Research Dashboard
Energy Security Foresight & Partner Intelligence
This dashboard combines ESS-based scenario analysis, Eurostat-derived historical indicators, and a source-aware
energy-partner module designed to preserve methodological coherence while keeping the visual workflow quick to scan.
Formula definitions are centralized in Method & Formula Atlas.
ESS4 Directional Security Index
ESS4 applies the paper's composite rule to the EU-adapted indicators: the ESS1 oil/gas exposure ensemble,
ESS2 industrial energy intensity, and ESS3 domestic production ratio are averaged into one composite trajectory.
Because component directions differ, use ESS4 as a scenario trajectory to inspect with its components rather than
as a standalone ranking.
Scenario Indicator Panels
The scenario panels below prioritize analytical clarity: consistent axes, lighter backgrounds, and high-contrast traces
to keep interpretation fast during research and presentation.
Forecast shock overlay:
S1 | Red Sea shipping disruption: 2025-2026 | maritime_trade | medium | included_source_backed_manual
S2 | Russian gas phaseout and LNG substitution pressure: 2025-2027 | gas_supply | medium | included_source_backed_manual
S3 | Accelerated EU renewables package: 2026-2030 | renewables_policy | medium | included_source_backed_manual
S4 | Industrial decarbonisation funding: 2026-2030 | industrial_efficiency | medium | included_source_backed_manual
S5 | Mediterranean corridor infrastructure agreement: 2026-2030 | domestic_resilience | low | included_source_backed_manual
S6 | Digital corridor investments: 2026-2030 | digital_infrastructure | low | included_source_backed_manual
Forecast Quality Backtest
This section tests the forecasting engine against history. For each cutoff year, the model uses only the data available
up to that point, projects the next years with the same baseline formula logic, and compares those predictions with the
Eurostat observations that actually followed.
Backtest window
1995-2024
Rolling forecasts
800 metric-year tests
1-year MAPE
4.01%
1-year direction accuracy
61.93%
Backtest source: eurostat api extended aggregate proxy. This validation layer uses a separate extended EU27 Eurostat aggregate-proxy history, not the production ESS workbook merge. Each prediction is produced using nested model selection and only observations available up to the cutoff year shown in the hover text.
Selected Baseline Models
Each baseline metric is forecast with the candidate model that performed best in the rolling historical backtest before scenario and shock overlays are applied.
| Metric | Selected baseline | Backtest MAPE | Direction accuracy |
|---|
| ESS1 Oil Exposure Channel | Long linear trend | 2.76% | 100.00% |
| ESS1 Gas Exposure Channel | Recent linear trend | 13.19% | 66.67% |
| ESS2 Industrial Energy Intensity | Recent weighted annual change | 6.98% | 98.00% |
| ESS3 Domestic Production Ratio | Recent weighted annual change | 3.38% | 58.00% |
| Total Energy Consumption (ktoe) | Last observed value | 2.95% | 0.00% |
| Renewable Energy Consumption (ktoe) | Average annual change | 6.17% | 89.00% |
| Renewable Energy Production (ktoe) | Recent weighted annual change | 8.18% | 89.00% |
| Oil Imports (thousand tonnes) | Last observed value | 4.81% | 0.00% |
| Gas Imports (ktoe) | Last observed value | 5.76% | 0.00% |
EU Main Energy Partners
This source-aware section starts from raw Comext replication for publication-grade product-partner evidence,
then keeps the official annual API and workbook routes as reproducible comparison layers with different scopes.
Partner Data Reliability Summary
Source hierarchy for this project: Raw Comext as the preferred scholarly replication layer
when product-partner evidence is the focus, Official API as the reproducible annual quantity route,
and Workbook as a published supplement.
Partner validation: 36 pass / 0 blocking fail / 9 not comparable. Metadata reports 0 failed checks. Workbook supplement passed blocking checks; Q1 article rows are kept as diagnostic context.
| Source |
Available |
Reliability |
Reproducibility |
Coverage |
| workbook | True | medium | published_workbook | 2024-2025 shares |
| official_api | True | higher | official_api | 2024 annual quantity |
| raw_comext | True | highest_when_populated | raw_comext | depends_on_local_raw_extract |
Preferred scholarly source: Raw Comext replication. Use this block for publication-facing product-partner evidence when the selected Comext product scope matches the claim being made. Unmapped aggregate or special Comext partner codes are kept as source codes rather than relabeled by inference.
Comparison source: Official Eurostat annual API. Use this block as the reproducible annual quantity route when raw Comext value/mass replication is not the focus.
Supplementary source: Workbook-based published figures. Keep this as a communication and comparison layer rather than the default empirical backbone.
Interpretation: This comparison uses 2024 quantity shares only. Official API bars represent the reproducible annual quantity route, while workbook bars remain a published comparison layer. Gaseous natural gas is omitted here because the current official API route in the project does not yet isolate the same published workbook definition one-to-one.