Introduction to Spatial Data Science -Specialization in Google Earth Engine (GEE)

Spatial data science (SDS) is a subset of Data Science that focuses on the unique characteristics of spatial data, moving beyond simply looking at where things happen to understand why they happen there. Essentially, Spatial data science treats location, distance & spatial interactions as core aspects of the data using specialized methods and software to analyze, visualize and apply learnings to spatial use cases. Having skills in spatial data science helps to enhances performance, ease weighted decisions and allows more accurate analysis. It inspires a greater spectrum of applications because of data accuracy, better precision leading to increased productivity. Additionally, geospatial technologies enhance the performance of AI in the industry. The development of geospatial technologies initiates a corresponding upgrade of related industries. Improvement in process is not likely to stop, thus ensuring even greater precision, credibility, performance, quality, and security.

Google Earth Engine Introduction

Google Earth Engine (GEE) is a free Google cloud platform used by scientists, researchers and developers to create solutions for sustainability. It is useful in detecting changes, quantifying differences on the earths’ surface. It contains multi-petabyte catalog of satellite imagery and geospatial datasets. It is useful in academic research, non-profit, business and government. The platform hosts real-time and historical public data spanning more than 40 years back and was started in 2010 by Google’s non-profit division. Google’s main objective to establish GEE was to:

• To Address the Imagery computational power challenge.
• To overcome the processing time – it’s uses Google Cloud GPUs.

Google Earth Engine (GEE) supports listed APIs and integrations

1. JavaScript API
2. Python API
3. Climate Engine
4. Bigquery
5. Google Maps

Google Earth Engine Components

1. Cloud platform
2. Cloud Integrated Development Environment (IDE) Code Editor
3. Google cloud compiler integrations
4. Data Catalogue hosted in Google servers
5. Interactive User Interface

Ways to import data

1. Reading from the GEE Data Catalogue
2. Import manually in the assets

Types of datasets used in Spatial Analysis

1. ShapeFile
2. GeoTiff
3. Excel
4. Raster data

Industrial Applications of GEE

1. Urban planning
2. Agriculture mapping
3. Forest degradation monitoring
4. Land use land cover classifications
5. Disease mitigation
6. Disaster response – droughts/floods
7. Water resource mapping
8. To infinity…

Approaching Spatial Data Science project with GEE

1. Choose a challenge and analyze it extensively
2. Pick dataset – Understand the most appropriate data by considering . Application domain – is it Agriculture, Forest Mapping or.. . Time aspect – Do you want Daily, of Monthly time updates
3. Find and Area of Interest (AoI) - Case Study
4. Data Preprocessing - Date ranges and filter clouds
5. Clip area
6. All Geospatial Analysis here
7. Visualizations
8. Export – to use with Python Colab or validation with in-situ data

Challenges while Working on GEE Projects

• Spatial-temporal dependence estimation is computationally expensive- data querying and analysis.
  • Limited open source contribution
  • Expensive learning resources
  • Data mining techniques, (where, how and when)
  • Scarce Spatial analysis skills

Comments

[WanjohiChristopher]

Nice on bro