📚 DOCUMENTATION OFFICIELLE DU PROJET
🔗 CLIQUEZ ICI POUR ACCÉDER AU RAPPORT EN LIGNE
---
---
# âš¡ Morocco 2030 WC Demand Forecast
### *Forecasting Morocco's Electricity Demand for FIFA World Cup 2030 using SARIMA and Transfer-Learned Neural Event Kernels*
Figure 1 — Morocco Electricity Demand Forecast 2026–2030: SARIMA Baseline + Neural Event Kernel Uplift (85% Prediction Interval)
---
## 📋 Table of Contents
1. Project Overview
2. Research Question
3. Methodology
4. Donor Events
5. Model Architecture
6. Experimental Design
7. Results
8. Morocco 2030 Deployment
9. Repository Structure
10. Installation & Usage
11. Future Work
12. Citation
13. Acknowledgements
14. References
---
## 1. Project Overview
### Motivation
Morocco is co-hosting the **2030 FIFA World Cup** alongside Spain and Portugal, with Morocco's venues potentially hosting a substantial share of the matches. This represents an unprecedented electricity demand planning challenge: the Moroccan grid operator (ONEE) must reliably supply electricity to stadiums, hotels, transportation networks, fan zones, and media infrastructure—many of which are entirely new constructions—for a twelve-month period of elevated, event-driven demand.
Standard forecasting frameworks, including SARIMAX, are designed to extrapolate historical growth trends and seasonal patterns. They do not possess any mechanism to model the **demand shock** that accompanies a mega-event—a systematic, time-localized upward perturbation that begins months before the tournament kickoff (infrastructure commissioning, logistics mobilization), peaks during the event, and decays in the weeks following the final.
### Why Standard Models Are Insufficient
The fundamental challenge is one of **data scarcity combined with structural novelty**:
- Morocco has **never hosted a FIFA World Cup**; its electricity demand series contains no such event signature.
- The demand lift produced by a World Cup is a **rare, high-impact shock** …