Challenges of manual climate risk assessment

The biggest problem with today’s physical climate risk models isn’t the hazard data or climate variables, it’s how buildings are treated. In many models, building vulnerability is an afterthought, if it’s considered at all. Why? Because the current generation of climate risk tools relies on loss models originally developed for the insurance industry or publicly available resources like FEMA’s HAZUS. These models use top-down historical claims data, such as NFIP claims, and are built for portfolio-level or regional loss estimation, not for understanding specific losses or damage causes at individual properties. HAZUS itself acknowledges this limitation. When you look at the underlying claims data, the issue becomes clear: losses are highly scattered even for the same hazard intensity (e.g. flood depth vs loss data from NFIP claims collected since the 1970's). The best these models can do is group buildings into broad archetypes and fit loss curves through a wide cloud of outcomes to represent the average. But, there is no explanation of why one claim results in zero loss while another leads to a total loss for the same flood depth because that data isn't being collected. The insurance industry can live with this (for now) because the logic works good enough at scale. Across thousands of properties, severe over- and under-estimates tend to balance out. But at a single-asset level, that variance matters enormously. It can significantly alter perceptions of risk and more importantly, influence your decisions about individual buildings including retrofits, capital planning, acquisitions, or site selection. The disconnect between portfolio logic and asset-level decision-making is where many climate risk assessments fall short today. If we want climate risk analysis to inform real decisions, we need to model vulnerability very differently.

Their almond supplier faces 30% water cuts next season. Their risk assessment said: not material. (a due diligence story in 3 acts) I was assessing a food brand's climate risks for an investor. The brand had done the homework, or so it thought. A big consultancy had prepared its sustainability assessment. - Polished.  - Confident. - Professional. The conclusion? Water wasn't material. One small detail: a key ingredient was almonds from southern Spain. A water-intensive crop, grown in a region facing severe allocation cuts. 📌 Act 1: The gap I started where I always start - at sourcing regions. - Their assessment used generic industry benchmarks. - Nobody had mapped where ingredients actually came from. - Nobody had cross-referenced suppliers with climate projections. "Spanish almonds" was a line item.  It didn't show as a drought risk. 📌 Act 2: The conversation I flagged the water exposure in my assessment. The investor asked the brand directly: "Walk us through your water risk." Brand: "Our consultants said it wasn't material." Investor: "Your almond supplier is in a region with forecasted 30% allocation cuts. How is that not material?" Silence. 📌 Act 3: The lesson The deal didn't fall apart. But confidence took a hit. The brand had paid for an assessment that missed what a 10-minute sourcing conversation would have caught. Because generic templates don't ask "where do your almonds grow?" What I've learned from doing these assessments: Climate risk lives in the specifics. - Which seasons, not which years - Which regions, not which countries - Which suppliers, not which categories Before your next investor conversation, ask your team: → Do we know the climate risks in our top 5 sourcing regions? → Has anyone mapped our suppliers to water stress data? → Would our answers survive a 10-minute follow-up? Investors are starting to ask these questions. Better to discover the gaps yourself first. P.S. What's ONE climate risk you discovered hiding in your supply chain?

I was talking to the sustainability reporting lead at a Fortune 500 company a few weeks back. They did their climate risk assessment with a Big Four consultant back in 2022. [Insert generic expensive, slow, painful experience here.] Here's the thing: they're actually one of the lucky ones. The consultant handed over the underlying Excel models. Most companies just get a PDF or a PowerPoint deck. Static. Untouchable. "Here's your risk, good luck." But even with the spreadsheets, they're stuck. Since 2022, they've acquired another company — massive footprint expansion, new supply chains, new facilities across multiple regions. Their climate risk model hasn't caught up. The Excel files are still sitting there, frozen in time, while the business has completely changed shape. This is the pattern I keep seeing: → Climate risk work happens once, as a consulting engagement → The deliverable is static — a report, a deck, or (if you're lucky)a spreadsheet → The business keeps moving — acquisitions, divestitures, new suppliers, new offices → The risk model doesn't Climate risk is inherently dynamic. Your assets change. Your suppliers change. Natural disasters happen. Regulations shift by jurisdiction. But most companies are still relying on a snapshot from years ago. The problem isn't the quality of the original work. It's that the format can't keep up with the pace of the business. Static deliverables for dynamic work. That's the structural mismatch.