Introduction to the course

Welcome!

This course contains lessons and material about AI, Geospatial Data Analysis, and Urban Transformation Research. You will gather hands-on experience in topics such as Urban Growth Modeling, Flood Detection, and Air Quality Analysis using Python and Jupyter Notebooks. The curriculum emphasises combining data from diverse sources, including Earth Observation, Volunteered Geographic Information (VGI), and Social Media Geographic Information (SMGI). By mastering modern geospatial tools and methods, you will be prepared to tackle interdisciplinary challenges in sustainable development and Earth System Science.

Please make sure, to always check provided python-scripts for correct file-paths. Either change the path or the file locations of here downloaded files so that the scripts work as intended. Use the hands-on videos provided by this course to guide you through each exercise.

The materials were created by the NFDI4Earth project and can be accessed openly via openRUB. The content is licensed as CC by-NC attribution.

Course Contents

Part 1: The future is urban, the data is smart – Analysis of transformation processes with VGI, SMGI, and EO data. 

1. Python and EO data
   • Filter and select feature and image collections in Google Earth Engine (GEE)
   • Simple analysis metrics in GEE
   • Detect floods with Sentinel-1 data in GEE
   • Creating timelapses from Landsat satellite data
   • Calculate nighttime light trends in GEE
   • Measuring air quality with Sentinel-5p in GEE
     
     
2. Python and VGI
   • Utilising ohsome API for building development and extraction
   • Combining osm amenities and GEE population data
   • Utilising osmnx API for road network analyses
     
     
3. Python and SMGI
   • Filter Flickr image collection with tags, geotags, and free text
     
     

Part 2: Artificial Intelligence: Basics and Geographic Applications.

1. Introduction
   • Introduction to python and relevant libraries (e.g. pandas, NumPy, matplotlib)
   • Introduction to machine learning and artificial intelligence in Geography
     
     
2. Machine Learning
   • Scripting and utilisation of Random Forest
   • Random Forest for classification and regression
   • Scripting and utilisation of Suppport Vector Machines
   • Suppport Vector Machines for classification and regression
     
     
3. Neural Networks
   • Basics of neural networks, bias, activation functions, feed forward, gradients, loss, backpropagation
   • Introduction to convolutional neural networks and recurrent neural networks
     
     
4. Geographical Applications
   • Ship detection with CNN
   • Weather Prediction with recurrent neural networks
   • Prediction of water levels
   • Estimation of soil sealing