GFT & Google Earth Engine

GFT invited by Google as partners to be part of the Sustainability Hackathon using Google Earth Engine.

Google have gone into sustainability in a big way from setting a target to be carbon neutral in all data centres by 2030 to providing tools to customers to view Carbon Foot metrics for services they consume. To keep in the vein of sustainability they have now released a service Google Earth Engine for commercial use. As a Google partner this gave us an opportunity for Team “GFT Green Coding” to get hands-on with Google Earth Engine for use-cases like sustainable value chains, climate risk and resilience, digital sustainability and ESG process transformation. To get a feel on how to use Google Earth Engine and apply it to a use-case we chose to look at pollution levels around schools across London using the data from Google Earth Engine in a particular time frame. An initial model required little effort from a technical perspective and most of the effort was around understanding how to work with Google Earth Engine and consolidate multiple different “layers” into a usable view.

In our case Google Maps, Government Schools data and “LandSat” data. See below for output and code snippet.

What do all the dots tell us?

The analysis revealed in central London (darker Green) pollution was higher and in the west where Heathrow (Darker Green) is situated. The hypothesis being these areas were more polluted due to Vehicle or Airplane Pollution.  This would form the basis of the further analysis to see if that can be correlated to poor health and performance in School children in those areas. The resulting analysis can be used to form strategies and solutions to tackle these problems. Using Google Earth Engine and additional data sources you can continually monitor to see if these strategies are affective over time as data is continually being added by Google. The key benefit, more importantly,  allows solutions to be implemented to improve the school children’s health and performance.

Java Code Snippet to produce the (n.b Schools data needs to be import into GGE for below to work) :

// Load air quality data from a sensor in London
var collection = ee.ImageCollection(“COPERNICUS/S5P/NRTI/L3_NO2”)
  .filterBounds(ee.Geometry.Point(-0.1278, 51.5074))
  .select(‘tropospheric_NO2_column_number_density’);

// Define a date range to filter the data
var startDate = ‘2021-01-01’;
var endDate = ‘2022-12-31’;
var filteredCollection = collection.filterDate(startDate, endDate);

// Get the mean value of NO2 for each image in the collection
var no2Mean = filteredCollection.mean();

// Display the NO2 level in the London area
var visParams = {
  min: 0,
  max: 0.0002,
  palette: [‘black’, ‘blue’, ‘purple’, ‘cyan’, ‘green’, ‘yellow’, ‘red’] };
Map.addLayer(no2Mean, visParams, ‘Mean NO2 Level’);
var pollutiontable = no2Mean.reduceRegions(table, ee.Reducer.mean(), 30)
print(pollutiontable)

// Add the pollutionBySchool table as a map layer
Map.addLayer(pollutiontable, {min: 0, max: 0.0005, palette: [‘white’, ‘yellow’, ‘red’]}, ‘Pollution by School’);


So, what is Google Earth Engine? 

It is geospatial data and analytical platform combining more than 40 years of historical imagery and scientific data sets ready to use for analysis for sustainability:

  • Data Catalog – which has more than 40 years of historical imagery and scientific data sets which are curated and includes near-real-time satellite imagery. 
  • Computation Platform – A tool with a front end which enables you to analyze and visualise the Earth data at scale by providing parallel processing speed with machine learning built in
  • 50K Monthly Active Users (MAU) ( academics, NGOs, researchers) – Who use the data sets to publish new data sets, these users build their own data products, maps, create models and techniques used to solve problem which are put back into Google Earth Engine.

The key aspect of the Data Catalog is that Google has consolidated with the MAU’s expertise into a useable dataset for analysis, and here’s the good bit, it has been validated and cleansed ready to use.

And now for the real techie bit…

  1. Google Cloud Account
  2. Google Earth Engine Service
  3. Client libraries in JavaScript & Python
  4. Code Editor: Easy interactive experimentation with one-click collaboration
  5. Earth Engine Apps: Wire up and share custom interactive dashboards
  6. One Platform API for direct integration via HTTP or Stubby
  7. 80+ PB and growing of data

 The business angle…

The use of Google Earth Engine is still in the early days as it has been on general availability from a commercial perspective only recently. New ventures and existing businesses are just starting to understand the use and potential.

Several examples of use cases for Google Earth Engine  

  • Sustainable Sourcing – Enable global supply chain transparency and traceability to footprint.
  • Agricultural Technology – Precision Agriculture, Increase Yield, improved visibility of the Food Supply Chain
  • Environmental Impact – Monitor and act (ex. Methane Detection)

To use Google Earth Engine within the financial industry it can potentially help Financial Firms meet targets on sustainability given by existing or new legislation. An example being for Climate Risk, financial services companies (banks, insurance, funds) are increasingly being put under scrutiny and regulations. They need to mitigate climate risk for the companies they invest in, which will potentially require solutions using Google Earth Engine to help meet the current and future legislation.

Find out more on UK Finance & UNEP here.

Conclusion

The use of Google Earth Engine is going to continue to grow, and a lot of our competitors have interest in creating  in house expertise on using Google Earth Engine to support the ESG framework (here) or partnering with companies who have produced solutions using Google Earth Engine to create more tailored datasets for the customers. GFT can have opportunities by engaging with existing client base to create a POC with clients or accelerators to showcase to clients. A couple of uses-cases.  

  • Risk Modelling for Insurance using satellite ( An example of a  company has  provided a solution and was the supporting google patterner at the Hackathon io)
  • Using Google Earth Engine to develop a solution to enable sustainable lending 

 

Find out more about our partnership with Google here

Hybrid and multicloud

Learn how cloud and multicloud drive transformation!

Download now