Dynamic LULC (Land Use Land Cover) Application: Google Earth Engine

Land Use and Land Cover (LULC) classification plays a pivotal role in environmental research, urban planning, agriculture, and natural resource management. Understanding how land is utilized and its associated cover types provides valuable insights into ecological processes, climate change, and sustainable development. In this blog, we delve into the significance of LULC, explore the capabilities of an innovative application, and introduce the key datasets driving this field.

Why LULC Matters for Research?

1. Environmental Monitoring: LULC data enables us to monitor changes in land patterns over time. Whether it’s deforestation, urban expansion, or wetland loss, researchers can track alterations and assess their impact on ecosystems.

2. Biodiversity Conservation: Identifying land cover types helps protect critical habitats. By pinpointing areas with high biodiversity, conservation efforts can be targeted effectively.

3. Climate Modeling: LULC influences local and global climate. Accurate classification aids climate models by accounting for land surface properties such as albedo, vegetation, and water availability.

4. Disaster Management: During natural disasters like floods or wildfires, LULC data assists emergency responders in assessing affected areas and planning relief efforts.

The Application: Simplifying LULC Image Downloads

Our newly developed application harnesses the power of Google Earth Engine, Sentinel-2 imagery, and the Dynamic World dataset. Here’s how it benefits researchers:

1. Easy Access: Researchers can effortlessly download pre-classified LULC images for specific regions. No complex processing is required—just select your area of interest and retrieve the data.

2. Time Efficiency: Instead of spending hours on manual classification, our application streamlines the process. Researchers can focus on analyzing the data rather than wrangling with algorithms.

3. Customization: Users can tailor their LULC queries based on their time range of interest and custom area of interest. This flexibility enhances research precision.

Sentinel-2 and Dynamic World Dataset

1. Sentinel-2: Developed by the European Space Agency (ESA), Sentinel-2 is a constellation of Earth observation satellites. It captures high-resolution multispectral imagery, making it ideal for LULC classification. With its frequent revisit times, Sentinel-2 provides up-to-date information on land dynamics.

2. Dynamic World Dataset: This comprehensive dataset integrates the Sentinel-2 dataset. As the satellite has a spatial resolution of 10m and a temporal resolution of 5 days, it is most suitable for land use land cover analysis. The dynamic dataset covers the entire globe and offers consistent land cover information. The dataset is prepared by using deep learning which is cutting-edge technology in remote sensing. Researchers can tap into its rich archives for historical analyses (Click here for more information).

In summary, our application bridges the gap between LULC research and practical implementation. By leveraging Sentinel-2 and the Dynamic World dataset, we empower researchers to explore land cover dynamics efficiently. Let’s unlock the potential of LULC data for a sustainable future! 🌍🔍

User interface

The application Dynamic LULC have a very user-friendly interface.

Dynamic LULC: GEE app interface

This is the main dashboard of the application. Here (1) is the command manu and (2) is the result manu. User can use different parameters for their preference.

1. Study area

Select the area of interest. For demo purposes, we have only used three areas. Rangpur, Dhaka and Sylhet. The dropdown box will give you options to select your preferred location

2. Select filters 

After selecting the study area, you have to edit your time preferences. Now one thing to make sure that, the dynamic world dataset filters satellite images greater than 35% and removes them from the main image collection. The cloud-contaminated part is masked from the images to increase the model performance. April to September is the monsoon period in Bangladesh so selecting the date range will provide some masked images. October to February is the dry season in Bangladesh so it is preferable to use this period to work with optical satellite images.

After editing your date range click apply to filter your image collection.

Filter date options for Google Earth engine application

3) Select visualization

The select visualization gives you the option to select a vegetation index to know your area better. The layer is added under the classified layer. You can uncheck the classified layer to see the vegetation index layer.

Vegetation indexes for visualizaiton

Visualize the vegetation layer by unchecking the classified layer

4) Start an export

These two buttons will create a download link to download the classified image or vegetation indices.

Download image from Google Earth Engine Application

In the map, some additional features will help the user to interpret the result much better. A bar chart will be generated every time the filter options are changed in the top-right corner. This will show the square kilometre of each class.

Bare chart of the classified image

A legend in the bottom-right corner will show the classes of the classified layer.

By clicking any pixel in the map, a line chart will be generated in the middle to show how the land cover in that particular pixel has changed over that particular year.

Line chart of the land cover change over the year

By refreshing the application every parameter will be reset to its pioneer stage.