---
language:
- en
license: apache-2.0
size_categories:
- 1K<n<10K
task_categories:
- text-generation
- visual-question-answering
- time-series-forecasting
- text-classification
tags:
- climate-change
- natural-disasters
- causal-reasoning
- multimodal
- earth-observation
- sentinel
- open-meteo
- em-dat
---

# DisasterChain: a multimodal global disaster dataset linking Earth observation, meteorology and structured narratives

## Dataset Description

- **Homepage:** [http://disasterchain.icar.cnr.it](http://disasterchain.icar.cnr.it)
- **Paper:** [Coming Soon...]
- **Repository:** https://github.com/Franco7Scala/DisasterChain
- **Point of Contact:** [Francesco Scala](francesco.scala@icar.cnr.it)

### Dataset Summary

**DisasterChain** is a high-fidelity, multimodal dataset designed to advance research in causal reasoning, environmental monitoring, and disaster impact assessment. Building upon the EM-DAT database, it provides a comprehensive reconstruction of over 1,400 global extreme environmental events from 2014 onwards. 

Unlike traditional disaster catalogs that only provide static metadata, this dataset introduces **Grounded Causal Impact Chains**. Using advanced Large Language Models (Qwen 2.5 72B and Llama 3 70B), we extract step-by-step causal sequences of physical and socio-economic impacts (e.g., *Extreme Precipitation -> River Overflow -> Flood -> Infrastructure Damage -> Displacement*). Crucially, every step in the causal chain is strictly grounded to real-world news reports via verified supporting quotes.

Furthermore, each event is enriched with extensive multimodal data, including 21-day weather time series and multispectral satellite imagery availability, making it a unique resource for bridging Natural Language Processing (NLP) and Earth Observation (EO).

### Supported Tasks and Applications

- **Causal Reasoning & Event Extraction:** Training models to identify cascading effects in unstructured text;
- **Multimodal Disaster Assessment:** Combining satellite imagery (Sentinel-1, Sentinel-2), weather data, and text to predict physical damage or human casualties;
- **Climate Impact Analysis:** Analyzing the socio-economic downstream impacts of specific weather triggers across different global regions.

## Dataset Structure

The dataset is provided in `JSON` format. The core JSON structure contains the following nested objects for each disaster event:

### 1. Root Metadata & Impact Data
Standardized identifiers, geographical coordinates, dates, and quantitative human impact metrics (deaths, affected population, financial damage) sourced from EM-DAT.

### 2. `weather_data`
Retrieved via the Open-Meteo API, containing:
- Pre-event and post-event statistical summaries (mean rainfall, snowfall, min/max temperatures);
- A **21-day daily time series** capturing the meteorological evolution before, during, and after the event.

### 3. `satellite_data`
Sourced via Copernicus Data Space Ecosystem (CDSE) / Sentinel Hub:
- Precise bounding boxes (`bbox`) for the affected areas;
- Temporal availability arrays for Sentinel-1 (SAR), Sentinel-2 (Optical/False Color), and Sentinel-3 (Thermal);
- Integration with ESA WorldCover land use data.

### 4. `news_data`
Metadata regarding the information retrieval process:
- Sources queried (Google News, ReliefWeb, IFRC, Wikipedia);
- Relevance scores and filtering penalties ensuring only high-quality, event-specific journalistic text is retained.

### 5. `summary` & `causal_chain` (The LLM Reasoning Layer)
- **Summary:** A concise, factual narrative of the event generated by Llama 3 70B.
- **Causal Chain:** Generated by Qwen 2.5 72B-Instruct. A normalized JSON array containing the chronological steps of the disaster. Each step includes:
  - `type_event`: A semantically normalized category (e.g., *Flood*, *Casualties*, *Infrastructure Damage*);
  - `description`: The specific manifestation of the impact;
  - `supporting_quote`: An exact or fuzzy-matched quote from the retrieved news context, ensuring a **zero-hallucination rate** in the causal extraction;
  - `token_usage`: Computational footprint tracking.

## Methodology and Data Pipeline

The construction of this dataset followed a rigorous, multi-stage pipeline designed to eliminate LLM hallucinations and ensure physical consistency:

1. **Event Anchoring:** Sourcing base events from EM-DAT (post-2014) and geocoding locations via Nominatim;
2. **Multimodal Retrieval:** Automated fetching of historical weather data and satellite imagery footprints;
3. **Textual Context Gathering:** Scraping, parsing, and algorithmically scoring news articles for relevance;
4. **Causal Extraction:** Using Qwen-72B to extract causal graphs. A strict Python-based fuzzy-matching algorithm (threshold ≥ 0.80) was applied to cross-reference every LLM-generated quote against the raw news text. Any unverified causal step was systematically dropped;
5. **Semantic Normalization:** Over 500 raw event types were algorithmically normalized into a clean taxonomy of ~390 standardized categories.

## Bias, Risks, and Limitations

- **Media Coverage Bias:** The richness of the causal chains and the availability of news data are inherently biased toward events that received significant international or English-language media coverage. Events in remote areas may have shorter causal chains.
- **Satellite Data Availability:** Cloud cover heavily impacts the usability of Sentinel-2 optical imagery during severe weather events (e.g., hurricanes). Sentinel-1 SAR is provided to mitigate this.

## Citation

If you use this dataset in your research, please cite our paper:

```bibtex
Coming soon...