Climate Risk Chat: Get Decision-Ready Insights in Plain English

Christopher DuMont
4 days ago
3 min read

Stop wrestling with jargon. Ask a question about climate risk for any place or asset and get a clear, sourced answer—fast.

What is our Climate Risk Chat?

Climate Risk Chat is a conversational interface inside the Aldabra dashboard. You type a question the way you’d ask a colleague—“Flood risk for a warehouse in Miami in 2050?”—and the chat returns a concise summary, a hazard details table, future projections (when available), and often an interactive map. Every answer includes sources and assumptions so you can trust what you’re seeing.

Note: For now, each question is analyzed fresh. Previous messages aren’t remembered until login is enabled.

How to use it (3 minutes)

Open the chat
In the dashboard, click the panel titled Climate Risk Chat.
Ask a question
Type your question and press Enter to send. Use Shift+Enter for a new line.
Include a location (required)
Paste a place name, an address, or coordinates like 34.05, -118.25.
(Optional) Add future horizons or scenarios
Mention years like 2030, 2050, or 2100.
Say “moderate” (SSP2-4.5) or “high” (SSP5-8.5). If you say “Paris-aligned” / “1.5°C”, the tool may include SSP1-2.6.
Review the results
You’ll see a plain-language summary, hazard details, sources, and sometimes a map. Toggle the map with Show/Hide Map.
Start over
Click Clear to reset the conversation pane.

What kinds of questions can I ask?

Ask about current conditions and future projections for a location or an asset. Here are starter ideas:

Risk overview

“What are the top climate risks for this location?”
“Summarize climate risk at 34.05, -118.25 for 2030 and 2050.”

Flooding

“Show me 2030 and 2050 flood risk for Miami Beach.”
“What is the 100-year flood depth near 29.95, -90.07?”

Heat

“How many heatwave days by 2050 in Phoenix?”
“How will extreme heat change by 2030 at 33.45, -112.07?”

Coastal hazards

“Compare high vs moderate scenarios for coastal flooding in Norfolk, VA.”
“Sea-level rise impacts by 2100 for Charleston.”

Wind, wildfire, drought (where available)

“Wind risk by 2050 for Houston, high scenario.”
“Wildfire likelihood in Sonoma County in 2030.”

Asset-specific

“Flood risk for a warehouse at 40.71, -74.00 in 2050.”
“Risks to a salmon farm near 58.3, -134.4 under high scenario.”
"Wheat farming risk in Iowa in 2030."

Evidence & details

“Show hazard details and data sources for my last answer.”
“What assumptions and limitations were used for these projections?”

Tip: If the chat asks for more info (like a location or asset type), reply with the missing detail and resend your question.

What you’ll see in the answers

AI-Generated Summary
A plain-language explanation (often with an overall risk indicator) that translates the numbers into business-friendly implications.
Hazard Details Table
Probability, means, units, thresholds (where available), plus any notes/errors so you know the caveats.
Future Projections
Metrics for 2030/2050/2100 under selected scenarios when supported.
Sources & Citations
The datasets and tools used for the analysis.
Interactive Map
Spatial overlays (e.g., hazards or FEMA flood zones) when the analysis includes map layers.

Best practices for better answers

Be specific about location: city, address, or coordinates.
Include time horizons: “2030 and 2050” or “by 2100.”
Add a scenario if you care: “moderate (SSP2-4.5)” or “high (SSP5-8.5).”
Mention asset type: warehouse, port, neighborhood, fish farm.

Limitations & transparency

Fresh per question: Prior messages aren’t remembered yet during Beta.
Data coverage: Some hazards or places may have limited data.
Uncertainty: Projections include uncertainty—see Limitations and Sources when shown.

Need ideas? Try these prompts

“What are the future hazards at 37.78, -122.42?”
“Compare high vs moderate scenarios for 2030 and 2050 at 29.76, -95.37.”
“How will heat change by 2050 at 40.73, -73.94?”
“Show 2030 and 2050 flood risk for 34.05, -118.25.”

FAQ

Does the chat remember my previous question?

Not yet. Each question is analyzed fresh during Beta.

Do I need coordinates?

No—an address or place name works. Coordinates simply reduce ambiguity.

Can I choose climate scenarios?

Yes. Say “moderate (SSP2-4.5)” or “high (SSP5-8.5).” If you say “Paris-aligned” or “1.5°C,” the tool may include SSP1-2.6.

Will I see a map every time?

Maps appear when the analysis includes spatial layers (e.g., flood zones). Otherwise you’ll still get a summary, table, and sources.

Where do the numbers come from?

From well-known climate and hazard datasets. Each answer lists the sources and methods used.

Ready to try it?

Open the Climate Risk Chat in your dashboard, paste a location, add a time horizon (2030/2050), and ask your first question. You’ll get an explainable, sourced answer you can share with your team.