ICIMOD Conducts First Training on Google Earth Engine (GEE) in Nepal

07 Jun 2017

Thirty-eight participants—13 of them women—from Nepal and Kenya, convened for a four-day training workshop on the Google Earth Engine (GEE) platform in Kathmandu, Nepal. The participants were selected from an open call for applications that received over 250 applications from professionals and students interested in learning more about GEE.

The training was organised by the International Centre for Integrated Mountain Development (ICIMOD) in partnership with Google Earth Engine Outreach, under its SERVIR Hindu Kush Himalaya (SERVIR-HKH) initiative. 

At the opening session of the workshop, Birendra Bajracharya, Regional Programme Manager MENRIS and Chief of Party, SERVIR-HKH said, “Technology has changed the way we work with remote sensing and geographic information science. Back in 2000, we had to buy Landsat imagery at a premium, and processing satellite imagery would take at least a month. Landsat and many other satellite imagery are now available for free. The Google Earth Engine hosts and organizes petabytes of geospatial data including Landsat, on their servers and offers a powerful computing environment for data processing and producing outputs very quickly. This has made the transition from a desktop environment to a cloud based ecosystem seamless.”

Nicholas Clinton, Google Earth Engine Outreach, at the workshop  Photo: ICIMOD

Nicholas Clinton, Google Earth Engine Outreach, at the workshop

In his introductory lecture, Nicholas Clinton with Google Earth Engine Outreach, provided participants an overview of how earth observation technology and Google Earth Engine in particular, is being used by scientists, governments and non-profits to quantify changes in the environment. With real world examples, he showcased how the platform was also incrementally aiding the development of science across research in the environment. Nicholas led the course, using a combination of lectures and hands-on exercises, followed by a group project.

Over the course of the training, participants got an overview of the myriad applications that can be developed using the GEE platform and Google’s public data catalogue that holds over five petabytes of data. They learned how to write scripts using GEE’s JavaScript Application Programme Interfaces (APIs), to carry out digital image processing and geospatial analysis, ranging from basic to intermediate levels (in terms of the complexity of tasks), using various kinds of datasets—MODISLandSat, and Sentinel, among others. 

In addition to the introduction and usage of GEE, the training also covered a session on mapping gender and socioeconomic issues. Kamala Gurung, a gender specialist at ICIMOD, presented a session on how gender studies generate evidence based knowledge which can be integrated with geospatial knowledge while developing early warning systems, agricultural advisory services, and land cover and land use change products. Geographically referenced gender disaggregated data can be a strategic variable to revealing socioeconomic as well as ecological issues in depth.

The participants identified ten different projects and worked in teams with resource persons from Google and ICIMOD to fine tune their applications further. They presented an exciting array of applications within a short period of time.

A team developed a prototype to map glaciers using time series LANDSAT data, while another attempted a flood mapping tool in the GEE platform. Other prototypes ranged from applications that displayed district information, mapped urbanised areas, and visualized the urban heat island effect in cities.

Steve Omondi Otieno and Simon Teddy Mwangi Kuria, from SERVIR Eastern & Southern Africa hub with their team members Photo: ICIMOD

Steve Omondi Otieno and Simon Teddy Mwangi Kuria, from SERVIR Eastern & Southern Africa hub with their team members

Steve Omondi Otieno and Simon Teddy Mwangi Kuria, from the Regional Centre for Mapping of Resources for Development (RCMRD)SERVIR’s Eastern & Southern Africa hub, were part of a team that developed a prototype for land use land cover classification and vegetation index. They attended the training as part of the exchange agreement between SERVIR hubs.

A training workshop on GEE in Bangladesh is planned for August 2017.

What is GEE?

Google’s GEE platform can be used for large and small-scale scientific analysis and visualization of geospatial datasets. It stores, organizes, and provides access to a wide variety of satellite imageries and geospatial datasets, and offers global-scale environmental data analysis capabilities. In addition to the tools and cloud computational powers necessary to analyze large datasets, the GEE platform also offers Application Programme Interfaces (APIs) in JavaScript and Python. GEE is widely used by researchers, non-profit organizations, educators, and governmental agencies to analyze large-scale geospatial data, and is available free of cost for non-commercial users by signing up here.