29 Mapping Biodiversity Through Google Earth Engine
Practical Module | International Summer Course 2026
This module introduces participants to satellite-based vegetation analysis using Google Earth Engine (GEE), complementing the drone-based multispectral survey conducted at the Faculty of Biology, Universitas Gadjah Mada. Participants learn how cloud-based geospatial analysis can extend the ecological information obtained from local drone surveys to broader regional and landscape scales.
29.1 Learning Objectives
By the end of this module, participants should be able to:
- Understand the role of Google Earth Engine in environmental analysis.
- Access and visualise satellite imagery through the GEE Code Editor.
- Calculate vegetation indices using Sentinel-2 satellite data.
- Compare drone-derived and satellite-derived ecological information.
- Interpret environmental patterns across multiple spatial scales.
- Understand the strengths and limitations of both drone and satellite remote sensing approaches in biodiversity and conservation studies.
29.2 Study Context
The Faculty of Biology forest at Universitas Gadjah Mada is located within the densely developed urban environment of Yogyakarta. It is surrounded by roads, academic buildings, residential areas, and the broader urban landscape of one of Java’s largest cities. Although the drone survey provides highly detailed information on canopy structure, the satellite analysis conducted in this module reveals how this forest patch relates to the surrounding urban green space network and broader landscape-scale vegetation patterns.
29.3 Why Combine Drone Mapping and GEE?
Drone mapping and satellite remote sensing provide complementary ecological information.
Drone surveys provide extremely high spatial resolution and detailed local observations, allowing participants to observe canopy structure, understory vegetation, habitat boundaries, and fine-scale ecological variation. However, drone surveys are generally limited by flight duration, coverage area, operational complexity, and data acquisition frequency.
Satellite imagery provides broader spatial coverage and repeated observations over long time periods. Through Google Earth Engine, participants can access historical and near-real-time environmental datasets that support large-scale monitoring and temporal analysis.
By combining both approaches, participants can connect:
- Detailed local ecological observations from drones,
- With broader regional environmental patterns observed from satellites.
This integration supports more comprehensive biodiversity assessment and conservation planning.
29.4 Introduction to Google Earth Engine
Google Earth Engine is a cloud-based geospatial analysis platform developed for large-scale environmental data processing. GEE allows users to access global remote sensing datasets without downloading massive amounts of data locally.
The platform provides access to:
- Sentinel-2 imagery,
- Landsat imagery,
- Digital elevation datasets,
- Land cover products,
- Climate data,
- Forest change datasets,
- And many other environmental datasets.
Because processing occurs in the cloud, GEE can analyse large geographic regions efficiently while supporting interactive visualisation and scripting.
Participants in this exercise will use Sentinel-2 multispectral imagery, NDVI analysis, and regional vegetation assessment around UGM.
29.4.1 Recommended Platform
Participants may access:
29.5 Exercise: Satellite NDVI Analysis
This exercise introduces participants to vegetation analysis using Sentinel-2 satellite imagery processed within Google Earth Engine. Participants will calculate NDVI and compare the resulting vegetation patterns with drone-derived multispectral products from the Faculty of Biology forest.
The exercise demonstrates how spatial resolution influences ecological interpretation and how satellite imagery provides broader environmental context beyond local drone surveys.
29.5.1 Step 1 – Access Google Earth Engine
Participants create a free Google Earth Engine account and access the GEE Code Editor environment.
Participants should familiarise themselves with:
- Map display,
- Dataset catalog,
- Layer visualisation,
- And script execution tools.
29.5.2 Step 2 – Load Sentinel-2 Dataset
Participants load Sentinel-2 multispectral imagery covering:
- Universitas Gadjah Mada,
- The Faculty of Biology forest,
- And surrounding urban vegetation.
Participants should select imagery with:
- Low cloud cover,
- Recent acquisition date,
- And suitable environmental conditions.
29.5.3 Step 3 – Calculate NDVI
Participants calculate NDVI using the near-infrared and red spectral bands from Sentinel-2 imagery.
Participants visualise the resulting NDVI map using colour gradients representing vegetation condition and density.
Higher NDVI values generally indicate healthier or denser vegetation, while lower values may indicate sparse vegetation, built environments, or stressed ecological conditions.
29.6 Comparing Drone and Satellite Observations
Participants compare:
- Drone-derived multispectral NDVI,
- Sentinel-2 NDVI,
- And ecological interpretation at different spatial scales.
Drone imagery provides highly detailed local observations capable of revealing:
- Canopy gaps,
- Understory vegetation,
- Small habitat patches,
- And individual tree structure.
Satellite imagery provides broader regional context showing:
- Urban vegetation distribution,
- Ecosystem fragmentation,
- Ecological corridors,
- And landscape-scale vegetation patterns.
Participants should evaluate how differences in spatial resolution influence ecological interpretation and conservation analysis.
29.7 Ecological Interpretation and Policy Relevance
The final stage of the exercise focuses on ecological interpretation and environmental decision-making. Participants analyse how the Faculty of Biology forest functions within the broader urban ecosystem and discuss how remote sensing products may support biodiversity conservation and urban environmental management.
Potential discussion topics include:
- Urban green-space connectivity,
- Habitat fragmentation,
- Ecosystem resilience,
- Vegetation stress,
- Restoration priorities,
- And biodiversity conservation within urban environments.
Participants may evaluate whether the Faculty of Biology forest acts as:
- An ecological refuge,
- A biodiversity corridor,
- Or an isolated habitat patch within the surrounding urban landscape.
The exercise demonstrates how drone mapping and satellite analysis can support evidence-based environmental planning and the development of conservation policies.
29.8 Discussion Questions
Participants may discuss:
- How connected is the campus forest to surrounding vegetation?
- Are there signs of ecosystem fragmentation?
- Which areas appear ecologically stressed?
- How might future urban development affect biodiversity?
- Which areas should be prioritised for ecological protection or restoration?
29.9 Summary
This exercise introduces participants to multiscale biodiversity analysis by integrating drone mapping with Google Earth Engine. By comparing high-resolution drone-derived observations with regional satellite imagery, participants learn how local ecological conditions relate to broader landscape-scale environmental processes.
The exercise also demonstrates how cloud-based geospatial analysis supports environmental monitoring, ecosystem assessment, and evidence-based conservation decision-making.
29.10 Full Workshop Module
The complete practical module document is embedded below for reference and offline reading.