Historical physical-risk diagnostics

Heatwave Hazard and Portfolio Translation

Rank DFS heatwave events with HPRI-G, inspect the physical signature behind each score, and translate hazard into an explicit portfolio screening model without confusing percentile rank with probability or financial loss.

HPRI method Event diagnostics

Loading the HPRI-G event catalog...

Source: ERA5 daily maximum 2 m temperature
Event method: DFS space-time connected objects
Coverage: 1940-2025
Current mode: HPRI-G / global full-event catalog
Score: 100 ? PercentileRank(HWMS)
Diagnostics: P(HWMM), P(HWMA), P(HWD)
Risk layer: Acute physical hazard screening
Archive release: 2026-01-29

Shared analytical scope

Define the historical hazard screen

Preparing controls...

AR6 region Analysis from Analysis to Minimum HPRI-G Event structure Event lookup Evidence rows

HPRI-GHistorical percentile hazard rank

Each full-event HWMS is ranked against the complete global DFS event catalog. Filtering changes the analytical view, not the score reference distribution.

Reference-domain detail

Filtered events--

Median HPRI-G-Within the filtered event set

Severe tail-HPRI-G ??90

Extreme tail-HPRI-G ??95

Compound events-Two or three P90 traits

Highest selected rank--

Hazard-rank distribution

Filtered occupancy across HPRI-G intervals

Near-uniform occupancy in the unfiltered global catalog is expected because HPRI-G is an empirical percentile rank. Departures after regional, temporal, or structural filtering show where a subset concentrates in the global tail; they do not estimate annual probability or return period.

Selected event

No event selected

HPRI-G-/ 100

Select an event row to inspect its physical signature.

Event figure

AR6 regional diagnostics

Global-rank tail structure by region

Find region Sort by

Region	Type	N	HPRI median	P90	P99	Maximum	Severe ??0	Extreme ??5	Compound	2011-2025

These are global HPRI-G ranks grouped by event AREA. They are useful for catalog screening, but they are not a substitute for HPRI-R or asset-local HPRI-L.

Reproducible event evidence

Highest-ranked events in the selected scope

Rank	Event	Region	Year / DOY	HPRI-G	Band	HWMS	P(HWMM)	P(HWMA)	P(HWD)	Event structure

Select a row to update the event-signature panel. Original DFS metrics remain unchanged in every export.

Transparent risk translation

Portfolio screening workspace

Fictional template 繚 screening only

1Hazard

HPRI with an explicit domain and baseline

100 ? 峉 w_j ? (HPRI_j / 100) ? V_j ? (1 ??AC_j)
Use asset-local HPRI-L for decision-grade applications. Applying the selected HPRI-G to assets is provided only as a preliminary sensitivity screen.

Asset Sector Weight % Hazard basis HPRI Vulnerability Adaptive capacity Asset score Contribution Remove

Covered weight-Sum of entered portfolio weights

Screening score-Contribution on a full-portfolio basis

Exposure-weighted HPRI-Normalized over valid covered weight

Top contribution--

Sector calibration register
Required evidence by sector
Sector template

Model boundary

Inputs are assumptions, not inferred facts

HPRI
Observed or externally supplied acute hazard rank

Vulnerability
Sector/asset calibration supplied by the analyst

Adaptive capacity
Documented resilience assumption supplied by the analyst

Output
Screening score contribution; not currency loss or loss probability

Product readiness
Available calculations and data gaps

Product Status Current basis Required next

HPRI v2.1 framework
Dual-baseline future mode
Method detail

DetectScenario-period P90
Preserve separable future DFS event objects.

Historical calibration register
Financial-loss evidence still to be sourced

Schema readiness is not evidence availability

Historical case Tracker record Required calibration fields Current status

Data and provenance
Reproducible outputs

Original DFS fields are retained beside derived HPRI fields

HPRI method metadataReference modes, bands, sector requirements, and cautions Portfolio CSV templateStatic input schema with fictional example rows Asset GeoJSON templatePoint-overlay schema for future HPRI-L production

Event source
Heatwave/data/list_data.json

Reference domain
-

Baseline
-

Metric basis
-

Interpretation guardrails
What this risk workspace can and cannot establish

HPRI is not a probability
HPRI 100 marks the top of the current empirical catalog; it does not establish a return period or 100% event probability.

Global rank is not local asset risk
Raw HWMS and HPRI-G can reflect event footprint. Cross-market decisions require HPRI-R or preferably HPRI-L.

Screening score is not expected loss
Financial interpretation requires calibrated vulnerability, adaptive capacity, exposure, and loss functions.

HPRI method Event diagnostics Source data Spatial search