEX-GDDP-CMIP6 projections. Differences in source methods and reference periods, along with variability within the six-year baseline, can contribute to the measured change from 2025. * **Uncertainty in projections:** Our models use IPCC-endorsed scenarios (SSP245, SSP370, SSP585) to explore a range of plausible climate futures. These do not represent precise predictions but rather structured scenarios. * **Downscaling methods:** We translate global or continental-scale climate model outputs into local impacts using statistical downscaling and proxy measures (e.g., degree days, damage ratios). These methods balance accuracy and practicality but inevitably smooth over local variability. *** ### 2. Model scope and use * **Asset classes:** Our methodology covers a broad set of asset classes (residential, commercial, data centers, infrastructure). It does not account for unique bespoke designs or adaptive management strategies at individual sites. * **Temporal scope:** Most projections are benchmarked against 2025–2050 horizons. Beyond this, uncertainty compounds rapidly, and results should be viewed as indicative rather than predictive. * **Systemic interactions:** Our model assesses direct asset-level impacts. Wider economic feedback loops (e.g., supply chain disruptions, policy shifts, insurance retreat at systemic scale) are not explicitly modeled. *** ### 3. Methodological assumptions * **Two-Step framework:** Our results combine (1) location-based climate exposure metrics with (2) asset-specific sensitivity adjustments. This simplifies complexity but may underrepresent unique micro-level asset conditions. * **Sensitivity coefficients:** Asset class vulnerabilities are represented by multipliers on a 1–5 scale. This approach ensures transparency and scalability but does not capture every engineering nuance. * **Aggregation choice:** Metrics (e.g., Maintenance Cost Increase) are calculated as blended scores before sensitivity adjustment, rather than adjusting each underlying climate driver separately. This avoids false precision but may mask cases where opposing sensitivities exist. *** ### 4. Financial modeling assumptions * **Insurance premiums:** The model sums annualized growth contributions from inland-flood, coastal-flood and wildfire resilience-adjusted impacts over 2025–2050, using a baseline floor to limit sensitivity to very small starting impacts. Stable or declining impacts contribute zero growth, so high but stable exposure can have a low growth estimate. Observed premium growth can differ because of rebuilding-cost inflation, market conditions, reinsurance, regulation and policy terms. A 25-year horizon does not guarantee mean reversion or convergence to the climate-driven projection. Historical rebuilding-cost inflation and premium trends are shown in the UI at the sub-market level where available to provide local context; they do not enter the projection. * **Climate Risk Discount:** Combined resilience-adjusted impacts across nine hazards are scaled against the global reference range for each horizon, using reference rates of 0.8% for 2035, 1.2% for 2050 and 2.0% for 2100. These are modelling assumptions and reference scaling rates, not enforced upper limits. The resulting adjustment is added to the exit capitalization rate; the DCF discount rate remains unchanged. * **CapEx for retrofits:** Additional capital expenditures are estimated on an exponential scale (up to 10% of annual income per hazard). Actual retrofit costs may vary widely depending on design choices, regulatory changes, and supply chain dynamics. * **Operational impacts:** Metrics such as downtime or workforce productivity loss are simplified as percentage changes in annual outputs. Real-world impacts may be non-linear and event-specific. *** ### 5. Uncertainty beyond the model * **Extreme events:** While acute hazard metrics are included, unprecedented “black swan” events cannot be reliably quantified. * **Policy & regulation:** Future changes in building codes, carbon pricing, or adaptation subsidies may materially alter cost structures. * **Market dynamics:** Insurance, financing, and valuation markets are dynamic. Our assumptions represent today’s best estimates but may diverge from future market behavior. *** ### Ethical and transparency commitment AlphaGeo subscribes to the principle of **“decision-useful transparency”**: providing clear, interpretable, and bounded insights rather than overstated precision. We recommend using our analytics as one input among others — complemented by local expertise, engineering assessments, and market intelligence — for robust decision-making. --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation by asking a question. Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters: ``` GET https://docs.alphageo.ai/climate-resilience-suite-crs/climate-resilience-suite-overview/limitations-assumptions-and-data-transparency.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.