convez.ai
PluginDados & AnalyticsAvançado
1988Cameroun

Detecção de Ilhas de Calor Urbano com PyQGIS + Claude AI

1988Cameroun/urban-heat-qgis

Detecta ilhas de calor urbano em satélite + gera insights de planejamento via IA

Ver no GitHub
InicialPluginScore · de 0 a 100

Plugin QGIS que integra dados reais de satélite Landsat 9 com IA Claude para detectar e quantificar ilhas de calor urbano. Processa bandas espectrais (NDVI, LST) e gera relatórios estruturados de planejamento urbano automaticamente, com visualização em heatmap interativo.

Instalação

$npx convez-ai 1988Cameroun/urban-heat-qgis

1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.

Métricas

Estrelas
0
Forks
0
Acompanhando
0
Saúde
28%
Último commit
há 6 meses
Versão
—
Issues abertas
0
Contribuidores
2

Casos de uso

  • Análise térmica urbana com dados geoespaciais reais
  • Relatórios automatizados de planejamento ambiental via IA
  • Visualização 3D de sobreposição térmica em QGIS

Veredito

Projeto bem-estruturado que combina SIG profissional (PyQGIS), processamento espectral real (Landsat 9) e IA generativa para geração de relatórios. Excelente para especialistas em geoanálise e planejamento urbano que querem automatizar pipelines complexos.

README

A tradução para o português deste README ainda está na fila.

3D GIS Flyover Visualization

This project also includes a cinematic 3D terrain flyover demonstrating thermal overlay integration and geospatial animation workflows.

▶ YouTube Demo: https://youtu.be/ku7GaLSLmMM 3D Flyover Preview

🌡️ Urban Heat Island Detection — PyQGIS + Claude AI

Live Demo

Interactive decision-support dashboard:

  • Explore heat zones across Houston
  • Adjust extreme heat thresholds in real-time
  • Visualize NDVI vs temperature relationships Open in Streamlit

**Detects and quantifies urban heat islands using satellite data, generating actionable urban planning insights via an automated geospatial + AI pipeline.**

Stack: Python · PyQGIS · Anthropic API

Before and After

Raw Landsat Inputs

Band 4 (Red)Band 5 (NIR)Band 10 (Thermal)

Houston Urban Heat Island Heatmap

---

What This Does

Landsat 9 Bands (B4, B5, B10)
        │
        ▼
  PyQGIS Pipeline
  ├── NDVI computation     (vegetation health index)
  ├── LST computation      (land surface temp in °C)
  ├── Heat zone classification (4 tiers)
  └── Heatmap visualization (blue→yellow→red ramp)
        │
        ▼
  Claude AI (claude-sonnet)
  └── Generates structured urban planning report in Markdown
        │
        ▼
  Output: styled QGIS map + UHI\_Report\_Houston\_TX.md

---

Quick Start

1. Clone

git clone https://github.com/1988Cameroun/urban-heat-qgis.git
cd urban-heat-qgis

2. Install dependencies

pip install anthropic numpy
# PyQGIS comes bundled with QGIS Desktop — install from qgis.org

3. Set your API key

export ANTHROPIC\_API\_KEY=sk-ant-...

4a. Run Demo (no Landsat data needed)

python scripts/demo\_no\_qgis.py

This generates a full AI report using realistic Houston baseline statistics.

4b. Run Full Pipeline (with real Landsat data)

Download Landsat data:

  1. Go to earthexplorer.usgs.gov
  2. Draw AOI over Houston
  3. Select Landsat 9 OLI/TIRS C2 L2
  4. Download Bands 4, 5, and 10

Place files:

data/
├── LC09\_B4.TIF   ← Red band
├── LC09\_B5.TIF   ← NIR band
└── LC09\_B10.TIF  ← Thermal band

Run inside QGIS Python Console:

exec(open('scripts/heat\_analysis.py').read())
run\_pipeline()

Or standalone:

export QGIS\_PREFIX\_PATH=/usr   # Linux; adjust for your OS
python scripts/heat\_analysis.py

---

Output Example

# Urban Heat Island Analysis Report
\*\*City:\*\* Houston, TX

## Executive Summary
Houston's summer thermal profile reveals a pronounced urban heat island
effect, with land surface temperatures ranging from 28.4°C to 58.7°C
and a city-wide mean of 41.2°C...

## Heat Zone Analysis
| Zone          | Threshold | \~Area Coverage |
|---------------|-----------|----------------|
| Extreme Heat  | ≥ 49.9°C  | 6.7%           |
| High Heat     | ≥ 44.1°C  | 30.9%          |
| Moderate      | ≥ 41.2°C  | 50.0%          |
| Cool/Vegetated| < 38.3°C  | 30.9%          |

## Mitigation Recommendations
1. Expand tree canopy in Loop 610 industrial corridor...
2. Mandate cool roofs for new commercial construction...
...

---

Project Structure

urban-heat-qgis/
├── scripts/
│   ├── heat\_analysis.py     # Full PyQGIS pipeline
│   └── demo\_no\_qgis.py      # Demo (no QGIS needed)
├── data/                    # Place Landsat TIF bands here
├── outputs/                 # Generated TIFs + MD reports
├── reports/                 # Sample AI-generated reports
├── requirements.txt
└── README.md

---

Why This Stands Out

Most QGIS projects stop at the map. This pipeline goes further:

  • PyQGIS scripting — fully automated, reproducible, no GUI clicks
  • AI interpretation layer — Claude converts raw raster stats into actionable planning language
  • Real-world data — uses actual Landsat Collection 2 radiometric scaling
  • Reusable — swap city name + data path to analyze any metro area

---

Tech Stack

LayerTechnology
Spatial analysisPyQGIS 3.x, QGIS Raster Calculator
Data sourceUSGS Landsat 9 OLI/TIRS C2 L2
IndicesNDVI, LST (Kelvin→Celsius)
AI reportAnthropic Claude (claude-sonnet)
LanguagePython 3.10+

---

License

MIT

Ver o original no GitHub

Bom para

Relacionados