Your forecast, the Moon and the risk-assessment status. Separate instruments, separate from reporting.
Weather
Forecast for St. Louis, MO · fallback
Loading this location's forecast.
Change forecast location
St. Louis is the fallback until you choose a location.
Use my location shares a position rounded to about 1 km with this site and NOAA/NWS for your forecast. No coordinates are saved by this card. When storage is available, it remembers only your choice or city name.
Location options
St. Louis fallback turns off automatic location for this card. Forgetting a choice returns to the browser-permission default on your next visit.
City search stays on this site. City data: GeoNames, CC BY 4.0. NWS forecasts cover the US and territories.
10-day weather view · NWS forecasts up to 7 days
NWS supplies day/night forecasts for about seven days. This ten-calendar-day view shows only supplied periods; later dates and partial days stay unfilled.
Loading this location's daily forecast.
Forecasts can change, especially later periods. These are supplied day/night temperatures, not observed daily extremes. Overnight periods after midnight belong to the preceding day's following night. Day and night precipitation chances remain separate; missing is not zero.
Deterministic approximate geocentric calculations adapted from SunCalc 1.9.0 Sun/Moon geometry. Whole percentages and phase silhouettes are schematic. They are separate from the USNO daily reading above, not uncertain weather forecasts, observations, visibility predictions or exact primary-phase event times. No moonrise/set or visitor coordinates are inferred.
No evidence-backed current assessment has been issued. There is no current analytical risk score to show.
The optional examples illustrate a scoring formula. They do not describe today's world conditions.
Explore illustrative scenarios
Illustrative example · not today's measured risk
elevated strain
Scenario pressure · ordinal points
6
The dependency trap. An existing example setting, not a forecast.
Meaningful agent capability or deployment is running ahead of available controls. Restrict permissions, verify behavior and preserve effective human override.
Inputs and signed contributions
Driver
Setting × weight
Points
AutonomyOpen-ended agents
2 × 2
+4
Deployment reachCritical infrastructure
2 × 2
+4
Independent oversightPartial
1 × -2
-2
Total ordinal points
6
Formula: autonomy ×2 + deployment reach ×2 − oversight ×2. Each input is 0–2. Scores below 3 are more contained, 3–6 elevated strain, and 7–8 critical strain. These are editorial teaching thresholds, not scientifically calibrated weights.
No measured current-world score or score-change history is recorded. Changing the setting explores assumptions; it is not a recent change in world conditions. Headlines do not update this score.
Model reviewed . Research citations below informed the wider scenario discussion; they do not validate these weights or predict catastrophe.
Independent experimental/editorial scoring. These are qualitative scenarios, not probabilities or a predicted apocalypse date. Separate from the Bulletin's symbolic clock; no affiliation is implied.
Evidence-backed assessment · proposed method
How strongly does current evidence support concern about human-caused, globally cascading catastrophe over the next ten years?
Proposed Global Risk Pressure Index, 0–100 editorial points. Scope, ten-year horizon and equal weights are provisional judgments, not scientifically estimated risk shares. This method has not yet been adopted or evaluated against a completed evidence packet.
Proposed domain weights
Domain
Weight
Nuclear
25%
Biological
25%
Climate & ecosystems
25%
Advanced technology
25%
Each domain would score hazard pressure, escalation or spread, and safeguard weakness from 0–4 using fixed, domain-specific evidence anchors. Domain points =25 × the mean of those three scores. Total points = the sum of each domain's points × its disclosed weight.
An issued assessment must show dated evidence, all domain scores and their contributions, the highest-pressure domain, reasons for changes, a low/central/high assessment range, and evidence confidence separately. A range is not a statistical confidence interval. Missing or stale required evidence withholds a new score; it never becomes zero or silently changes weights. A previous assessment, if one exists, keeps its own date and method version.
No current factor scores or assessment history exist here yet. The scenario points are not converted into this proposed index. No model or research run is claimed. A published index would remain distinct from a catastrophe probability, countdown or the Bulletin's official Doomsday Clock.
Kp is an estimate derived from magnetic observations, not a local weather forecast. These readings are separate from reported news and do not establish a trend or common cause.